Public Disclosure Authorized

WORLD BANK DISCUSSION PAPER NO. 365

Work

for
in

public
progress

discussion
WD P365

In IIovatiOnS ill I7lICltlhCare

Public Disclosure Authorized

"i iILicI i g
P/'o,(f
, /i,/Is (/,X/ It //,/ IP'k (Clfi'T/'I '''
Public Disclosure Authorized

(U ! I. /(L //! /(y

Public Disclosure Authorized
 Recent WorldBank Discussion Papers
No. 296 Reforming the Energy Sector in Transition Economies: Selected Experience and Lessons. Dale Gray
No. 297 Assessing Sector Instititions: Lessons of Experiencefrom Zambia's Education Sector. Rogerio F. Pinto and Angelous J.
Mrope
No. 298 Uganda's AIDS Crisis: Its Implications for Development. Jill Armstrong
No. 299 Towards a Payments System Lawfor Developing and Transition Economies. Raj Bhala
No. 300 Africa Can Compete! Export Opportunities and Challengesfor Garments and Home Prodiuctsin the EluropeanMarket.
Tyler Biggs, Margaret Miller, Caroline Otto, and Gerald Tyler
No. 301 Review and Outtlookfor the World Oil Market. Shane S. Streifel
No. 302 The Broad Sector Approach to Investment Lending: Sector Investment Programs. Peter Harrold and Associates
No. 303 Instituitional Adjustment and Adjuisting to Institutions. Robert Klitgaard
No. 304 Putting Instituitional Economics to Work: From Participation to Governance. Robert Picciotto
No. 305 Pakistan's PutblicAgricutltutralEnterprises: Inefficiencies, Market Distortions, and Proposalsfor Reform. Rashid Faruqee,
Ridwan Ali, and Yusuf Choudhry
No. 306 Grameen Bank: Performance and Stability. Shahidur R. Khandker, Baqui Khalily, and Zahed Khan
No. 307 The Urtguay Rouind and the Developing Economies. Edited by Will Martin and L. Alan Winters
No. 308 Bank Governance Contracts: Establishing Goals and Accountability in Bank Restructuring. Richard P. Roulier
No. 309 Public and Private Secondary Education in Developing Coluntries:A Comparative Stuldy. Emmanuel Jimenez and
Marlaine E. Lockheed with contributions by Donald Cox, Eduardo Luna, Vicente Paqueo, M. L. de Vera, and
Nongnuch Wattanawaha
No. 310 Practical Lessonsfor Africafrom East Asia in Indutstrialand Trade Policies. Peter Harrold, Malathi Jayawickrama, and
Deepak Bhattasali
No. 311 The Impact of the Urugutay Round on Africa. Peter Harrold
No. 312 Procurement and Disbursement Manuialfor Projects with ComnmuznityParticipation. Gita Gopal
No. 313 Harnessing Informationfor Development: A Proposalfor a World Bank Group Strategy. Eduardo Talero and Philip
Gaudette
No. 314 Colombia's Pension Reform: Fiscal and Macroeconomic Effects. Klaus Schmidt-Hebbel
No. 315 Land Quiality Indicators. Christian Pieri, Julian Dumanski, Ann Hamblin, and Anthony Young
No. 316 Suzstainability of a Government Targeted Credit Program: Evidencefrom Bangladesh. Shahidur R. Khandker, Zahed
Khan, and Baqui Khalily
No. 317 Selected Social Safety Net Programs in the Philippines: Targeting, Cost-Effectiveness, and Optionsfor Reform. Kalanidhi
Subbarao, Akhter U. Ahmed, and Tesfaye Teklu
No. 318 Private Sector Development Duiring Transition: The Visegrad Couintries.Michael S. Borish and Michel Noel
No. 319 Eduication Achievements and School Efficiency in Rural Bangladesh. Shahidur R. Khandker
No. 320 Houiseholdand IntraholuseholdImpacts of the Grameen Bank and Similar Targeted Credit Programs in Bangladesh. Mark
M. Pitt and Shahidur R. Khandker
No. 321 Clearance and Settlement Systemsfor Secufrities:Critical Design Choicesin Emerging Market Economies. Jeff Stehm
No. 322 Selecting Development Projectsfor the World Bank. Jean Baneth
No. 323 Evalutating Public Spending: A Frameworkfor PutblicExpenditutre Reviews. Sanjay Pradhan
No. 324 The Bangladesh Rural Advancement Committee's Credit Programs: Performance and Suistainability. Shahidur R. Khand-
ker and Baqui Khalily
No. 325 Instituitional and EntrepreneutrialLeadership in the Brazilian Science and Technology Sector:Setting a New Agenda .
Edited by Lauritz Holm-Nielsen, Michael Crawford, and Alcyone Saliba
No. 326 The East Asian Miracle and Information Technology:Strategic Management of TechnologicalLearning. Nagy Hanna,
Sandor Boyson, and Shakuntala Gunaratne
No. 327 Agricultzural Reform in Ruissia:A Viewfrom the Farm Level. Karen Brooks, Elmira Krylatykh, Zvi Lerman,
Aleksandr Petrikov, and Vasilii Uzun
No. 328 Insuiring Sovereign Debt Against Default. David F. Babbel
No. 329 Managing Transboutndanj Stocks of Small Pelagic Fish: Problems and Options. Max Aguiero and Exequiel Gonzalez
No. 330 China: Issutesand Options in GreenhoutseGas Emissions Control. Edited by Todd M. Johnson, Junfeng Li,
Zhongxiao Jiang, and Robert P. Taylor

(Continuiedon the inside back cover)

WORLD BANK DISCUSSION PAPER NO. 365

Innovations in Health
Care Financing
Proceedings
of a WorldBank Conference,
March 10-11, 1997

Editedby
GeorgeJ. Schieber
Tkt WorldBank
Washington,
D.C.
Copyright 1997
The International Bank for Reconstruction
and Development/THE WORLD BANK
1818H Street, N.W.
Washington, D.C. 20433,U.S.A.

All rights reserved

Manufactured in the United States of America
First printing July 1997

Discussion Papers present results of country analysis or research that are circulated to encourage
discussion and comment within the development community. To present these results with the least
possible delay, the typescript of this paper has not been prepared in accordance with the procedures
appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. Some sources
cited in this paper may be informal documents that are not readily available.
The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s)
and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to
members of its Board of Executive Directors or the countries they represent. The World Bank does not
guarantee the accuracy of the data included in this publication and accepts no responsibility whatsoever
for any consequence of their use. The boundaries, colors, denominations, and other information shown on
any map in this volume do not imply on the part of the World Bank Group any judgment on the legal
status of any territory or the endorsement or acceptance of such boundaries.
The material in this publication is copyrighted. Requests for pernission to reproduce portions of it
should be sent to the Office of the Publisher at the address shown in the copyright notice above. The
World Bank encourages dissemination of its work and will normally give permission promptly and, when
the reproduction is for noncommercial purposes, without asking a fee. Permission to copy portions for
classroom use is granted through the Copyright Clearance Center, Inc., Suite 910,222 Rosewood Drive,
Danvers, Massachusetts 01923,U.S.A.
ISSN:0259-210X

George J. Schieber is health sector leader in the World Bank's Middle East and North Africa Region.

Libraryof Congress Cataloging-in-Publication Data

Innovations in health care financing: proceedings of a World Bank conference, March 10-11, 1997 / edited
by George Schieber.
p. cm. - (World Bank discussion paper; ISSN 0259-210X; 365)
ISBN 0-8213-3964-8
1. Medical care-Developing countries-Finance-Congresses
I. Schieber, George. II. World Bank Conference on Innovations in Health Care Financing (1997:
Washington, D.C.) m. Series: World Bank discussion papers; 365.
RA410.55.D48I55 1997
338.4'33621'091724-dc2l 97-17183
CIP
Contents

Foreword v

Acknowledgments vi

Abstract vii

A Curmudgeon's Guide to Financing Health Care in Developing Countries 1

GeorgeSchieberand Akiko Maeda

Government Financing of Health Care 39

Bengt J5nssonand Philip Musgrove

From Beveridge to Bismarck: Health Finance in the Russian Federation 65

Igor Sheiman

Private Insurance: Principles and Practice 77

Deborah . Chollet and MaureenLewis

Private Health Insurance in Egypt 115

Nadwa Rafeh

Strategies for Pricing Publicly Provided Health Services 127

PaulJ. Gertler and Jeffrey S. Hammer

Cost Recovery Strategies in Sub-Saharan Africa 155

JosephWang'ombe

Rural Risk-Sharing Strategies 163

Andrew Creeseand SaraBennett

Rural Health Care Financing in Thailand 183

SirilaksanaKhoman

'iii
Market-BasedReform of U.S. Health Care Financing and Delivery:
Managed Care and Managed Competition 195
Alain C. Enthoven

Managed Care and Managed Competition in Latin Americaand the Caribbean 215
AndreCezarMedici,JuanLuisLondonio,OswaldoCoelho,andHelenSaxenian

Medical SavingsAccounts for Developing Countries 233

Len M. Nichols,NicholasPrescott,and KaiHongPhua

Medical SavingsAccounts and Health Care Financing in Singapore 247

Kai HongPhua

INNOVATIONS IN HEALTH CARE FINANCING

iv
Foreword
F inancing health care is a critical concern for rich
and poor countries alike, as health care systems
account for 9 percent of global production.
Developingcountriesfaceparticularlyseriouschallengesas
theyattemptto improvethe well-beingoftheir populations,
achieve economicdevelopmentobjectives,and integrate
themselveswith the globaleconomy.Health care financing
is a particularconcernfor these countries,which account
for 84 percentof the world'spopulationand 93 percent of
its diseaseburden but only 18percent of its incomeand 11
percent of its health expenditures.Imbalancesbetween
spendingand the diseaseburden will be exacerbatedas a
result of the changingcompositionof illnesstoward non-
communicablediseases and injuries,whichby 2020 will
account for almost 80 percent of these countries' disease
burdens, comparedwith just over 50 percent now.These
diseasesare more expensiveto treat and harder to prevent
than the infectiousdiseasesthat were previouslythe lead-
ing causesof illnessand death.
This volumecontains the thirteen paperspresented at
theWorldBank'sConferenceonInnovationsin HealthCare
Financing,heldinWashington, D.C., on March10-11,1997.
The conferencebroughttogether400participantsfrommore
v
than70 countriesto considerthe broadrangeof issuesrelat-
ing to the financingof healthcare systemsin low-and mid-
dle-incomecountries.Both the conceptualand operational
policycontextsforintroducingchangesin healthcarefinanc-
ingwere explored.Traditionalpublic and privatefinancing
approacheswereaddressedtogetherwithmorerecentmeth-
ods such as medicalsavingsaccountsand managedcompe-
tition.Issuesofparticularrelevanceto low-income countries,
includinguserchargesandinfonnalruralrisk-pooling schemes,
werealsodiscussed.Casestudiesfromeachdeveloping region
wereused to highlightthe variousapproaches.
It is hoped that this volumewillhelp countriesdevelop
effectivehealth care financingpolicies,and stimulatefur-
ther policydialogueand researchon this criticallyimpor-
tant socialand economicissue.
Davidde Ferranti
DirectorandHead
HumanDevelopmentNetwork
RichardG. A. Feachem
Director
Health,Nutrition,and Population
Abstract
T hisvolumecontainsthirteenpaperspreparedfor the
WorldBankConferenceon Innovationsin Health
CareFinancing,heldinWashington,D.C.,on March
10-11, 1997.Togetherthesepapersprovideconceptualand
practicalpolicyprescriptionsforfinancinghealthcaresystems
in developingcountries.Healthcarefinancingisa seriouscon-
cemforthesecountries:theycontain84percentofthe world's
populationand 93percentof its diseaseburdenbut account
forjust18percentof itsincomeand11percentofglobalhealth
spending.Moreover,developingcountrieshavethe capacity
to raiseless than60 percentof the revenuesraisedby indus-
trialcountries.With5 percentof theirgrossdomesticprod-
uctsdevotedto healthspending,splitalmostevenlybetween
publicand privatesourcesof spending,raisingandmanaging
health sectorrevenuesis a major challenge.The papers in
thisvolumeprovidean overviewof the healthcarefinancing
issuesthat aremostrelevantfordevelopingcountriesandpre-
sents casestudiesillustratingtheir experienceswithvarious
revenuegenerationand managementtechniques.
TheoverviewpaperbyGeorge SchieberandAkikoMaeda
describesdemographic, epidemiological,servicedelivery,
and
health expenditurepatterns for the world'sindustrialand
developingcountries,analyzesfrom both conceptualper-
spectivesand real-worldexperiencesthe rangeofpublicand
private revenue-raisinginstruments,and providesgeneral
perspectives forhealthfinancingreformsin eachoftheworld's
sixdevelopingregions.BengtJonssonand PhilipMusgrove
analyzethe issuesand experienceswith governmentfinanc-
ingofhealthservicesin both industrialand developingcoun-
tries.Theseissuesarehighlighted in IgorSheiman'scasestudy
of recenthealthinsurancereformin the RussianFederation.
DeborahJ.CholletandMaureenLewisdiscussprivatehealth
insuranceasa mechanism forfinancinghealthservices,empha-
vi
sizing underlyingcharacteristicsof insurance markets as
wellas the need for regulation.Nadwa Rafehpresents an
exampleofthe evolutionof privatehealthinsurancein Egypt.
Tworevenue-raisingmethods of particularimportance
to developingcountriesare user chargesand informalrural
risk-poolingarrangements.Paul J. Gertler and Jeffrey S.
Hammer analyzeuser chargesand reviewtheir effects on
revenue raising, equity,and efficiencyfrom conceptual,
empirical,and country perspectives.Joseph Wang'ombe
summarizes Sub-Saharan Africa's experience with user
charges.AndrewCreeseand SaraBennett discussthe con-
ceptualunderpinningsof and real-worldexperienceswith
informal rural risk-sharing arrangements. Sirilaksana
Khomanpresentsan exampleof one ofthese arrangements,
the health card used in Thailand.
Howrevenuesaremanagedhasimportantimplicationsfor
the efficiencywithwhichsuchrevenuesare used.Byprovid-
ingindividualswith strongincentivesto managehealthcare
funds,managedcompetitionand medicalsavingsaccounts
are tworecentinnovationsthat haveimportantdemand-side
effects.AlainC. Enthovendescribesthe basisfor managed
competition,discusseshow managedcare is a logicalcon-
comitantto managedcompetition,andanalyzesthe applica-
bilityofthese methodsto developingcountries.AndreCezar
Medici,Juan Luis Londoflo,OswaldoCoelho, and Helen
Saxeniandescribethe experiences withmanagedcompetition
and managedcare in LatinAmericaand the Caribbean.Len
Nichols,NicholasPrescott,and KaiHong Phua discussthe
conceptualandoperationalbasesformedicalsavingsaccounts,
describereal-worldexperienceswiththeseaccounts,andana-
lyzethenecessaryconditionsforimplementingtheseaccounts
in developingcountries.Phua alsoprovidesanin-depthanaly-
sisof Singapore'sexperiencewith medicalsavingsaccounts.
Acknowledgments
T he papers presentedin thisvolumewere commis-
sioned for the World Bank Conference on
Innovations in Health Care Financing, held in
Washington,D.C., on March 10-11, 1997.Specialthanks
are due to the Bank's conference cosponsors: the
CommonwealthFund, the U.S. Agencyfor Health Care
Policy and Research,the U.S. Agencyfor International
Development'sBureauforEuropeandNewlyIndependent
States,the U.S. Health CareFinancingAdministration,the
WorldBank's EconomicDevelopmentInstitute, and the
WorldHealth Organization.
The conferencewas initiatedby Armeane Choksi,for-
mer vicepresidentforHuman ResourceDevelopmentand
OperationsPolicy;David de Ferranti, director and head,
Human Development Network; and RichardFeachem,
director,Health, Nutrition, andPopulation.Their support
wasinstrumentalin ensuringhigh-quality, relevantpapers.
Many Bank staff contributed to the thematic content of
vii
the papersand helped ensuretheir relevancefor develop-
ingcountries.TheeffortsofXavierColl,EdwardElmendorf,
Theresa Ho, Eva Jarawan, Maureen Lewis, Sandy
Lieberman,Jo Martins,and NicholasPrescott are greatly
appreciated.
Specialthanks are due to the WorldBank staff mem-
bers who were peer reviewers for the papers: Shanta
Devarajan, David Dunlop, Charles Griffin, John
Langenbrunner, Chris Lovelace, William McGreevey,
AlexanderPreker, and Jacques van der Gaag. The Alpha
Center helpedthe Bank organizethe conference,and their
logisticand substantivesupport in meetingdeadlinesand
ensuringthe quality of the papers is greatly appreciated,
withspecialthanksto AmyBernsteinandDeborahChollet.
JillianCohen of the World Bank's Human Development
Department coordinatedthe peer reviewprocess.Finally,
the paperswereedited byPaul Holtz and laidout by Glenn
McGrath,both withAmericanWritingCorporation.
A Curmudgeon's Guide to Financing
Health Care in Developing Countries
GeorgeSchieberandAkiko Maeda
U nderstandinghow countriesfinance their health
care systemsis of critical importancefor indus-
trialand developingcountriesalike.The methods
used to mobilizethe resources that support basic public
health programs,provide accessto basic health services,
and configurehealth servicedeliverysystemsaffect peo-
ple's health status-as well as everyaspect of a country's
social,economic,andpoliticalwell-being(box1).Moreover,
health care systemsaccount for 9 percent of global pro-
duction and a significantportion of global employment.
Health caresystemsalsoaffectimportsand, in somecoun-
tries,exports.
Decisionson the methodsusedto raiserevenueforhealth
caresystemshaveimportantconsequencesforequityacross
income groups, the amounts of revenue raised, and the
lossesin consumerwelfareand production generated by
differentrevenue-raisingtechniques.Thus public and pri-
vate programs to finance and deliver health care affect
governmentbudgets,macroeconomic stability,employment,
imports,exports,and internationalcompetitiveness.
This paper focuseson howgovernmentsraiserevenues
to financetheirhealthcaresystemsaswellas on recentinno-
vationsfor public and private managementof these rev-
enues (includingmanagedcompetition, medicalsavings
accounts,private insurance,and communityrisk-pooling
schemes).It also discussesthe rationalesfor public and
privatefinance,assessesthe criteriathat shouldbe used to
evaluate differentrevenue sources,and identifiesappro-
GeorgeSchieberis healthsectorleaderin the MiddleEastandNorthAfricaRegionat theWorldBank.AkikoMaedais healthecon-
omistat theWorldBank.Thedatapresentedin thispaperarebasedon a revisedversionof theWorldBankHealthDataBase;assuch,
someof the statisticsdifferfromthosein the conferenceversionof thispaper.Theauthorsare gratefulto DeborahChollet,Shanta
Devarajan,NicoleKlingen, MaureenLewis, ChrisLovelace,
BillMcGreevey, LenNichols,
andJacquesvanderGaagforhelpfulcomments.
priate financingarrangementsfor developingcountries-
recognizingtheir underlyingeconomic and institutional
structures.
Conditionsin developingcountriesoften precludeuse
of the financingand managementarrangementsused in
industrialcountries.Thus this paper also drawsattention
to the costs associatedwith generatingpublic revenues-
coststhat usuallyfarexceedthe revenuethat is raised.These
costs, alongwith their distributionacross incomegroups,
are often overlookedin discussionsof health care financ-
ing.Yetinefficienciesand inequitiesin generatingrevenues
often compoundinefficienciesand inequitiesin allocating
expenditures.Suchissuesare ofcrucialimportancefordevel-
opingcountries,wherelow,incomelevelslimitthe scopefor
raisingrevenue.
Givenits emphasison financing,this paper focuseson
economicand administrativeissues.But noneconomiccon-
siderations, particularlyeach country's social, political,
and cultural environment,are of critical importance.An
attempt is made to capture these elements in the policy
and technicaldiscussionsbelow,particularlyin the discus-
sionson taxation and on the public-privatemix of financ-
ing,as wellas in the summariesof the healthcare systems
in each region.Still,since these discussionsare fairlygen-
eral,theymayunderemphasizethe importanceof noneco-
nomicfactorsin healthpolicymaking.
Althoughthispaperfocusesonsourcesof financing,such
discussionscannottakeplacewithoutconsideringthe demo-
AlexPreker,Nicholas Prescott,GailRichardson,
1
 tries giventheir institutionalcharacteristics.Sectionsixclis-
BOXI cussesoptions for private financing,with an emphasison
Goalsof a healthcaresystem
privatehealthinsurancemarketsand their implicationsfor
* Improvinga population's
healthstatusandpromoting social governmentregulation.Finally,sectionsevenprovidescon-
well-being cludingobservationson health care reformdebates in dif-
* Ensuringequityandaccessto care ferent parts of the developingworld.
* Ensuringmicroeconomic andmacroeconomic efficiency
in
the useofresources
* Enhancingclinical
effectiveness GlobalOverview
* Improvingquality
ofcareandconsumer satisfaction
* Assuring
the system's
long-run financial
sustainability Althoughthis paper'sfocusis on the sourcesof healthcare
financing,this sectionsummarizescurrent trends in such
financing,includingits relationto servicedeliveryoutputs
graphic,epidemiological, and servicedeliverycharacteristics and healthoutcomesbyregion.A comparisonof economic
of differentcountries.The next sectionsummarizesdemo- indicators, health outcomes, and health services across
graphicandepidemiological conditionsin developingregions. regionsand incomegroupsis shownin table 1.
Thethirdsectionanalyzesregionalhealthspendingin terms In 1994globalspendingon healthtotaled$2,330billion,
ofincomelevels,totalspending,andpublicandprivateshares or about 9 percent ofglobalincome(figure1).Of this, high-
of that spending.Thefourthsectiondiscussesthemainissues income countries (per capita income above $8,500)
concerningthepublic-privatenixofspending,the rationales accounted for justover $2,000billion-89 percent of the
forpublicandprivatefinancing,andthe advantagesandmar- total health expenditure.The populationsof these coun-
ket failuresassociatedwith financinghealth care through tries,however,accountedfor just 16 percent of the global
insurancemechanisms. Thefifthsectionprovidesanoverview population(figure2). The extreme disparitybetween the
of the differentsourcesfor publicfinancingof healthcare, amountof resourceslow-andmiddle-incomecountriesand
evaluatesthesesourcesin termsofeconomicefficiency, equity, high-incomecountries devote to health care reflects the
and administrativefeasibility,and discusseswhichpublic widelyvaryingcapacitiesofthese countrygroupsto provide
financingsourcesaremost appropriatefor developingcoun- healthservices.

TABLE
I
Economic
and healthindicators
byregionand incomegroup,circa1994
Economicindicators Healthoutcomes Health services
Percapita
Percapita GDP growth, Under-five Adult mortality, Hospital
GDP 1996-2005 mortality ages 15-60 Physicians
per bedsper
Region/income
group (1994 US$) (percent) (percent)a (percent)a 1,000people 1,000people
EastAsiaandthe Pacific 1,214 6.8 5.3 17.9 0.3 1.63
EuropeandCentralAsia 1,792 3.7 3.5 20.3 3.4 7.14
LatinAmericaandthe Caribbean 3,138 2.2 4.7 14.8 1.0 1.45
MiddleEastandNorth Africa 2,699 0.4 7.2 19.4 0.9 1.51
SouthAsia 440 3.7 10.6 23.5 0.2 0.53
Sub-SaharanAfrca 776 0.9 15.7 39.7 0.1 1.35
Low income 396 - 10.4 - - 0.87
Middleincome 2,707 - 5.3 - - 2.12
Low and middleincome 1,774 3.7 8.8 21.4 0.7 1.05
High income 18,611 2.4 0.9 9.7 2.5 6.29
Note:Regional
figuresarecountry-weighted
averages.
Income
groups
arebased
on 1994GDPpercapita:
lowincome
is$725or less,middleincome
is$726-8,500,
andhigh
incomeis$8,501or more.
a. Based
oncurrentlifetables.
Source:
WorldBank1996a and1997;
WorldBankdata.

INNOVATIONS
IN HEALTHCAREFINANCING
2
 The gapbetweenrichandpoor nationsis evenmoredra-
maticwhenthe distributionof the globaldiseaseburden is
considered.Of the estimated1.4trilliondisability-adjusted
life-years (DALYs)lost in 1990, industrial countries
accounted for just 7 percent (figure3). Of these, 81 per-
cent were attributable to noncommunicable diseases.
Developingcountries,whichaccounted for 93 percent of
the global diseaseburden, had a rather different disease
profile.Except for countriesin Europe and Central Asia,
whichhavedemographicand epidemiologicalprofilessim-
ilar to those in industrialnations,nearlyhalf of the DALYs
lost in developingcountrieswere causedby communicable
diseases,mainlyamongchildren.
Agingpopulationsand the risingincidenceof noncom-
municablediseaseswill continue to drive up the cost of
patientcare.In industrialcountriesa largeportionof health
spendingis used for a smallpercentageof patients in the
finalyearsof their lives.Most of these patients are suffer-
ingfromsomeformof noncommunicabledisease.Overthe
next three decades developingcountries will undergo a
majordemographicandepidemiological transition,with sig-
nificant increasesin the burden of injuries and noncom-
municablediseases (figure 4). These diseases are more
expensiveto treat andharder to prevent.Thistransitionwill
reorientdemandfor healthservicesand increasepressures
for new investmentin health care.
What are the prospects for narrowingthe disparities
between rich and poor nations? Some perspectiveon this
questioncanbe gainedbycomparingthe two groups'health
service capacities and prospects for economic growth.
Industrialcountrieshavethreetimesasmanyphysiciansper
capita and sixtimes as many inpatient beds per capita as
developingcountries (see table 1). To close the resource
gap,developingcountrieswillhaveto makesizableinvest-
mentsin healthservicesand increasespendingat ratesfaster
than those of high-incomecountries.

FIGUREI FIGURE2
Globaldistribution
of healthspending,1994 Globalpopulationand incomedistribution,1994
Globaldistribution Populationdistribution by region
LatinAmerica
andthe Caribbean

middle-income ad High income

11% G6o

89% Europeand
South~~~ ~outre Central Asia
18% _ca X
SubSahar.an
coutae
./_ Af
Totalglobal
health I / East 10%
expenditure:
$2,330 billio Middle Eas.n the Pacific
and North Afr 3ca

Low- and middle-income countries

SouthAsia
8% ~~~~~~~~~~~~~~~~Global
distribution of GDP
European
43~~~~~~~~~~~~~%
CentralZAiak

LatinAmnerica
M E W,\and the Caribbean
43%
Sub-Saharan
Afr'ica ~ Vihicm

6 \ countrie

Total healthexpenditure: MiddleEastand North Afnca

$250 billionl 9%Note: SouthAsiaincludesIndiaandEastAsiaandthe Pacific
includesChina.
Source:
World Bankdata. Source:
Wodd Bankdata.

A CURMUDGEON'S GUIDE TO FINANCING HEALTII CARE IN DEVELOPING COUNTRIES

3
FIGURE
3 FIGURE
Globaldiseaseburden,1990
life-yearslost)
(disability-adjusted
OECD countries
Comunicable
Total:99million
DALYA,
(7% of globaltotal)
Developing
countries
Total:1,280
million
DALYs (93%ofglobaltotal)
Source:
WHO i996b.
Such an adjustment maybe achievablein East Asia,where
per capita GDP is projected to grow by 6.8 percent a year
over the next decade. But in other regions-especially the
Middle East and North Africa and Sub-Saharan Africa-
annual economic growth rates will be less than 1 percent
(see table 1). These projections have particularly serious
consequences for Sub-Saharan Africa, where the base of
health infrastructure is already quite weak.
Strengthening health service capacities will require
expanding facilities and personnel as well as improving the
quality of services.Countries at similar income levels show
considerable variation in the performance of their health
systems-variation that can be partly ascribed to differences
in the equity, efficiency, and quality of health services. For
example, the average number of hospital beds and physi-
ciansper capita is higherin Europe and CentralAsiathan
in high-income countries (see table 1). Yet overutilization
and inappropriate clinicalinterventionsraise questions about
how effectivelythese resources are being used. Recent stud-
ies of OECD countries point to the importance of organi-
INNOVATIONS IN HEALTII CARE FINANCING
4
4
Changing
burdenof diseasepattem in developing
countries, 1990 and 2020
1990
Total: 1,280millionDAY
2020
Total:1,292million
DALYs
Source:
WHO I 996b.
zational arrangements in deterrmining the efficiency,qual-
ity, and equity of health delivery systems (OECD 1995).
Although there have been many studies of health deliv-
ery systems, less information is available on the inefficien-
cies and inequities associated with different health financing
systems. At least one study suggests that systems that rely
on social security financing might be more costly to admin-
ister than systems that rely on general revenue sources
(Poullier 1992). In developing countries the lack of infor-
mation on access to and distribution of services, utilization
rates (inpatient admission rates, physician visits per capita,
and so on), and quality measures have limited cross-coun-
try comparisons of what a dollar's worth of health expen-
diture buys in terms of effective health services.
Health Care Financing and Spending
Patterns
Policymakers face the perpetual challenge of raising suffi-
cient revenue for the health sector in an equitable and effi-
cient way.Althoughmost countriesrecognizethat health
care is a right for all citizens-as embodied in the World
Health Organization's(WHO) goal of "Health for Allin
the Year2000"-there are no clear guidelineson howthis
objectivetranslatesintohealthservicedelivery,andwhether
suchservicesare affordable.
What resourcescan a countrywith a per capita income
of $400 expectto raisefor its healthsector,and what kind
of servicescan it providefor its citizens?The WorldBank's
WorldDevelopmentReport1993:Investingin Healthoffered
a normativeresponseto that questionusingthe conceptof
a minimalpackageof care and services.In principlesucha
packagewould cut the number of lost DALYsrelativeto
availableresourcesin countries at differentincomelevels
(WorldBank 1993).Althoughthis approachprovidesone
set of objectivecriteriafor rationinghealth care services,
decisionsabout health spendingcannotbe isolatedfrom a
country's social,political,and economiccharacteristics.
Moreover,factorsbeyondthe controlofpolicymakersoften
affectspending.Understandingthe interactionofthese fac-
tors is essentialto designinghealth policy.
Policymakersmust have some wayto evaluatethe per-
formanceof their country'shealth systemsagainstthose of
other countriesor regionsat comparableincomelevels.One
approachis to dividethe performanceof healthfinancing
mnechanisms intothree broadcategories(figure5).The first
categoryis concernedwith how efficientlyand equitably
revenuesare raised, and what effect they haveon the size
and distributionofthe resourcesavailableto the healthsec-
tor.The secondcategoryinvolvesevaluatinghowefficiently
and equitably resources are used to provide health ser-
vices.The third categoryrelatesto the effectshealthexpen-
ditureshave on health outcomes.This last measure is tied
occurred only sporadicallyduring the 1970s and 1980s
(WorldBank 1987).The firstsustainedandreplicableeffort
to developsuch an informationbase for the twenty-four
member countries of the Organization for Economic
Cooperation and Development(OECD) began in 1977,
andthe informationisnowupdatedeachyear(OECD 1977,
1985,and 1993).The experiencesof the OECD countries
led to the development of a system of national health
accounts,whichonlybegan to be usedin developingcoun-
tries in the past two or three years.1 WorldDevelopment
Report1993was the firstcomprehensiveeffort to system-
aticallydevelopexpenditureinformationfor alldeveloping
countries (WorldBank 1993).
Despitetheseefforts,healthexpendituredata for devel-
opingcountriesare often incompleteor unavailable,espe-
ciallyfor private spending.Definitionsof health spending
vary by country, and disaggregationof health spending
beyond the broad categoriesof public and private is even
more problematic.As a result developingcountries lack
the basicinformationand toolsneededto assesshowhealth
systemresourcesare being raised and used. Withoutsuch
information it is extremelydifficult for policymakersto
understand the effects of their policiesand to determine
whichdecisionsare likelyto ensureequityin financingand
increasereturns on the resources devoted to the health
sector.
Moreover,withoutsuchinformationit is difficultto gauge
the effectivenessof pastinvestmentsor to evaluatecurrent
investments.Indeed, the importanceof these data and the
Differentwaysof measuring
performance
in healthfinancing
In_healthfinancin

to intersectoralfactors such as education,water and sani- Revenue

generation
Health
services
output
Health
andoutcomes
status
tation, and women's status, since health servicesare just Keyissues: Keyissues: Keyissues:
Efriciency,
stability,
and Eficiency
inservice
output Intersectoraldeterminants
one factor among many that determine a population's health sustainability
in revenue perdollarinvested;access; of healthinccuding
medical
status.Thispaperfocuseson issuesrelatedto the firstgroup generation;
effective- quality;patient
satisfaction care,householdbehavior,
nessin riskpoolingand and patientchoice. women'sstatusandeduca-
of measures. redistribution. tion, waterand sanitation,
Internationalcomparisonsof healthfinancingare diff- Examples of instruments: environment, nutrition,
aExomples
ofinstruments:Unitcostof effective andlifestyle
changes.
cult, partly becauseof the lack of reliabledata. Efforts to Burdenof taxation, services bydifferent
distribution
of public facilities,
payment systems, Examplesof instruments:
compilecomprehensive cross-countrydataon healthexpen- subsidies,
extentof organizational andmanage- Mortalityandmorbidy rate
ditures date back to Brian Abel-Smith's work in the 1960s insurance
coverage mentstnrctures. reduction
(increase
in life
(formalandinformal). expectancy,
reduction
in
for the World Health Organization (see Abel-Smith 1967). infantmortality,lifeyears
Efforts to update that information for developing countries savedper dollar).

A CURMUDGEON'S GUIDE TO FINANCING HEALTEI CARE IN DEVELOPING COUNTRIES

5
difficultyin compilingthem stronglysuggestthe need for collected,about 37 percent of the privatehealth expendi-
direct,systematic,andregularcollectionofhealthsystemdata ture data wereimputedfroma regressionmodel,compared
fromalldevelopingcountries, possiblybyadaptingthe national with 17 percent for public healthexpenditures.
healthaccountapproachused by OECD countries.Priority In this paperpublicexpenditureson healthreferto funds
shouldbe givento developingan accountingsystemthat is from governmentbudgets, compulsory(that is, publicly
affordableandeasyto use in the developingcountrycontext. mandated)healthinsurancefunds (socialsecurityschemes,
TheWorldBankandother intemationalorganizations could mutualfunds,sicknessfunds),and externalloansandgrants.2
supportthiseffortwith financialand technicalassistance. Privateexpendituresreferto directhouseholdexpenditures,
The WorldBankhas attempted to update globalhealth includingout-of-pocketpaymentsforservices,expenditures
expendituredata to 1994usingthe latestavailablesources through private health insurance plans, direct payments
(actual data range from 1990 to 1995). Data have been for healthservicesby firms and corporations,and charita-
collectedon total,private,and publichealthexpenditures, ble contributions.
and limited informationhas been obtainedon sourcesof
financing(socialhealthinsurance,donors) andtypesof ser- Regionalhealth expenditure patterns
vices(hospitals,pharnaceuticals).The followingdiscussion
is based on preliminaryanalysesof this database.Although Averageper capita healthexpendituresrange from $16 in
effortshavebeenmadeto ensurecomparabilityacrosscoun- low-incomecountries to $1,827 in OECD countries-a
tries,the analysisshouldbe interpreted withcaution.Data hundredfold difference (table 2). OECD countries also
were obtained from various sources (public expenditure spend moreon healthas a percentageof GDP Low-income
reviews,government budgets, household surveys,World countriesspend about 4 percent of GDP on health;OECD
Bank reports,sectorreports),so definitionsand collection countriesspend more than 8 percent.
methods varied.Moreover,data on private healthexpen- Amongregions,South Asiaspends the least on health
diture are evenmore proneto measurementerrors because as apercentageof GDP-less thanevenSub-SaharanAfrica.
of a lackof reliableinformationon householdsand private However,a significantportionofhealthcostsin Sub-Saharan
enterprises. Of the 202 economiesfor which data were countriesare financedbyexternalsources(seebelow).Other
TABLE
2
byregionandincomegroup,circa1994
PercapitaGDPand healthexpenditures
Publichealth
Per capita expenditure
Per capitaGDP healthexpenditure Healthexpenditure as a shareof total
aspercentage healthexpenditure
group
Region/income PPP$ US$ PPP$ US$ of GDP (percent)
EastAsiaandthePacific 4,554 1,214 158 38 4.1 52
EuropeandCentralAsia 3,847 1,792 346 154 7.2 72
LatinAmericanandthe Caribbean 5,729 3,138 367 200 6.1 49
MiddleEastand North Africa 7,181 2,699 353 116 5.2 50
SouthAsiab 1,887 440 65 12 3.7 39
Sub-SaharanAfrica 2,070 776 III 38 4.0 54
Low income 1,565 396 71 16 4.3 47
Middleincome 5,790 2,707 364 168 5.3 57
Highincome 20,615 18,611 1,521 1,468 6.9 67
OECDc 21,169 22,498 1,777 1,827 8.3 76
Note:Regional
fgLuresarecountry-weightedaverages.
International
dollars
(PPP$)
arelocalcurrencies
converted
to U.S.dollars theuseof purchasing
through powerparities
('exchange
rates'thatadjust
forcostdifferences
across
countries).
Income groups
arebased on 1994GDPpercapita:
lowincomeis$725or less,middleincomeis$726-8,500,
andhighincome is$8,501or more.
a. Indudes
China.
b. Indudes
India.
c. Exdudes
Hungary, Mexico,andTurkey.
Source:
WorldBankdata.

INNOVATIONS
IN HEALTHCAREFINANCING
6
regionaldifferencescannot be explainedby differencesin
per capita income alone.For example,althoughcountries
in Europe and CentralAsiahavethe third-highestincome
levelamonglow-and middle-incomecountries,theyspend
more than 7-percentof GDP on health, the highestamong
this group of countries.
Per capita incomesand the public share of health care
costs tend to rise together, indicating an expanding gov-
ernmentrole in health care financingas countriesdevelop
economically(figure 6). In OECD countries the public
sector accounts for, on average,more than 75 percent of
total health spending. Developing countries show consid-
erable variation in the public share of health spending.
This heterogeneity underscores the diversity of approaches
to health care financingin developing countries and reflects
these countries' historical, political, and economic struc-
tures. For example, the large public share of health spend-
ing in Europe and Central Asiais a legacyof highlycentralized
government structures under socialism.
Composition
of publicexpenditures
As noted, data on public health spending have been drawn
from government budgets, compulsory health insurance
funds, and external loans and grants. Finding a consistent
definition of compulsory health insurance funds is some-
what problematic; thus the data on this component of health
expenditures should be interpreted with caution. Still,some
interesting trends emerge.
Public insurance schemes play a limited role in most low-
income countries, where public health expenditures usually
come directly from government budgets. For example, in
the four Sub-Saharan African countries that reported on
public health insurance, the schemes accounted for 1-9 per-
cent of total health expenditure. In India they accounted
for just over 1 percent. Yet in China the Government
Insurance System accounted for nearly 30 percent of health
expenditures in 1993.
Social insurance schemes play a larger role in middle-
income countries. Again, though, the pattern that emerges
is fairly diverse. Some countries continue to rely on-govem-
ment budgets to finance public health systems,while others
are moving toward payroll tax-based insurance schemes. In
Latin America and the Caribbean and Europe and Central
A CURMUDGEON'S GUIDE TO FINANCING HEALTH CARE IN DEVELOPING COUNTRIES
FIGURE6
Publicshareof healthexpenditures
and percapita
GDP,variouscountries,circa1994
Publicshare(percent)
'00
. 1It*
60
40
20 e
___
100 o0,000
1.000 100.000
Per capitaGDP, US$ (log scale)
Source:
Wodd Bankdata.
Asia,for example, socialinsurance funds finance a significant
portion of health expenditures. In 1994socialinsurance costs
in the fifteen (of thirty-four)Latin American and Caribbean
countriesthat reported suchinformationrangedfrom 7 to
60 percent of total health expenditures (PAHO 1996). In
Europe and Central Asia social insurance funds accounted
for 20 to 80 percent of health expenditures. But in the Middle
East and North Africa just seven (of nineteen) countries
reported having compulsory insurance schemes, which
financed between 9 and 37 percent of health expenditures.
(Mosthigh-incomeoil-exportingGulf countriesfinancehealth
servicesdirectly from government budgets.)
External assistance continues to be an important source
of financingfor health servicesin low-incomecountries, espe-
ciallyinSub-SaharanAfricaand SouthAsia (exdudingIndia).
Since the mid-1980s, however,the share of bilateral official
development assistance (ODA) allocated to health has been
declining,while ODA from multilateral sources-including
the World Bank-has been increasing (Michaud and Murray
1996, p. 230). The tightening of ODA resources makes it
hard to predict how ODA to the health sector will evolve
over the next decade. For many middle-income countries
external assistance may fall, with the emphasis of aid shift-
ing from financial assistance to technical assistance. Yet
low-incomecountries,especiallythosein Sub-SaharanAfrica,
will likely continue to depend on external assistance for at
least the next decade, and possibly longer (table 3).
7
Composition of private expenditures
Data on privatehealth spendingare evenharderto obtain
than data on public spending.Most of the data on private
spendingare estimates drawn from household expendi-
ture surveys.Thesedata usuallyare not disaggregatedinto
differentformsof payment,such as directfees for service,
insurance premiumsor other forms of prepayment, and
cost-sharingpayments.And only a few countries report
expendituresfrom private insuranceschemes and direct
corporatespendingonhealth services,whichoften account
for a significantportionof health expenditures,especially
in low-income countries.Thesecategoriesshouldbe included
in future data collection efforts.
Incomeelasticities
provide benefits or externalitiesthat extend beyond the
individual)and personalhealt services(whichbenefit only
the individual;see Musgrove1996).Funds can be raised
(or derived)throughtaxes,mandates,privatehealth insur-
ance,directprivateout-of-pocketpayments(includinguser
charges for publiclyprovided services),grant assistance,
charitablecontributions,and domesticor foreignborrow-
ing. They can be managedby government (ministriesof
health)or quasi-goverrmentagencies(socialsecurityorga-
nizations,sicknessfunds), for-profitor nonprofitprivate
entities(privateinsurers,purchasingcooperatives,employ-
TABLE
3
External assistancefor health costsby region, circa
1990
(percent)

Extemalassistance
Income elasticitiesprovide a useful measure of how dif- asa shareof percapita
ferencesin countries' income levelstranslate into differ- Region healthexpenditure
ences in health expenditures (table 4 and figure 7). The EastAsiaandthePacific
(exduding
China) 3.7
globalelasticityis estimatedat 1.13.Thus for every10per- LatinAmericanandthe Caribbean 3.6
cent differencein per capitaincomethereis a 11.3percent MiddleEastand NorthAfrica 1.5
difference in per capita health expenditures-that is, coun- SouthAsia(excluding
India) 13.1
India 0.7
tries with higher incomes tend to devote a larger share of Africa
Sub-Saharan 16.4
those incomes to health expenditures. The income elastic- Note: Regional
dataarecountry-weighted
averages.
Not enoughdatawere avaibble
ity for the public component of health expenditures is 1.21; for Europe Bank
andCenturlriat9p9r4pWord data.
for private spending it is 1.02. This pattern suggests that
public health spending is more responsive to income dif- TABLE4

ferences than is private health spending, and is consistent Income elasticitiesfor total, public, and private
with the fact that high-income countries have larger public health care spending,circa 1994
shares of total health expenditures. Income Numberof Adjusted
Income elasticities for countries by income level are Spendingcategory elasticity(il) observations R2
shown in table 5. Income elasticities for per capita health Total
healthexpenditure 1.13 122 0.94
expenditures relativeto per capita GDP are highestfor high- Public 1.21 162 0.91
income countries (1.47), followedby middle-income (1.19) Private 1.02 126 0.85
Note:Dependent
variableispercapita
heakhexpenditure
(US$).
and low-income (1.00) countries. This pattern is also con- WorldBankdata.
Source:

sistent with the fact that higher-income countries devote a

larger shre
totheof resurces
healt sector.TABLE 5
larger share of resources to the health sector. Incomeelasticitiesby incomegroup, circa 1994

The Public-PrivateFinancingMix Income Numberof Adjusted

Incomegroup elasticity
(il) observations R2
Healh care systems are financed by many sources, public Lowincome 1.00 31 0.34
Middleincome 1.19 57 0.82
and private. These funds are managed by public and pri- Highincome 1.47 34 0.64
vate entities and spent on bothpublic healthservices(which Source:World Bank
data.

INNOVATIONS
IN HEALTHCAREFINANCING
8
ers, unions), or consumers (out-of pocket payments, med-
ical savings accounts).3 Funds are then used to purchase
publicly or privately provided health services (figure 8).
The basic issues relating to the appropriateness of pub-
lic or private sources of finance are predicated on govern-
ments' allocational,distributional,stabilization,and economic
goals and on the policies that are used to correct for mar-
ket failures and externalities in the financing, consumption,
and provision of health services. Particularly relevant are
insurance market failures and instabilities, which may pre-
clude people from obtaining the benefits of collective risk
reduction through efficient insurance provision.
Several other potential market failures affect the health
sector (see Hsiao 1995; Musgrove 1996; Jonsson and
Musgrove in this volume). One important market failure
involves extemalities in consumption, whereby the collec-
tive benefits from consumption of health servicesare greater
than the individual benefits. Market failure also can pro-
vide a rationale for public financing because of the effects
ill health has on income redistribution, income levels, and
poverty. Other areas of market failure that have implica-
tions for public and private financing deal with informa-
tion gaps and asymmetries,interdependence between supply
and demand, and supply-side market failures.4
Health services with collective benefits
Certain health services-vector control, clean air and water,
sanitation systems, environmental health, medical research,
most health promotion and education activities-are purely
public goods. That is, no individual can be excluded from
the benefit, and consumption by one individual does not
reduce the amount availableto others. Other goods, known
as merit goods, benefit individuals as well as communities.
Immunizations and treatments for contagious diseases are
examples. Left to their own devices, most people will spend
on public and merit goods only up to the point at which
their private (marginal) benefit equals the private cost, and
society as a whole will underconsume these services.5 Thus,
in order to ensure an appropriate collective consumption
level, such services should be publicly financed (or subsi-
dized). Around the world, most of the health services that
are consumed are personal health services, for which ben-
efits accrue largelyor exclusively to individuals.
A CURMUDGEON'SGUIDE TO FINANCINGHEALTIICAREIN DEVELOPINGCOUNTRIES
FIGURE7
Per capita health spendingand GDP, various
countries,circa 1994
Percapitahealthexpenditure
(logscale)
0,000
1.000
100
10
100 1P000 10(000 100,000
WorldBank
Source: data.
Although the taxonomy of services described above is
used in this analysis,other categorizations of health services
are relevant for discussions of risk pooling and private and
public responsibilities. In particular, health services can be
classifiedas preventive,curative for unexpected health prob-
lems, curative for chronic predictable health problems, and
curative for lifestyle-induced health problems. Once these
serviceshave been defined, societies can then choose appro-
priate direct subsidies and cross-subsidies(through risk pool-
ing) to finance the costs and determine individual
responsibilitiesfor lifestylechoices-substance abuse,promis-
cuity, obesity-that adversely affect health (Nichols 1996).
World Development Report 1993 outlines how govern-
ments can invest scarce public funds in cost-effective basic
public health services. For $12-22 per capita developing
countries will get the best return on public health spend-
ing by investing in a basic package of public health and
essential clinicalservices, includingimmunizations, school-
based health services, programs to reduce alcohol and
tobacco consumption, familyplanning services, tuberculo-
sis control, control of sexuallytransmitted diseases, and care
for childhood illnesses such as acute respiratory infections,
diarrheal diseases, measles, malaria, and acute malnutrition
(WorldBank 1993). In many of the poorest countries, how-
ever, governments have not mobilized sufficient resources
for such a package of care, and in some cases both gov-
9
ernments and individualshave chosen to spend funds on
other, lesscost-effectiveservices(WorldBank 1994).
AlthoughWorldDevelopment Report1993does not pro-
vide definitiveanswerson howto maximizethe returns to
publicand privateexpendituresabovethisthreshold,many
developingcountriescould obtain a better return on their
public healthinvestmentsby providingthis basic package
of publichealth and clinicalservices.Many of these coun-
trieshaveinvestedsignificantpublicresourceson state-of-
the-art curativeservices,whichare often availableonlyto
the ruling elite,the politicallyconnected,or those able to
buy such servicesfrom the private sector.Althoughthere
is no universalprescriptionon the appropriatemixof pub-
licandprivatefinancingor the appropriateroleof the state
in financingpersonalhealthcare services(other than those
for the poor and other vulnerablegroups,whichare justi-
fied on equityand incomeredistributiongrounds),there is
considerablescopefordevelopingcountriesto improvetheir
health investmentsand outcomes.
Redistribution
In most countriesthe bulk of health expendituresare for
personal(individual)health services.Althoughin principle

FIGURE8
Sources,management, and provisionof health care financing
Revenue Managed Provided
source by by

~~~~~~Taxes
- - - - - - -- - - - - - - -o_
.~~~~~~~~ Government _.
i ~~~~~~~~~~~~~~~~~~~~~~agencies

~~~~User charges --- _ __.

Public Public
- ~~~~~~~~~~~~~~~~~~~~~~~~
l 0 . ~~~~~~~~~Socia]
insurance/
l l _ ~~~~~~~~~~sickness
funds

Mnates "L%.

Private
hearPnvate organizabons/
Gnsurance 1X.. insurers

S | ~~~Lc,ans s 1ll l

0 | | = | ~~~~~~~~~~~~~
ll ~~Employers
l

Private IN E' C NANCINt P vateNNOVATIOS

|Charitable
} a . l l

W
=8ffi=Ri2=.tt*'.=. .
|
.
~~~~~~~~~~~~Individuals
l .

Out-of-pocket|lll
expenses |I

I insurrc . . .. . . ....

IN HEALTHCARPFINANCING
INNOVATIONS
- ~ ~~~~~
10
these servicescan be financedprivately,almostal societies
view accessto health care as a basic human right. Thus
governmentsoftenprovideaccessto healthservicesforpeo-
ple whocannotaffordthem.Moreover,manygovernments
fund health insurance for vulnerable groups or provide
personalhealth carebecauseof the directlink betweenill-
ness,earningpotential,qualityof life, and poverty,aswell
as the randomnature of ilness.
Yetequity considerationscan generateinefficiencyand
market failure in private insurance markets and impose
significantcosts on government.Put another way,equity
concernsoften reflectjudgmentsabout tradeoffsbetween
desiredredistributionandthe distortedincentivesthat often
accompanyredistribution(Pauly1996).If consumerschoose
not to spend any moneyon health-relatedgoods and ser-
vices,knowingthat the statewillpayfor theirmedicalcosts
on equitygrounds,risk poolingthrough privatemarketsis
undermined,and the state maybe left fundingthe bulk of
society's health risks-with deleterious fiscaland redis-
tributiveimplicationsforboth the governmentand the pri-
vate sector. Moreover,such actionsreduce the potential
gainsfrom risk poolingand could result in excessivetaxa-
tion sothat governmentcanfinancemedicalexpenses(see
below). Thus societiesmust weigh the welfare costs of
individualsbeing unable to privatelypurchase "needed"
health servicesagainst the butdens of publiclyfinancing
suchneeds.
Market failuresin healthinsurance
Insuranceis prepaymentforservicesthat willbe paid forby
a (publicor private)third party (the insurer)should a pre-
definedeventoccur.6 Insuranceisa substitutefor(orin some
cases a complementto) direct out-of-pocketpaymentfor
such services.As discussedbelow,insurancereduces risks
bypoolingthem.Whetherpublicor private,insuranceaffects
the distributionof healthcare financingamonghouseholds
and can alsoaffectthe deliveryof healthservices.7
The potentialmarket failuresthat arisefrom the insta-
bilitiesinherentin insurancemarketsprovideanotherpoten-
tial rationalefor governmentfinancingof personalhealth
services(as well as for governmentregulationof private
insurance).Much of thisinstabilityoccursbecauseof indi-
vidualaversionto risk,uncertaintyaboutthe randomnature
A CURMUDGEON'S GUIDE TO FINANCING
of ilness and medicalexpenses,informationasymmetries
between insurers and consumers,adverse selection,and
moralhazard. To understandhow these factorscan cause
market failure,it is essentialto understandthe conceptual
basis for insurance.8
Rationalesfor insurance.Most people prefer to avoid
facingrisksthat result in substantialeconomiclosses.Thus
they are risk averseand, giventhe opportunity,willavoid,
minimize,or shift risksto others. They are willingto pay a
relativelysmallcertainpriceto avoida relativelylargeunpre-
dictableloss.The degree of risk facedby an individual(or
insurer)depends on the accuracywith whichthe probabil-
ity of the adverse event occurringcan be predicted.The
greater the uncertainty,the higherthe risk.
Insurancereduces risksby improvingthe predictabil-
ityof the adverseeventthroughthe poolingof a largenum-
ber of similarrisks.Froma financialperspective,insurance
is an arrangement that redistributes the costs of unex-
pected losses (Dorfman 1982, p. 5). From a legal per-
spective,insuranceis a contract in which the third party
agrees to compensate the subscriber for specific costs
incurred when a specific loss occurs (Hall 1994, p. 6).
Insurance has two components: the expected loss (total
loss incurred times the probabilityof the loss occurring)
and the risk premium and load factors (amount the sub-
scriber is willingto pay to avoid the expected loss and,
from the insurer's perspective,costs of marketing, prof-
its, and administration).
Insurancelowersrisksin the aggregateand makesthem
more predictablebecausepoolinga largenumber of simi-
lar eventsincreasesthe predictabilityof the event.9 Thus
the larger is the insurance pool for a particular risk, the
greater is the likelihoodof correctlyassessingthe proba-
bilityof the lossoccurring.Risksthat are unpredictablefor
the individualbecome predictablefor the group and can
be estimatedaccordingly.Since there are large economies
of scalein terms of both administration(load factor) and
improvedaccuracyof lossprediction,a risk becomeseas-
ier to insure as the insurance pool for that risk grows.
Conversely,risk premiumsand load factorsincreaseas the
pool shrinks.For example,privatehealthinsurancepremi-
ums in the United Statesare much higherfor smallgroups
and individualsthan theyare for largeemploymentgroups.
HEALTII CARE IN DEVELOPING COUNTRIES
11
Undercertaincircumstancesthe loadfactorsmaybe sohigh
that insurancemarketsare not viable.
Successfulinsurancesystemsshare severalcharacteris-
tics:
Most individualsare risk averseand prefer to substitute
a smalloutlay(theinsurancepremium)fora largeuncer-
tain loss.10
* Bypoolinga largenumber of similarrisks, insurersare
able to predict lossesaccuratelyand chargea premium
appropriateto that risk.
* Individualspooltheirpotentiallossesand payarelatively
smallpremiumfor the right to collectindividuallosses
from the pool.
* The premium is small relative to the potential loss
(Dorfman 1982,pp. 22-23).
Severalother supplyand demandfactorsaffectthe via-
bility of insurance markets. The main factors affecting
demand for insurance are the size and predictabilityof
risk. Peoplewill insure againstlarge unpredictablelosses
(Pauly1986).Peoplewillnot insureagainstlossesthat are
certain to occur, since in that case poolingdoes not pro-
vide anyadvantage.Similarly,whilein a perfectlyrational
worldmostindividualsshouldnot insureagainstsmalllosses
(sincetheycanself-insureagainstsucheventsand avoidthe
riskpremium),manyindividualswant to be insuredagainst
smallpredictablelossesas wellas largeones.11
An important distinctionregardinghealth care financ-
ing and the appropriateroles of the public and privatesec-
tors followsfrom these observations.In its purest form
insuranceisa mechanismthat reducesrisksbypoolingthem.
Insurancecanalsobe usedasa financingmechanismthrough
whichpublic or private entitiescollectpremiums(includ-
ing administrativecosts) to cover highlyprobable or com-
pletely predictable losses. Where losses are completely
predictable,insuranceprovidesno collectivebenefitin the
form of risk reduction. Still,there may be equity reasons
for a government to cover servicesin this manner, and
individualsmaydemandthat suchbenefitsbe includedin
basic insurancepackages.The inclusionof such benefits,
however,underminesinsurancemarkets(Hall 1994,p. 25;
Cholletand Lewisin this volume).
P'roblemsin
insurancemarkets.
Twosupply-sideaspectsof
insurance create major problemsfor insurancemarkets:
INNOVATIONS IN HEALTEI CARE FINANCING
adverse selection and moral hazard. Adverse selection,
alsoknownas biasedselection,occursbecauseof an infor-
mation asymmetrythat ariseswhen insurancesubscribers
have better informationabout their individualrisksthan
the insurer.As a result there is a higherthan averageprob-
abilityofthe adverseeventoccurring,sincepeopleare more
likelyto purchaseinsurancethat is offered at an actuarially
fair price for the entire community.Thus higher-riskindi-
vidualspayan averagepremiumthat is wellbelowwhat an
actuariallyappropriaterate for their risk group would be
(thatis,insuredsarenot chargedarate fortransferringtheir
exposureto loss that fairlyreflectsthe cost of the transfer;
see Dorfman 1982, p. 24). Such actions can destabilize
voluntaryinsurancemarkets,sincehealthierindividualswill
eventuallydrop out as premiumsrise, creatingyet higher
premiumsand further healthy dropouts. This phenome-
non is referredto as the premiumdeathspiral.
Thereare severalways to dealwithadverseselection.Since
adverseselectionoccurslargelyin voluntaryprivate insur-
ancemarkets,one solutionis to create a mandatorypublic
insurancesystem.Byrequiringeveryoneto join,the adverse
selectionproblemis elimninated. But so too is a greatdeal of
consumerchoice,whichresultsin a welfarelossto society.
Privateinsurersdealwithadverseselectionin threeways:
by obtaininginformationaboutthe underlyingmedicalrisks
of individualsubscribers,.bynot covering some of these
underlyingrisks,andbysellinginsuranceproductsthatpre-
dude selectionon the basisof risk. Insurers use a variety
of methodsto obtaininformationaboutthe underlyingmed-
ical risks of individualsso that they can set appropriate
premiumsor not cover these risks. These methods,some-
timesclassifiedunderthe generalrubricof riskselection(as
wellasriskratingandmedicalunderwriting), includerequir-
ing medical examinations,examining previous medical
claims,establishingwaitingperiods,excludingpreexisting
medicalconditions,not guaranteeingrenewabilityof the
insurancepolicy,and refusingto insurethe individual.
Marketinginsuranceto individualsor groupsformedfor
a purposeother than to obtain healthinsurancealsomiti-
gates the risk of adverseselection.Marketingto employee
groupshas this characteristicbecauseindividualsjoin the
group for employment,not health insurance.12 Employee
groups also tend to be healthier and are generallylarge
enoughto effectivelypool risks.
12
 A second supply-sideinstabilityin insurance markets
occursas a resultof moralhazard.Moralhazardis the ten-
dencyfor insuranceto increasethe probabilityof the occur-
renceof the risk being insured against(Arnott and Stiglitz
1988). (In more general terms, moral hazard is the risk
that individualswillchangetheir behaviorbecauseof the
existenceof a contract.)Becauseinsurancelowersthe cost
of serviceto the consumerat the time of use, individuals
tend to consumemore health servicesthan they would in
the absence of insurance.Similarly,insurancemay cause
individualsto use less preventive servicesor take fewer
precautions to avoid accidents or deterioration in their
health. Moreover,sincethe costs of excessuse are spread
over all other purchasersof insurance,individualshavelit-
tle financialincentiveto restraintheir use.
Moralhazardoccursfor alltypesof insurance-publicor
private,voluntaryorcompulsory. Thebenefitpackagesofmost
insuranceprogramsare designedto dealwith moralhazard.
Featuresthat mitigatemoralhazardin healthinsuranceinclude
costsharing(deductibles,copayments,coinsurance),physi-
cal (for example,maximumof twentymentalhealthvisits)
andfinanciallimitsonbenefits(forexample,$10,000in cov-
eragefor pharmaceuticals), and total expenditurelimitson
policies(for example,lifetimeinsurancecoveragelimitedto
$500,000).Frequentrenewability is alsoused to deter moral
hazardsince premiumscan be adjustedto accountfor the
actualandprojectedexperienceof the group.
Anotherinnovationto dealwith moralhazard is the uti-
lizationmanagementpracticed by managedcare organiza-
tions (Enthovenin this volume).In their most complete
form managed care organizationsintegratethe functions
of managing financing and providing care. As a result,
throughvariousforms of utilizationmanagementby med-
icalprofessionalsemployeddirectlyor undercontractto the
managedcare organization,subscribershave far less dis-
cretionto demandservices.3 Managedcare principlescan
be adoptedby both public and privatefinancingentities.
It is not clearwhethertheseinsurancemarketinstabilities
justifycompulsorypublicfinancing(orprovisionof publidy
sponsoredinsurance)forallindividuals.4 Assumingthatpub-
lichealthservicesand redistributiveconcernsare dealt with
separately,the questionbecomesone of whetherregulated
privateinsurancemarketscan operateefficiently.On eco-
nomicwelfaregroundsthereislittledoubtthat efficientmar-
A CURMUDGEON'S GUIDE TO FINANCING HEALTI-I CARE IN DEVELOPING COUNTRIES
13
ketsthat provideinformedconsumerswitha rangeof choice
of insurancepoliciescan maximizesocialwelfare.But risk
poolingfor individualsand smallgroups,adverseselection,
riskselection,and moralhazard cancreate economicineffi-
ciencyand inequityin insurancemarketsthat justifygovern-
mentintervention in publicfinance,provision,
andorganization
ofinsurance,andpublicregulationofprivateinsurance(Aaron
1991,pp. 11-19;Arnottand Stiglitz1990).In anyeventthe
costsof governmentinterventionmust be weighedagainst
theseinherentprivatemarketinefficiencies. 15
Still,it is clearfromthe experienceof the UnitedStates-
the one majorindustrialcountryto relyon voluntary(tax-
subsidized)private insurance-that there are significant
problemswith risk pooling,access to insurance,and the
costs of health insurance and health services. Many of
these problemsoccur as a result of the rating factorsused
to set premiumsas wellas the risk selectionmethods used
by insurersto prevent adverse selection.1 6 Rating factors
includehealthstatus,age,sex,industryor occupationgroup,
group size,and geographiclocation.Ratingand risk selec-
tion often discriminateagainstsmallemployergroupsand
individualsas well as vulnerablepopulationssuch as the
handicappedandthe elderlyTheyresult in significantmar-
ket segmentationand premiumdifferentiationthat under-
mine effectiverisk pooling.
Sincethe purposeof insuranceis to poolrisks,someana-
lystsargue that the most effectivewayto do so is through
communityrating,inwhich(exceptfor adjustmentsforfam-
ily status,geography,and benefitsdesign)everyonein the
communitypays the same average premium (American
Academyof Actuaries1994).17But communityrating in
voluntaryinsurance markets increases the likelihood of
adverseselectionand creates"death spiral"instability,with
healthier individualsoptingout. The argumentsfor com-
munityrating are that risk.poolingis the purposeof insur-
ance, that youngerhealthier individualswho opt out may
end up needing medicalcare that willultimatelybe pro-
vided at publicexpense,and that these sameindividualsat
a later point in their life cyclewillend up benefitingfrom
thisarrangement.Unlikethe United States,mostindustrial
countries have dealt with these problemsthrough public
financingand provisionof health insuranceor health ser-
vices.Some U.S. reformers(includingPresident Clinton)
have arguedthat public financeand compulsorycoverage
for all is the only way to deal with this problem. Opponents
of public finance have proposed dealing with it through reg-
ulation of private insurance markets, particularly by regu-
latingrating factors and medical underwriting practices and
through public subsidies to or taxes on the insurance indus-
try to subsidize the premiums of high-risk groups (Helms,
Gauthier, and Campion 1992).
Market failure and direct consumer purchase of
health services
Direct consumer purchase of health services-whether basic
services,supplementation of insurance benefit packages, or
higher-quality services than are offered under an insurance
program or by the public sector-are an important source of
health care revenues and expendituresin allsystems,but espe-
ciaily in low-income countries. It can be argued that insur-
ance,whether publiclyorprivatelyfinanced,is the best vehicle
for financing health services,since by pooling risks, overall
risks are reduced. But since no societypredudes its citizens
from buying legal goods and services and since in many low-
income countries governments cannot raise sufficient rev-
enues to finance personal health care services for their
populations, it is important to consider whether market fail-
ures result in inefficientconsumptionand provisionwhenindi-
vidualspurchase servicesdirectly on an out-of-pocket basis.
Several market failures affect the direct consumer pur-
chase of personal health services. Some of these failures,
such as entry barriers and decreasing costs of production
(which provide a rationale for public provision or public
utility-type regulation), are onthe supply side. Others, such
as consumers lacking information about what services they
need, about what works in medicine, and about the prices
and quality of competing medical care providers, are infor-
mational.Because of such informationgaps,the care provider
often acts as the consumer's agent-yet the financial moti-
vations of the provider may not be entirely consonant with
maximizingthe welfare of the patient. As a result no indus-
trial country relies exclusively on free markets to produce
and allocatehealth care (Aaron 1991, p. 8). Moreover,direct
purchase of services cannot contribute to equity objectives
(that is, it is a pure benefit approach) and, as noted, pri-
vate purchase without government subsidies will result in
less than optimal consumption of certain health services.
INNOVATIONS IN HEALTH CARE FINANCING
14
In addition, direct purchase does not result in the gains from
risk pooling obtained through insurance.
Implications for the public-private mix
These concems lead to several prescriptions for public and
private financing of health services:
* Public health services should be financed publicly.
* Personal health services that have collective benefits
should be publicly subsidized.
* Personal health services that have no collective benefits
can be publicly or privately financed.
* Personal health services for vulnerable populations are
generally financed publicly on equity grounds.
* Insurance reduces overall risks by pooling them and is
thus a preferred method for fnancing health services.
* Insurance for personal health services can be publicly or
privately financed.
* Instabilities in insurance markets necessitate government
regulation and under certain circumstances public
financing.
* Market failures create inefficiencies relating to individ-
ual out-of-pocket purchase of health services.
* In some cases government regulation and provision of
information may be a viable alternative to public financ-
ing of personal health services.
Public Financing Sources
As noted, health care systems can be financed by a variety
of public and private sources, and these funds can be spent
on many types of public and personalhealth services.Certain
public health services (such as immunizations) are quasi-
public goods and should be publiclyfinanced or subsidized.
Public fnancing of personal health services is justified on
redistributive and equity grounds.
Twoother potential rationales for public financing of per-
sonal health services for the nonpoor as well as the poor,
either directly or through publiclyfinanced health insurance
systems, result from market failures in private health insur-
ance markets that preclude effective risk pooling and in
the consumption and provision of personal health services.
Public "insurance" systems can be of two general types:
national health service approaches (Beveridge model) and
TABLE
6
Government
revenuesbyregionand incomegroup
(percentage
of GDP)

Standard Range Numberof

Region/income
group Mean Median deviation Minimum Maximum observations
EastAsiaandthe Pacific 23 20 9 8 37 11
EuropeandCentralAsia 39 41 12 12 60 13
LatinAmercaandthe Caribbean 24 25 8 8 42 21
MiddleEastandNorth Arica 32 31 10 12 48 10
SouthAsia 27 20 15 10 47 6
Sub-SaharanAfrica 26 22 15 11 63 20
Developingcountries 28 26 12 8 63 76
Industrialcountries 45 44 8 31 62 21
Low income 20 19 9 8 44 22
Middleincome 31 30 12 8 63 54
Highincome 42 44 11 12 62 24
Note:Dataareforthelatest
available
yearbetween 1990and1995.
Incudes
central,
state,
andlcalgovemment
revenues.
Alsoincludes
taxand
nontax
revenues,
aswellasgrant
assistance.
IMF,
Source: Govemment Finance various
Statistics, issues.

social insurance funds (Bismarck model; see OECD 1992). Government revenues
In both models collective risk pooling is achieved through
compulsory taxation. Total government revenues (central, regional, and local) as
There are, however, two general distinctions. First, a percentage of GDP vary significantlyby region and by
national health services tend to be financed from a mix of income level (table 6). The relationship between per capita
general taxes and other public revenue sources, while social income and governments' ability to raise revenue is shown
insurance funds tend to be financed with earmarked pay- in figure 9. Several patterns are apparent:
roll taxes.18 Second, because national health services are * Revenue-raising capacity increases with income.
financed from the general budget, they are subject to annual * Relativeto their GDPs, low-incomecountries and regions
budget processes. Social insurance funds tend to be more can raise less than half the revenues that high-income
independent of such annual political machinations. There countries can raise.
is no one "right" approach for developing countries to use
on this issue, and various mixes of public and private financ- FIGURE9
ing are possible (seeJ6nsson and Musgrove in this volume). and per capitaGDP, various
Governmentrevenues
Although much of the discussion of health financing c
focuses on the benefits of health expenditures and on what of GDP)
Revenues(percentage
services should be purchased, cost and equity issues are also
involved in raising revenues to finance public expenditures. 80
These costs make it of critical importance for governments
to evaluate the benefits of public expenditures against the 60 0 *
costs (both economic and equity) of raising these revenues., 4 o *
As stated in World Bank (1991), "incremental changes in 40 OO
the level of taxation should reflect, among other things, 20 44
the benefits derived from incremental changes in the pub- *o o
lic expenditure program and the relative costs of financing o
100 1,000 10,000 100,000
it by means of taxation or non-tax revenues"(p. 18).Thus PercapitaGDP (US$)
policymakers must consider both the sources and uses of
funds as they consider health care financing reforms. FinanceStatistics.
Source:IMFGovernment

A CURMUDGEON'S GUIDE TO FINANCING HEALTH CARE IN DEVELOPING COUNTRIES

15
* Asa groupdevelopingcountriesraiselessthan two-thirds
of the revenuesraisedin industrialcountries.
Regionalvariationsalsosuggestseveralinterestingpat-
tems. First,the high ratioof revenueto GDP in the Middle
East and North Africalikelyreflectsoil and gas revenues.
sinceboth regionshavesignificantly higherpercapitaincomes
than Sub-SaharanAfricaand SouthAsia,yet their ratiosof
revenueto GDP are similar.This similaritymayreflectthe
disparatemixof low-and middle-incomecountriesin both
regions,incompleterevenuedata forsomecountries,1 9 inef-

Second,highrevenuegenerationin EuropeandCentralAsia fectivetax administration,or societalpreferencesfor indi-

likelyreflectsthe legacyof and continuedrelianceon the
centralizedrevenue-raising systemscreatedundersocialism,
in whichtaxes are establishedby the nationalgovernment
and each regionreceivesa percentageof that base, and in
which high payroll taxes are used to finance social pro-
grams.Third, the resultsfor East Asia and the Pacificand
LatinAmericaandthe Caribbeanare somewhatanomalous,
vidualrather than governmentresponsibility.
Still,the basic conclusionon developingcountries'rev-
enue-raisingcapabilityis clear:it is significantlylessthan in
industrialcountries,and the pooreris the country,the less
is the capacity.This limited capacityhas important impli-
cationsfor developingcountries' abilityto finance health
and other public services.The differentsourcesof public

FIGURE10
Taxationclassification

gIncome
(direct)
taxes | Property
(direct)
taxes

Company Personal
taxesl
taxes

Corpoatiand ancome Cogapin Gift Inhentance Death Wealth

taxtandsax Polltax gains taxes taxes duty tax

Ad valorem

Source:
AJlan
1971.

INNOVATIONS IN HEALTH CARE FINANCING

16
revenues,the criteria by whichthey should be evaluated, the financingmodelsfor industrial countriesmaybe less
the institutionalrealitiesof developingcountriesthat cir- relevantto developingcountries.The recent literatureon
cumscribethe use of various revenue-raisingmodalities, taxation theoryand public financein developingcountries
and the policyimplicationsfor financinghealth systemsin providesimportantinsightsinto this issue.
developingcountriesare discussedbelow. Governmentshave many optionsfor raising(or mobi-
lizingthrough compulsion)revenues:direct and indirect
Sourcesof publicrevenues taxes, user charges,mandates, grant assistance,and bor-
rowing.20 Directtaxesare taxes on individuals,households,
The sourcesof health care financingin developingcoun- and firms and include personal income taxes, corporate
trieshavenot receivedmuch in-depth treatment in the lit- profits taxes, payrolltaxes, social securitytaxes, property
erature. This is unfortunatebecause the cost, equity,and taxes, and wealthtaxes (figure10).Indirecttaxesare taxes
administrativeimplicationsof raisingrevenuesin develop- on transactionsand commoditiesand includegeneralsales
ingcountriesaredifferentthan in industrialcountries.Thus taxes,valueaddedtaxes,excisetaxes,turnovertaxes,import

Consumptontaes

Sales
taxes |Factor taxes

Ttrnwer
Salesadded
Purd
se inlds.ditreStamp
and roil Mre estaekactr|
IDrnl l
I
l
I
I

l ~~~~~~~~~~~~~I
I I I

_____
____ ____
____ ____ ___l l l l l _ _ _ _I__ _

______ _____ _____ ______ _____ _ l I l I l- - - -- -

+ ~ ~~~ l-

II I I I
I~~~~~~~~~~~~~~~

Specik l

A CURMUDGEON'S GUIDE TO FINANCING HEALTH CARE IN DEVELOPING COUNTRIES

17
duties,and exporttaxes.Expenditurescan alsobe financed
throughuserchargesand taxes on state-ownedenterprises.
Anotherway to provide servicesthrough public interven-
tion is with employerand individualmandates.Charitable
contributions(discussedbelowunder privatefinancing)can
alsoaugmentgovernmentrevenues.Grantassistance in the
formof grantsto governments(ornongovernmententities)
is another potential sourceof health revenues.Borrowing
fromdomesticor foreignsourcescan alsobe usedto finance
publichealth spending.In evaluatingthe altemativemeth-
odsused to financehealthcaresystemsin developingcoun-
tries, the economic, equity,political,and administrative
effectsof these methods must be assessed.
All countries use a combinationof these methods to
financetheir healthsystems.As discussedbelow,there are
significantdifferencesbetweenthe methodsused in indus-
trial and developingcountries. Some OECD countries
(Canada,Sweden,the UnitedKingdom)financetheirhealth
systemsthroughgeneralrevenues.Others(France,Germany)
relyon socialsecurity(payroll)taxes.Payrolltaxes are also
used in manytransitioneconomiesand in LatinAmerica.
Other countries use multiplemethods of financing.The
U.S. Medicareprogram,a nationalpublicprogramfor the
elderly and the disabled, is financedwith social security
taxes, general revenues, and premium payments from
enrollees.Userchargesare a commonfeatureof most pub-
liclyfinancedprograms,especiallyfor goods (likepharma-
ceuticals) and services where most countries require
consumersto share the costs of coveredservices(OECD
1992 and 1994). PresidentClinton's proposed reform of
the U.S. healthcare systemwasbased on a mandatefor all
employersto financecoveragefor theiremployees(Zelman
1994). A variant of this approach proposed in previous
unsuccessful U.S. reform efforts is the "play or pay"
approach,in whichemployerswouldeitherprovideemploy-
ees withhealth insuranceat a leveldeterminedby the gov-
ernment or pay an additional tax that would be used to
finance coveragefor employees(Moffet 1993,p. 2).
Taxation
Taxationis the mainsourceofrevenuein nearlyeverycoun-
try.In assessingthe equityeffects of differenttaxes, a dis-
tinction must be made between the impact or statutory
INNOVATIONS IN HEALTEI CARE FINANCING
18
incidenceof a tax (that is, the entitylegallyresponsiblefor
the tax) and the economicincidenceof the tax (that is,the
changein the distributionof privatereal income resulting
fromthe tax or the entitywhoultimatelybears the burden
of the tax).Impact and economicincidencediffer through
a processknown as taxshifting.Economicincidenceis the
more appropriatemeasurefor assessingthe overalleffects
of alternativetaxes (Rosen1995,pp. 273-77).
The basic criteria for evaluatingalternative financing
modalitiesare:
* Econornicefficiency:the tax systemshouldnot interfere
with the efficientallocationof resources.
* Equity (fairness):the tax systemought to be fair in its
treatment of differentindividuals.
* Administrative simplicity:the tax systemoughtto be easy
and inexpensiveto administer.21
Taxpolicy often involvestradeoffs among these crite-
ria.22Each is discussedin turn.
Economicefficiency. Wheneverthe governmentusestax-
ation to extract resourcesfrom the privatesector,there is
an economiccostthat in most casesexceedsthe amountof
resourcesextracted.This additionalcost is referred to as
the excessburden, efficiencycost, or deadweightloss of
taxation.In other words, excessburden results from the
productionandconsumptioninefficienciesassociatedwith
taxes that distort the decisionsmade by firms and house-
holds.This excessburden occursbecausevirtuallyalltaxes
create distortions(substitutioneffects)in economicdeci-
sionmakingthat reduce incomesand production by more
than the amount of revenue transferred to the govern-
ment.Taxesgenerallydistortthe decisionsfirmsmakeabout
production,trade,and investment,andthe decisionshouse-
holdsmakeabout consumptionand savings,both currently
and over time. For example,one study of the U.S. tax sys-
tem found that a 1 percent increasein alltax rates would
generateefficiencycostsof 17-56 percentover the amount
of revenue raised.A study of Sweden'stax systemfound
efficiencycoststo be 70-130 percentof the amountsof rev-
enueraised(WorldBank 1991,p. 3;JbnssonandMusgrove
in thisvolume).
Onlytaxes that do not changebehaviordo not result in
an excessburden. Suchtaxesare generallylimitedto lump-
sum and poll taxes (that is, a flat tax on each individual),
whichare hard to administerin developingcountriesand
performpoorlyon equitygrounds.Taxesonentitiesortrans-
actionsin whichpriceresponsiveness(that is, elasticity)is
limitedor nonexistent(that is, the quantitiesdemandedor
supplied are not affectedby the changein priceresulting
from the tax) alsohavelowerexcessburdens.Atax'sexcess
burden is also related exponentiallyto the tax rate-the
higheris the tax rate, the higheris the excessburden.2 3 Thus
taxes with low rates on entitiesor commoditieswith little
priceresponsiveness havelowerexcessburdensorefficiency
coststo the economythan taxeswith high rates on entities
or commoditieswith considerableprice responsiveness.
Excess burdens are minimizedwhen the ratio of the tax
rates is equal to the ratio of the relativeprice responsive-
ness of the commodities. 24
Equity. A second criterion for evaluatingtaxes is the
system'sfairnessor equity.This aspect of revenueraising
receivessignificant attention in the literature on health
care financing (van Doorslaer and Wagstaff 1995; van
Doorslaer,Wagstaff,andRutten1993;WHO 1996a).Equity
has two dimension:horizontalequity and verticalequity.
Horizontalequitymeans treatingtaxpayerswith the same
amountof incomeequally,irrespectiveof the sourceof the
income.Verticalequitymeans treating taxpayerswith dif-
ferent incomes differentlyor, more specifically,distribut-
ing the tax burden amongtaxpayerson the basis of ability
to pay. Definingincome and abilityto pay is not straight-
forward.Althoughhorizontalequityisoften consistentwith
efficienttaxesthat are simpleto administer,suchtaxesoften
violateverticalequity considerations(seebelow).
Usingthe tax systemto redistributeincomefromthe rich
to the poor is difficult because a sizable portion of rich
people'sincomescomesfromcapitalthat is internationally
mobile,because much of their income is entrepreneurial
and thus is hard to measure and tax, and becausethe rich
are politicallypowerful(Newberyand Stern 1987,p. 187).
More generally,there has been increasingacceptance of
the notionthat in developingcountriesthe tax systemis an
extremelypoorinstrumentforredistributingincome.Income
redistributioncan be better achievedbyusingthe revenues
from efficienttaxesto providesubsidiesto the poor (World
Bankforthcoming).Nevertheless,the economicincidence
of taxes on the distributionof income is one of the most
A CURMUDGEON'S GUIDE TO FINANCING HEALTH CARE IN DEVELOPING COUNTRIES
19
importantcriterionfor evaluatingvariousrevenue-raising
measures.
Administrative simplicity.Runninga tax systemimposes
significantcostson both governmentauthoritiesand tax-
payingunits. But perhaps more importantare the signifi-
cant equity, efficiency,and revenue implications of an
inefficientlyadministeredsystem.Administrativeand com-
pliancecosts dependon the typesof personneland equip-
ment needed, the types of records kept and information
needed, the complexityof the tax system (for example,
specialprovisionsandexemptions),and rate differentiation
acrossindividualsandcategoriesof income.Aninefficiently
administeredtax systemcan generate high administrative
costs,cause lossesof tax revenues(throughtax avoidance
and evasion),increaseexcessburdens of taxation,and fos-
ter inequity.It is becomingincreasinglyrecognizedthat suc-
cessful changes in tax policy require an effective tax
administration.
In developingcountries,whereinstitutionalcapacityis
oftenweakandtax systemdesignoverlycomplex,tax admin-
istrationand tax reform are closelyrelated.Simplicityhas
been advocatedas the fundamentalprincipalfor effective
taxation.This suggests:
- Eliminatingunproductivetaxes.
* Keeping differential tax rates and provisions to a
minimum.
Draftingclearlyand communicatingeffectivelythe pro-
visionsof tax laws(Khalilzadeh-Shirazi and Shah 1991,
p. xx).
These requirementsfor administrativeefficiencyhave
importantimplicationsfor economicefficiencyand equity.
In fact, in designingoptimal tax systems,all three criteria
must be consideredand explicittradeoffsmade.
The literature definesan optimal tax structure as one
that maximizesa country's social welfareby making the
explicittradeoffsbetween equity and efficiencythat best
reflectthe country'sattitudestowardachievingthese goals.25
As such there are no universalguidelinesfor designingan
optimal tax system;rather, countriesmust make explicit
choicesamongcompetingefficiencyand equityobjectives.
In principle,if allindividualswere identical,nondistorting
lump-sumtaxes could be imposed,horizontaland vertical
equity wouldbe preserved, and excessburdenswould be
avoided.Butsinceindividualsare not identical(andindeed,
informationon their differencescan be only imperfectly
obtainedby tax authorities),lump-sumtaxes result in fur-
ther incomeinequality.Sincesuchtaxes are undesirableon
equitygrounds (and possiblyon administrativegrounds),
governmentsmust use distortingtaxes. Thus tax systems
reflect the different attitudes that countries have about
equityand efficiencyobjectivesas wellas their basicpolit-
icalstructuresandadministrativecapabilities(Stiglitz1988,
pp. 479-80; Newberyand Stern 1987,p. 167).
Institutionalstructureof taxationin developingcountries.
A country's abilityto raise revenues, or its tax capacity,
depends on its per capitaincomelevel,the distributionof
income,the availabilityof "tax handles,"and the openness
BOx 2
Institutionalfeatures of developingcountries
that affect their tax capacity
* Mostofthepopulation isseif-employed insubsistencesmall-
scaleagriculture,wheremuchoftheincomeisinkind,trans-
actionsarehardto trace,andhighratesofilliteracy andpoor
accounting andrecord-keeping limitthe use of personal
incomeor profits taxes.
* Inurbanareasthereisa largeinformal sectorofsmalland
transientfirms,andevenindividuals employed intheformal
manufacturing sectorworkinsmallfirms,
* Large firmstendto begovernment enterprisesorextractive
industriesthatareoftenownedbyforeigners.
* Agriculturalproducts andmineralresources faceunstableand
unpredictable worldprices.
* Thedualism ofamodernurbansectoranda traditional rural
sector,andthemarket segmentation itcreates,distorts
com-
moitan,ao
modity akt,icraigtxbres
and labor markets, increasingtax burdens.,
* Highlevelsofincomeinequality tendto resultin highertax
rates,greatertaxavoidance, andhigherefficiency losses.
* Tradedistortions-import tariffs,
quotas,exporttaxes,dif-
ferential
exchange rates,foreign
exchange ratoning-abound,
resultinginresourcemisallocations andinequity.
* Theinfluence of state-ownedenterprises,coupledwith
nonoptimal userchargestructures,oftenresultsinineficient
publicinvestment decisions.
* Taxadministration capacityislimited.
Source:World Bank1991and forthcoming,Newbery andStern 1987;
Khalilzadeh-Shirazi
andShah1991.
INNOVATIONS
of the economy(Musgrave1987,p. 244).Industrialcoun-
tries have far more flexibilityin designingoptimal tax sys-
tems because, relativeto developingcountries,they have
higherincomelevels,moreequalincomedistributions,more
"tax handles,"lessopen economies,and more efficienttax
administrations.In fact,the institutionalfeaturesthat char-
acterizedevelopingcountriesseverelylimit their abilityto
implementthe fullrange of instrumentsthat couldbe used
to executeoptimaltaxpolicies(box2).As countries'incomes
rise, these impedimentsto effectivetaxationdisappear.
Taxationpracticesin industrialand developingcountries.
As noted, industrialcountriesraise almosttwice as much
revenuerelativeto GDP as developingcountries.Taxcapac-
ity growswith income,and urbanizationfacilitatestax col-
lection and increases demand for public services. In
attemptingto dealwith the efficiencyand equitytradeoffs
inherentin optimaltax theory,industrialcountriesrely on
incometaxes,both personal (one of the best instruments
forredistribution)and corporate,as theirprincipalsources
of revenue, followedby indirect taxes. Developingcoun-
triesrelyon trade (especiallyimport) taxes,indirect taxes,
and to a lesserextentcorporate incometaxes.Developing
countriesarefar lessreliantonpersonalincometaxesbecause
these are not an easilyaccessible"tax handle"for the rea-
sons stated above. In fact, in many developingcountries
personalincome taxes are essentiallywage taxes on gov-
ernment employeesand employeesof largecorporations.
In many cases these corporationsare foreignowned and
engagedin the exportationof mineralresources.As such
they generate largeprofits and are convenientand visible
entitiesto tax.Personalincometaxesbecomemore impor-
tntias coutrs' incomgecorpore
es income
tant as countries' incomesgrow,while corporate income
taxes followa bell-shapedcurve, initiallyrisingin impor-
tance and then falling.
Domestictaxes on goods and services(such as excise
and salestaxes) are an importantsource of revenue in all
countries.In developingcountriessuchtaxes are the most
ta ntso rceo rene aftererad taxes, wici
taxes playinga larger role than general salestaxes (which
are often imposedwith multiplerates and stronglyresem-
ble excisetaxes). Excisetaxes on alcohol,cigarettes,and
1 p
petroleumproductsare particularlyimportant,accounting
for almostthree-quartersof excise tax revenues.There is
IN HEALTH CARE FINANCING
20
no relationshipbetween these taxesand growthin a coun-
try's income.
Tradetaxes,especiallyimportduties,arethe mostimpor-
tant sourceof revenuein developingcountries.Importduties
fall sharply as countries' incomes increase (as do export
duties, whichare far less important).
Socialsecuritytaxes are a minor sourceof revenuesin
developingcountries comparedwith industrialcountries,
wheretheyare severaltimesas important.Moreover,social
securitytaxes become increasinglyimportant as incomes
grow.Thebasicinstitutionalweaknessesof developingcoun-
tries described above preclude widespreaduse of social
securitytaxes.Sociopoliticalfactorsoften lead to their use,
however,as evidencedby the importanceof these taxes in
LatinAmericaand the Caribbeanand Europe andCentral
Asia.Both industrialand developingcountriesoften ear-
mark suchtaxes to support healthinsurancesystems.
Wealthand propertytaxes account for a smallpercent-
age of tax revenuesin industrialand developingcountries,
althoughthey increasein importanceas incomeincreases
(Tanzi1987,pp. 205-36; WorldBank 1991,p. 16).
Implicationsfortaxrevenuein developingcountries.
In pur-
suingoptimaltax policies,industrialcountriestend to bal-
ance equity and efficiencyconsiderationsby usingbroadly
based progressiveincometaxes and (more or less)propor-
tionalpayroll,corporateprofits, and generalconsumption
(sales,valueadded) taxeswithlittle tax rate differentiation
and relativelylow tax rates. Inequalityin the distribution
of income is often handledwith cash transfersand subsi-
dized servicesfor the poor. Becausethe supplyof labor is
relativelyinelastic,the basesfor differentcommoditytaxes
are broad, and tax rates are relativelylow (withlittle rate
differentiation),these typesof taxesgenerallydo not result
in the largeexcessburdensfound in the tax systemsof devel-
opingcountries-where relianceon trade taxes, commod-
ity taxes, and corporate income taxes with widely
differentiatedandhightax rates,market segmentation,and
limitedadministrativecapacityresult in significantexcess
burdensand inequity.
Given the institutionalweaknesses(and socialwelfare
functions)in developingcountries,whichrevenue-raising
mechanismsbest approximateoptimaltax policies?
* Commoditytaxes are an importantsource of revenue,
A CURMUDGEON'S GUIDE TO FINANCING HEALTH CARE IN DEVELOPING COUNTRIES
21
and equityand efficiencyconcernscan be addressedby
broadeningthe tax base, keepingtax rates low,limiting
tax rate differentiation,and not discriminatingby the
source of production (foreign or domestic).The best
instrumentfor thisis a single-ratevalueadded tax (VAT)
that exemptsitemsthat are a significantcomponentof
expendituresby the poor.
Equity would also be fostered by introducingluxury
and excisetaxeswith just three or four rateson income-
elastic goods that are not distinguishedby source of
production (foreignor domestic).
Giventhat the supplyof landis inelasticandownerstend
to be wealthy,land taxeswould scorehigh on both effi-
ciencyand equitygrounds.Suchtaxes generallyare not
politicallyfeasible,however.
Corporate incometaxes, especiallyon mineral extrac-
tion, are an importantand accessible"tax handle" and
mayscorehigh on equity and efficiencygrounds, espe-
ciallyif the countryis a majorsupplierofthese resources.
But forcorporateincometaxes more generally,tax rates
that exceedinternationalstandardswouldlikelylead to
capitalflight,retardingeconomicgrowrth.Sincecapital
is more mobileinternationallythanlabor,capitalshould
be taxedlessheavilythan labor.Corporateincometaxes
with a singlestatutoryrate should be considered.
* Personalincometaxeswould likelyfall onlyon govern-
ment employeesand employeesof largefirms,discour-
aging saving and limiting their potential for raising
revenue.Moreover,givenlowformalemployment,taxes
on labor (such as social securitytaxes) are unlikelyto
raisesubstantialrevenue and maygenerateexcessbur-
dens (dependingon the supplyelasticityof labor and
whethersuchtaxesencourageinappropriatecapitalsub-
stitution). Such distortionsin productioncould retard
economicgrowth.But sincepersonalincometaxesscore
high on equity grounds,considerationshouldbe given
to a single-ratepersonal income tax combined with a
generouspersonalexemptionin lieu of the more com-
mon systemsof sharplyprogressiverates and a variety
of deductionsand credits.
* Import taxes result in larger excessburdens than gen-
eral consumptiontaxes. However,theyare easyto col-
lect at the point of entry and to the extent that imports
are highly income-elastic(that is, responsive)luxury
 goods,suchtaxesmayhavefavorableequityimpacts.Yet
if importshavea high capitalcontent,importtaxes may
generatesignificanteconomicdistortions.Countriesmust
carefullyweighthese tradeoffsas well as the potential
long-runeconomicdevelopmenteffects of these taxes
(WorldBank 1991,pp.57-58; Newberyand Stern 1987;
Thirskforthcoming).
Taxsubsidiesfor purchaseof healthinsuranceand medical
services.Another tax-basedmechanismthat warrants fur-
ther discussion,given its widespread use in many coun-
tries,is taxsubsidizationforthepurchaseofprivateinsurance
and medicalexpenses.Thesesubsidiesare generallyin the
formof deductionsor creditsagainstpersonalor corporate
incometaxes and, in the caseof employer-provided health efit). Fourth, such subsidiesmute the cost consciousness
insurance,are providedbynot countingthe employer'scon-
tributionto healthinsurancepremiums(ormedicalservices)
as incometo the employee.
Subsidizingemployersto purchasehealthinsurance(or
to self-insureby payingfor or directlyprovidingservices)
has severaladvantages.First, when private firms provide
the insuranceor service,the expendituresare not a gov-
emmnentbudget expense. Second, as discussed above,
employergroupsare alogical"community"to insurebecause
(for largeemployers)theyprovidea largeenoughgroup to
effectivelypool risks,employerscanbe moreinformedpur-
chasersthan individuals,and adverseselectiontends to be
mitigated by the fact that employeesjoin the group for
employmentpurposes.Moreover,if employersalso cover
the employee'sdependents(asin the more advanceddevel-
opingcountries),largesegmentsofthe populationwillhave
privatelyprovidedhealth insurance.
Developing countriesalsomustconsiderissuesoftaxadmin-
istrationand industrialstructure.If the bulk of the popula-
tion worksin subsistenceagricultureand the formalurban
manufacturingsectoris composedof many smalltransient
firms,a significantportionof the populationis unlikelyto be grantassistance,andborrowing.Thesefinancingsourcescan
coveredbyemployer-based insurance.Moreover,if muchof
the modemurban sectoris workingat or nearthe minimum
wage,requiringemployersto provide healthcoveragewill
increaseunemployment.Althoughtax subsidiesthroughthe
corporateincometax are generallyeasyto administer,there
isastrongpossibilityoffraudincountrieswithweaktaxadmin-In a traditionalpublicfinancecontext,individualspayuser
istration.Thesupplyof insurancealsomaybeaproblem,espe- charges(basedon the principleof costrecovery)for a pub-
ciallyin the absenceof developedfinancialmarketsand gov-
emient regulation.Similarly, firmsmaynot havethe exper-
tiseto self-insureor to purchaseinsurance.Nevertheless,as
theformalsector,financialmarkets,andadministrative capac-
ity develop,usingtax subsidiesto encourageemployer-based
insurancemaybe an effectivewayto encourageprovisionof
insurancethat is largelyprivatelyfinanced.
There are, however,significantdrawbacksto usingtax
subsidiesin this manner.First, there is a potentiallylarge
revenuelossto the government.Second,dependingon how
the tax subsidyis structured,it maybe an open-endedloss
ofrevenuesbasedsolelyon the decisionsof privateemploy-
ers. Third, the benefit incidencemaybe highlyregressive
(withhigh-wagefirmsandindividualsgettingthe mostben-
of employers(in terms of the comprehensivenessof the
insurancethat they purchase)and employees(since they
are payinglessthan the fullcost of the servicesused,by the
amountof the subsidy).In other words, suchsubsidieswill
increasehealth care costs and may encouragecoverageof
healthservicesofmarginalbenefit.Fifth,employer-financed
insurancewillresult in lowercash wages(see CBO 1994;
Hall 1994;Hoff 1996;Arnett 1996).
In summary,developingcountriesneed to replacenar-
row,distortingtax bases that have high and widelydiffer-
entiated rates and numerous loopholeswith broader tax
bases that generatehigherrevenuesat lowerrates and do
not discriminateagainst the various sources and uses of
income.Doingsowouldresultin the simultaneousachieve-
ment of efficiency,administrativesimplicity,and horizon-
tal equity (Thirsk forthcoming).Moreover,effectivetax
reforms can only take place if they are accompaniedby
reformsin tax administration.
Countriesalsoneed to considerusingsourcesof public
revenueother than taxes to finance their health systems,
includinguser charges,employerand individualmandates,
be evaluatedusingthe same criteria-economic efficiency,
equity,and administrativesimplicity-describedabove.
User charges

INNOVATIONS IN HEALTH CARE FINANCING

22
lidy providedgood or service.Suchchargesshouldbe used
whenevera publiclyproducedgood or servicehas benefits
that canbe assignedto an individual.Taxesare more appro-
priate to finance publicgoods with collective(rather than
individual)benefits,to compensatefor marketfailures,and
to achievedistributionalobjectives(WorldBank 1988).
User chargesshouldbe evaluatedusingthe same crite-
ria used to evaluatetaxes (economicefficiency,equity,and
administrativesimplicity).Althoughuser chargesfor pub-
liclyprovidedservicesare a sourceof governmentrevenue,
from a nationalhealh accountsperspectivethey are tech-
nicallyprivate expendituresfor publiclyprovidedservices.
In this context they are no different from direct private
out-of-pocketpaymentsfor nongovernmentservices,dis-
cussed below.But since they haveimplicationsfor public
financingand governmentpolicy,user chargesfor publicly
providedservicesare analyzedin this section.
User-charges embodythe benefitapproachof taxation-
that individualswho benefit from a serviceshould payfor
it. Userchargescanbe usedwhenpublicserviceshavelargely
private benefits (most personalhealth servicesfulfillthis
criteria,althoughbasicpublichealthservicesdo not). User
charges differ in one important aspect from taxes: if set
appropriately(that is, so that the private marginalbenefit
equals the marginalcost of providingthe service), user
charges can generate revenueswithoutany efficiencyloss
(WorldBank 1991,p. 20). Yetuser chargesset in thisway
wouldnot be particularlyequitable,sincethe poor andnon-
poor would pay the same rate for services.This outcome
can be offsetbyexemptingthe pooror bybasingusercharges
on abilityto pay.As in the caseof optimal taxes,however,
this willresult in some efficiencyloss.
In practice,user chargesare often set below marginal
costs,resultingin excessconsumption,overproduction,and
inefficiency.As a result some analysts have argued that
user chargesshouldbe set at levelsabovethe marginalcost
of producingthe service-forincome-elasticgoods (that is,
servicesfor which consumptionincreasesdisproportion-
atelyrelativeto increasesin income),and that the profits
be used to subsidizeservicesfor the poor (Ray 1975).As
in the previouscase, there is a tradeoff between excess
burden and equity.
Userchargesare an importantsourceof revenuein cases
where it is politicallyunfeasible to reduce spending or
A CURMUDGEON'S GUIDE TO FINANCING HEALTH CARE IN DEVELOPING COUNTRIES
23
increasetaxes. Household surveysin a number of coun-
triesindicatethat peopleare willingto payfor servicesthat
they deem of benefit to them, makinguser charges a less
coercivewaythan taxationto raiserevenuesto financepub-
licservices.Moreover,efficiency,equity,and socialwelfare
can be improvedif the additionalrevenuesgeneratedare
usedto providepublicserviceswiththe highestsocialreturn.
User chargesfor publiclyfinancedor providedgoods
and servicescantake a varietyofforms.If the serviceis pub-
liclyprovidedhealthinsurance,individualsmaybe required
to pay premiums or share costs (for example, through
deductibles,coinsurance,andcopayments). Whilethe health
insuranceis providedpublicly,the health servicescould be
providedthrougha public or privateinstitution.Under the
more traditional concept of user charges,individualsare
requiredto paya feefor publiclyprovidedservicesin a pub-
lic facilityThe different types of user chargesrelating to
publichealthinsurance(for example,premiums)add some
complexityto the traditionalanalysisof user charges.Still,
the basic argumentsand analysisare the same as they are
for less complexpubliclyprovidedservices.
Thepremiumsandcostsharingpaidforapubliclyfinanced
insurancepackageare sourcesof publicrevenue,but they
alsoinfluencethe allocation(use)of resources.Premiums
havethe virtue of spreadingcosts(in atrue insurancesense
of spreadingrisks) over the entire populationcoveredby
the program.Cost sharingessentiallytaxesindividualsatthe
point of serviceuse, affectsfar fewerindividuals,and may
be viewedas a tax on the sick.But becauseit counteracts
the moral hazardinherentin healthinsurance,cost sharing
is an importanttool for achievingallocativeefficiency.For
example,the U.S. Medicareprogram covershospitaland
inpatientservicesusingsocialsecuritytaxes,but outpatient
servicesare coveredby a voluntarymedicalinsurancepro-
gram(MedicareSupplementaryMedical Insurance)inwhich
individualspay a highlysubsidizedpremiumequal to just
25 percent of the actuarialcosts of this insurance(the rest
of the costsare coveredby generalgovernmentrevenues).
Cost sharing, both deductibles and coinsurance, is also
required at the point of serviceuse to encourageefficient
use of servicesunder both programs.
In designingpremiumsand cost-sharingarrangements,
severalimportantequity,efficiency,andadministrativecon-
cerns emerge. To ensure equity in most cases the poor
must be exempt from such charges. To ensure efficiency,
copayments must be set high enough to discourage frivo-
lous use, but not so high as to discourage the use of needed
or cost-effective services. In fact, some countries have dif-
ferential cost sharing depending on the type of service, so
that preventive services generally do not have any cost
sharing (and indeed may be provided directly by the gov-
erninent as a pure public good outside the health insur-
ance system), while less important or more discretionary
services may require high levels of cost sharing.
Another aspect of premiums and cost sharing relates to
the need to exempt individuals with catastrophic medical
expenses. In industrial and some developing countries 10
percent of the population may incur 70 percent of health
care costs. Exempting these people on equity grounds results
in a substantial loss of potential revenues. This adjustment
also violates the efficiency rule that user charges should be
set equal to marginal costs (a rule more easily applied to
more traditional publicly provided services, such as water
and electricity).
Administrative issues are also important. If the costs of
collection exceed the user charge (unless there are signifi-
cant allocational savings), revenues will actually decline.
Another administrative compliance issue concerns who gets
to keep the additional revenue. In a number of developing
countries, unless the health facilityor provider gets to keep
some or all of the additional revenue, they have few incen-
tives to collect the cost sharing. Furthermore, as long as
the facilities and providers use these revenues for quality
improvements or service enhancements, individuals will
continue to use these facilities.
Another issue concerns policymakers relying on user
chargesto avoidconsidering altemative financingapproaches
that might generate larger revenues more easily and equi-
tably. For example, social insurance systems might have a
much larger impact and longer-lasting ability to finance
underfunded public systems and provide the government
with an opportunity to promote efficiencyby creating incen-
tives for better service delivery (Lewis 1993).
Many industrial and developing countries have extensive
experience with various types of cost sharing-both formal
user charges and direct out-of-pocket payments for private
services-including controlled experiments in the United
States and China. In the United States a $70 million health
INNOVATIONS IN HEALTII CARE FINANCING
24
insurance experiment conducted by the Rand Corporation
found that well-designedcost sharing can provide additional
revenues and increase efficiencywithout discouraging nec-
essary service utilization. The evidence from Canada and
Europe regarding the equity and efficiencyeffects of user
charges is more mixed, however. Gertler and Hammer (in
this volume) summarize much of the evidence and discuss
the links between user charges and social insurance.
Evidence on user charges in developingcountries is exten-
sive and varied. The findings indicate:
* A strong utilization response to user charges, with dis-
proportionately larger reductions in utilization on the
part of the poor.
- Exemptions from user charges often benefit high-income
groups.
* User charges need to be considered in the context of
time costs and managerial and administrative capacity
* Although user charges can be used to supplement pub-
lic revenues for curative care at the facility level, total
revenue generation has not met expectations (being on
the order of 5 percent of total revenues).
* Effects on health status and efficiency are less clear,
although there is some evidence of adverse effects on
health outcomes.
* People appear more willing to pay for tangible benefits
(such as drugs).
* User chargesfor public serviceshave elicited higher prices
for private services, resulting in declines in total utiliza-
tion (see Gerder and Hammer in this volume;Wang'ombe
in this volume; Gertler and van der Gaag 1990; Creese
and Kutzin 1995).
In other words, user charges can increase efficiency and
ensure equity if the systems are designed carefully and
generate tangible benefits for individuals. But time costs,
administrative capacity,and impacts on private pricing poli-
cies must be analyzed carefully. Moreover, while revenue
generation can have important effects at the facility level,
overall revenue generation may be lower than original
expectations.
Mandates
When fiscal constraints are tight and market failures pre-
dude public provisionof a particular benefit-such as health
insurance-government mandates to require provisionof
such benefits, either by employersor by individuals,can
help governmentachieveits policygoals.2 6 Employerand
individualmandates can be evaluatedusingthe same cri-
teria used to evaluatealternativerevenue-raisingsources.
In fact, under certain conditionsmandates are more effi-
cient than taxes, though they tend to be less equitable.
Employermandates can be more efficient than public
provisionfinancedby inefficienttaxes in severalways.The
costsof the mandatedbenefitwillbe treatedbythe employer
as an additionallabor cost and willbe shifted back on to
employeesin the form of lowerwages,discretionarybene-
fits, employment,or future wage increases.If employees
increasetheir labor supplyto offsetthe costsof the bene-
fit, the increasein labor supplywillreduce the deadweight
loss.If the increasein labor supplyfullyoffsetsthe costsof
the benefit (that is, there is fullvaluationof the benefit by
employees),thenthereis no excessburdenandthe employer
mandate is a more efficientmeans of providingthe bene-
fit thentaxation(Cutlerand Madrian1996).In effect,man-
dates are a benefit tax.
There may,however,be institutionalconstraintsthat pre-
cludethis employmentresponseand wageor discretionary
benefit adjustment from happening, including antidis-
crirninationrules,workplacerules, union rules, minimum
wagelaws,and soon. Thus evenif everyonein the employ-
mentgroupfullyvaluesthe benefit,these rigiditieswillcause
efficiencylossessincethe fullwageadjustmentwillbe inhib-
ited. Firms unable to reduce compensationwould close,
and those in noncompetitivemarketswould shift costs to
consumers.High-cost(elderly)andlow-wageworkerscould
lose their jobs. In fact, in certain cases these efficiency
costs can exceed those that would occur under a payroll
tax fallingon all workers to finance the benefit (Gruber
1994).
The mainproblemwithmandatesrelatesto equity,since
the valuationofthe benefitand the subsequentwageadjust-
ments generally bear no relationship to ability to pay.
Mandatesare a benefittax,not an abilityto paytax.Indeed,
in a world of rigiditieshigh-costand elderlyworkerswould
likelysuffer the most, makingthe mandate a regressive
approachto financinghealth insurancebenefits.
Mandatescan alsobe imposedon individuals.Mandates
on individualsdo not have anyadverse effectson jobs or
A CURMUDGEON'S GUIDE TO FINANCING HEALTII CARE IN DEVELOPING COUNTRIES
25
wages,and imposelower direct costs on businesses.But
suchmandatesmaybeevenmore regressivethan mandates
on employers;they provide incentivesfor employerswho
currentlyprovidehealthinsurancecoverageto drop it and,
unless accompaniedby subsidiesto the poor to offset the
equity problem,imposehigh marginaltax rates and notch
effectsif subsidiesare phasedout as an individual'sincome
increases.Moreover,other issues unique to health insur-
ance marketsarise, such as the higherpremiumscharged
for individualpoliciesrelativeto grouppolicies(Committee
on Energyand Commerce,House of Representativesand
SpecialCommitteeon Aging,U.S. Senate 1989).
Mandatescan be an efficientmeansof correctingfor the
failureof privatemarketsto providehealthinsurancewith-
out increasingpublic expenditures and introducing the
distortivetaxes needed to financesuch increasedexpendi-
tures. But equity considerationsand the effects on work-
ers andfirmsmustbe carefullyconsideredin lightof possible
labor market rigiditiesthat can result in adverse employ-
ment and production effects.And mandate enforcement,
adifficulttaskin industrialcountries,wouldbe evenharder
in developingcountries.Enforcementmechanismssuch as
withholdingwelfarechecks,driverslicenses,or schoolenroll-
ment wouldnot be effectiveincentivesin developingcoun-
tries because most developing countries have no such
programs or lack the administrativemechanismsneeded
for enforcement.
Grantassistance
Grants fromforeigndonors(borrowingis discussedbelow)
are a major source of health care financing and of total
expendituresin low- and some middle-incomecountries.
In 1994officialdevelopmentassistance(ODA)was more
than $47 billion,or about 1 percent of developingcountry
GDp.27 In Africa (excludingSouth Africa) donor assis-
tanceaccountsforan averageof almost20percentof health
spending,and in severalcountries for more than 50 per-
cent (WorldBank 1993).These revenuescannot be read-
ily classifiedas public or private.If they are tied to public
activities,the revenues essentiallyaugment government
efforts.Iftheygoto nongovernmental organizations(NGOs)
or directlyto individuals,they augmentprivate financing.
Donors are concerned about whether such assistance
increasesnet expendituresin the targeted sectorsor sim-
ply substitutesfor governmentspendingthat maythen be
used for unproductivesectors, such as militaryspending
(Feyzioglu,Swaroop,and Zhu 1996).
Donors typicallyfinance large shares of both capital
and recurrent expenditures.Donors should both coordi-
nate their efforts and get their prioritiesright, especially
since such activitieshave importantimplicationsfor pub-
lic and privatefinancingand for current and future expen-
ditures and revenue needs (WorldBank 1993, p. 167).
Donorsshouldtargettheir assistanceto high-priorityareas,
andwithbetter coordinationcouldavoidfragmentationand
conflictingimpactsfrom their assistance.
Grant assistanceis subjectto the budgetarysituations
andpoliticalagendasof donors.Asa result,whilegrantassis-
tance is an importantfinancingsourcein the short run for
many countries to purchase essentialhealth servicesand
developcriticalinfrastructurecapacity,it is not a reliable
long-runsource of financing.Thus it cannot be relied on
to ensurelong-termfinancialsustainability.
* growthand developmentobjectives.Given the poor bal-
Borrowing
Borrqwing,like grant assistance,can be either a publicor
privatefinancingmechanism.When the governmentis the
borrower,it is a publicrevenuesource.When aprivateentity
is the borrower,it is a private revenuesource.Funds can
be borrowedfromeitherdomesticorforeignsources.Foreign
sourcesof health financinginclude internationaldevelop-
mentorganizations,bilateraldonor assistanceagencies,pri-
vate commercialinstitutions,and foreignmedicalsuppliers.
Such lendingmayinclude a significantgrant (donor) ele-
ment or maybe at full market rates. The essentialfeature
of borrowingis that the funds eventuallymust be repaid.
Borrowing,in effect,imposesa burdenonfuturegenerations.
Domesticand foreignfinancinghave differentimplica-
tions. Domesticborrowinghas a clear opportunitycost-
namely,the opportunitycostof the alternativeusesof those
funds. If, in the absenceof borrowing,those funds would
havebeen used for domesticconsumption,whilethe bor-
rowedfunds areused forcapitalinvestment,then economic
growthand the future well-beingmayimprove.But if the
borrowedfundsare usedforlow-priority or inefficientrecur-
rent health expenditures rather than productive capital
INNOVATIONS IN HEALTH CARE FINANCING
26
investments,overallgrowth and future well-beingmay be
reduced. More generally,public investmentsfinancedby
domesticborrowingdo not contributeto capitalformation
if they simplydivert these funds from private investment
opportunities.In either case the loan willhave to be paid
backwith interest out of future production.28
One essentialdifferencebetween domesticand foreign
borrowingis that foreignborrowingdoes not require that
current consumptionbe reduced. Moreover,whilethe net
gainto future generationsis lowerthan it would havebeen
had governmentspending been financed out of current
tax revenues,the cutbacksin future consumptionneeded
to pay off and servicethe loan comeout of a higher level
of-income (assumingwise investment choices) because
foreignborrowingdoes not imposeanyimmediateoppor-
tunitycosts (Musgraveand Musgrave1976).
Borrowing,likeother potentialformsof revenueraising,
requires careful considerationof intertemporaltradeoffs.
Borrowingimposesa burden on future generationsand so
must be carefullyconsidered,in the context of economic
ance of paymentssituationsin many low-incomecountries
aswellas higher-priority developmentneeds,borrowingfor
socialprograms,exceptwhere suchborrowingis on a con-
cessionalbasis,is likelyto be quitelimited.Moreover,con-
cessionalborrowingfrom itnternational organizations(such
asthe WorldBank)generallyhas policyconditionsattached.
Publicfinancing implications for developing
countries
Governmentscan use their revenue-raisingand regulatory
powersto publiclyfinance or to require private financing
ofhealthinsuranceandhealthservices.Yetthe government's
abilityto raise revenuesand the private sector's abilityto
complywithgovernment mandatesand userchargerequire-
ments are inverselyrelatedto countryincomes.Economic
bases,institutionalstructures,and administrativecapacities
stronglyinfluence publicrevenue-raisingpotential
andinstru-
ments.In developingcountriesseveralissuesmeritpartic-
ular attention:
* Taxesand other public revenuesourcesmust be evalu-
ated in termsof economicefficiency,equity,and admin-
istrativesimplicity.
* The inherentlyagrarian/ruralnature of much economic
activity,the smalland transientnature of urban formal
sector firms,and the opennessof their economiesgen-
erallyprecludeheavyrelianceon the more efficientand
progressive incometaxesusedin mostindustrialcountries.
* Limitedtaxablecapacityand inefficientadministration
limitrevenuegenerationpotential.
* Broadlybasedtaxeswithlimitedrate differentiationand
few exemptions,deductions,and exclusionsappliedto
commodities,aswellas entitiesthat displaylittlerespon-
sivenessto the tax,are preferablefor economicefficiency
and equity
* A broadlybased valueadded tax with limitedrate dif-
ferentiation,exemptionsfor the poor, and nondiscrim-
ination between domesticallyproduced and imported
goodsrankshighin terms of equity,economicefficiency,
and administrativefeasibility.
* Publicfinancingof personalhealth servicespoolsrisks,
eliminates adverse selection, and ensures equity but
reduces choiceand imposesexcessburdens.
* Sincedevelopingcountriescan onlygenerateabout two-
thirdsthe revenuegeneratedin industrialcountries(rel-
ative to GDP), in the absence of regulated private
insurancemarketsdevelopingcountriescould improve
welfareby organizingthese marketsso that peoplewho
are ableto paycan purchasehealthinsuranceat an actu-
ariallyfair price.
* User charges can be used to enhancepublic revenues
where the servicesin question have individual(rather
than collective)benefits. But such chargesneed to be
carefullystructuredto balanceequity,efficiency,and rev-
enue generationobjectives.
Developingcountries will continue to rely on public
and private revenuesto financetheir health care systems.
The instrumentsused will affect economicefficiency,the
distributionof income,andthe revenuesgenerated.Limited
administrativecapacity and infrastructurehinder imple-
mentationof effectivepolicies.Dealingwith these imped-
iments to fiscaland health financingreform should be a
priority;otherwisethe effects of economicallyand politi-
cally "correct" policy prescriptions will be marginal.
Moreover,implicitlyunderlyingthese technicaland admin-
istrativeconcernsis the need for a stable and sustainable
politicalenvironment.
A CURMUDGEON'S GUIDE TO FINANCING
27
Private Financing Sources
Thissectiondiscussesprivatesourcesof revenuefor financ-
inghealthcare,focusingonprivateinsurance,out-of-pocket
paymentsfordirectpurchaseof medicalservices,and char-
itablecontributions.Borrowingandgrant assistance,which
can be treated as either a publicor privaterevenuesource
dependingon the entityreceivingthe funds,werediscussed
in the previoussection.
Private health insurance
The nonpoor mayhaveseveralreasonsto preferusingpri-
vate health insuranceto financepersonalhealth services.
First, by poolingrisks, overallrisks are reduced. Second,
consumersovereignty in choosingan insurancepackagethat
best fits that individual'spreferencesmaximizeswelfare.
Third,byrelyingon privatemarketsratherthan government
coercion,the benefitsof riskpoolingand consumerchoice
can be achievedwhile the efficiencycosts of taxation are
avoided.Fourth,privateinsurerscannegotiatewithproviders
over cost and qualitymore effectivelythan individualcon-
sumers.
On the other hand, adverseselectionbyindividualsand
risk selectionby insurersmayprecludeeffectiveand equi-
table risk pooling. Moreover,large numbers of insurers
andinefficientinsuranceoperationsmayresultin highadmin-
istrativecosts(suchasmarketingcosts),and individualinsur-
ers maybe too smallto effectivelynegotiatewith medical
care providers over prices. If this latter situation is the
case, insurerswill only be able to compete on their load
factor, which tends to be a small percentage of the total
premium (in the United States, for example,the average
load factor for large group policies is about 10 percent;
Enthovenin this volume).
A number of regulationsand innovativeapproachesto
riskpoolinghavebeenimplementedto dealwiththese prob-
lems.Examplesincludethe insurancereformsundertaken
by many U.S. states, managed competition as a mecha-
nism for organizinginsurancemarkets,and informal(gen-
erallyrural) risk-poolingschemesbeing implemented in
developingcountries.
Each U.S. state regulatesprivatehealthinsurancecom-
panies.29 Suchregulationsgenerallyapplyto solvency,mar-
HEALTH CARE IN DEVELOPING COUNTRIES
keting, information disclosure,premium levels,reinsurance,
and so on (box 3). Although such regulations are needed to
correct forthe market failuresinherent in insurance markets,
they also raise the costs of private health insurance policies
(because of, for example, mandated benefits; GAO 1996).
To ensure better functioning of state insurance mar-
kets, forty-four U.S. states have recently augmented their
insurance regulations, and some states (such as New York)
now require pure community rating (box 4). These new
requirements are designed to offset favorable risk selec-
tion by insurers (guaranteed issue, renewability, continuity
of coverage), redress informational gaps for consumers
(information disclosure, standardized benefits), and pro-
mote community rating (rating band restrictions).
Managedcompetition. Managed competition is a recent
innovation that organizes public and private insurance mar-
kets, fosters competition among insurers, and addresses

BOX 3
U.S. statesandinsurance
regulation
Marketconductrequirements
Planbenefitcoverage
anddescription States
reviewandapproveinsurance
policies
to ensurethattheyarenotvagueor
misleading
andmeetstaterequirements
(suchasmandatory benefit
provisions).

Small-group
reforms Moststates
requireinsurers
selling
to smallemployers
to acceptandrenewemploy-
eeswhowanthealthinsurancecoverage,
establish
shortwaiting
periods
for preexist-
ingconditions,andrequireportabilityof coveragewhenan individual
changes
jobsor
insurers.
Consumerprotectionsandcomplaints Statesmonitor insurers'actionsto ensurethatthey are not engagingin unfairbusi-
nesspracticesor otherwisetakingadvantage of consumers. They also assistcon-
sumersby investigating
complaints, answering
questions.andconducting educational
programs.

Financiol
requirements
Licensing Stateslicenseinsurance
companiesandthe agentswho sellinsurance
to ensurethat
companies arefinancially
soundandreputableandthatagentsare qualified.
Financial
solvency Statessetstandards
for andmonitorthe financial
operationsof insurersto determine
whetherthey haveadequatereservesto paypolicyholders'claims.
Statesrestricthow insurersinvesttheir funds.
Ratereviews Statesreviewandapproveratesor requireactuarialcertification
to ensurethat rates
are reasonable
for consumersand sufficient
to maintainthe solvencyof insurance
companies.
Somestatesregulateinsurerratingpracticesinthe small-groupmarketto determine
the factorsinsurerscanusein settingpremiums.
Taxrequirements
Premiumtaxes Nearlyall statesassess
premiumtaxeson insurers.
Guarantyfunds Statesrequireinsurersto financeguarantyfundsthat providefinancialprotectionto
enrolleeswho haveoutstanding medicalclaimsin caseof insurerinsolvency.
High-riskpools Somestatesrequireinsurersto financelossesin high-riskpoolsthat providehealth
coveragefor individuals
who hadotherwisebeendeniedcoveragebecause of a med-
icalcondition.

Source:
GAO1996,pp.5-6.

INNOVATIONS
IN HEALTHCAREFINANCING
28
several aspects of insurance market failure. The basic goal
of managedcompetitionis to organizeinsurancemarkets
so that individualscan makeinformedchoicesabout their
purchaseof insurancewhilereducingmoralhazardbymak-
ing them financiallyresponsiblefor the consequencesof
theiractions.Moreover,manyof the inappropriaterisk selec-
tion practicesof privateinsurerscan be mitigatedby orga-
nizinginsurancethrough purchasingcooperativessubject
to rules such as standard benefit packages,coordinated
open enrollment periods, guaranteed issue, guaranteed
renewability,limitson the number of rating bands, infor-
mation disclosure on plan performance, and marketing
throughthe purchasingcooperative(notthe insuranceplan).
By eliminating(or precluding)tax subsidiesfor the pur-
chase of insurance,subscribersare more cost-conscious,
and moral hazard is reduced. Purchasingcooperatives-are
essentiallya supermarketin whichinformedand cost-con-
sciousconsumerscompareprices and servicesbeforebuy-
ing health insurance. This approach also creates strong
incentivesfor insurersto be efficientpurchasersof services
from medicalcare providers(Enthovenin this volume).
For managedcompetitionto work effectivelyand equi-
tably,the revenuesthat financethe system(fromwhatever
source) should be pooled in a fund that distributesthem
to the insurerswho chose to participate (throughthe pur-
chasing cooperative)on a risk-adjustedcapitation basis.
Risk-adjustedcapitationpaymentsare essentialto ensure
that insurerswillenrollindividualswith greaterhealthrisks.
Developingoperationalrisk adjustmenttools is far from
simple (Enthovenin this volume). The CaliforniaPublic
EmployeesHealth Plan, the U.S. Government's Federal
EmployeesHealth BenefitsPlan, health plansin the U.S.
states of Connecticut,Florida,Iowa,Kentucky,Minnesota,
andWashington,U.S.PresidentClinton'sfailedhealthcare
reforminitiative,the now-suspendedDutch healthreform,
and the recently established Russian health insurance sys-
tem all relyon managedcompetitionorganizedby a pub-
lic entit with
bypu*hc
participation
or privateinsurersandRestrictions
lrovicentiwitheparticipationbyis
polubl rpan n determinepremiumrates(forexample,limitson the num-
Medicalsavingsaccounts.Medicalsavingsaccountsalso
provideincentivesforindividualsto be effectivepurchasers
of health services(see Nichols,Prescott, and Phua in this
volume). Simplyput, medical savingsaccounts are indi-
A CURMUDGEON'S GUIDE TO FINANCING HEALTH CARE IN DEVELOPING COUNTRIES
29
vidualsavingsaccountsfrom which individualspay for health
care, coupledwith a backup financingmechanism(along
the linesof a catastrophicinsurancepolicy).The sourceof
the savingsaccount and the backupfinancingcan be pub-
lic or private.Like directout-of-pocketpayments(includ-
ingusercharges)andmanagedcompetition,medicalsavings
accountsprovide consumerswith strong incentivesto be
cost conscious.Such accounts also preserve freedom of
choiceof medicalcare providerand offsetthe moral haz-
ard implicitin insurancearrangements.In addition,med-
ical savingsaccountsmayhelp achievebasic development
objectivesby encouragingdomesticsavings.
There are at least two basic models of medical savings
accounts:the systemusedin Singapore,whichsupplements
other publiclyfunded health programs, and the models
used in the United States, which cover a much broader
range of services.Both modelsare backed by a public or
privateinsurancemechanismthat coverscatastrophiccosts
abovesome thresholddeductibleif the savingsaccountis
exhausted.Toencouragecostconsciousness,thisdeductible
must be significant-perhaps 5-10 percent of family
income.
One of the biggest advantages of a medical savings
accountis that individualshavea strongincentiveto be pru-
dent consumersof medicalservices,sincethey can use any
unspentfunds in a varietyof ways (Nichols,Prescott, and
Phua in this volume).In other words, the benefits from
being a prudent consumeraccrue to the individual-not
BOX 4
RecentU.S.reformsin stateinsurance
regulation
* Guaranteed
issue(everybodymustbe offereda policy).
* Guaranteed
renewabilityexceptforfraud,nonpaymentof
premiums,
andthe like.
continuity of coverage provisions, includinginsurer limita-
C
tonsonpreexisting
condition exclusionsandwaiting periods.
limitingthe ratingfactorsthat can be used to
ber of ratingbandsbasedon ageor limitson premium dif-
ferentialsacross groups).
* Fulldisclosure
ofinformation.
* Standardized
planbenefits.
Source:Institutefor HealthPolicySolutions1995.
I
the government,employer,or private insurancecompany
(or its subscribers,collectively).Unspent savingscan be
usedfornonmedicalconsumption,passedonas a cashtrans-
fer at the time of death to the deceased'sheirs, or rolled
overfromyearto yearand used forfuturemedicalexpenses
(so that in a senseindividualsare poolingnoncatastrophic
risks over their life cycle).Althoughthere have been few
rigorousevaluationsof the experiencesof U.S. corporations
that have switchedto medicalsavingsaccounts,recently
passedU.S. healthinsurancereformsprovidetax subsidies
to individualsestablishingsuch accounts,makingit likely
that their use willincrease.
Informalrisk-poolingarrangements. Informal, generally
rural, and usuallyvoluntaryrisk-poolingschemesare one
variant of the health insurancemodel common in devel-
opingcountries,particularly-inAsiaand Africa(seeCreese
and Bennett in this volume). These schemes generally
encourageprepaymentof individualpremiumsinto an iden-
tifiablefund, provide somenotion of entitlementto bene-
fits,and work with a defined set of serviceproviders.The
schemesare usuallyorganizedby and often linked to gov-
emnment providers,althoughin somecases theyare run by
communityhealthcommittees.Zaire'sCommunityHealth
InsuranceScheme,Thailand'sHealth Card,andthe United
Nations Children'sFund's (UNICEF) BamakoInitiative
(inmorethanthirtycountries)areexamplesofsucharrange-
ments (see La Forgia and Griffin 1993;Khoman in this
volume; Tibouti). Creese and Bennett (in this volume)
describethe underlyingconceptualbasesforthese schemes
and experiencesin developingcountries.
Out-of-pocket paymentsfor healthservices
Asnoted above,consumerout-of-pocketpaymentsforhealth
servicesare private expenditures.But such expenditures
can take a variety of forms, includingdirect purchaseof
private services,direct purchaseof publiclyprovided ser-
vices (for example,public user charges),and cost sharing
for publiclyor privatelyfinancedservices.When suchpay-
ments are in the form of user chargesforpubliclyprovided
services,they are a public revenue source.When they are
directpaymentsfor privatelyprovidedservices,they are a
privaterevenuesource.(Theconceptualissuesand empir-
INNOVATIONS IN HEALTH CARE FINANCING
30
icalevidenceregardingdirectout-of-pocketpaymentswere
discussedabove.)
Much of the discussionabout health insurancefocuses
on howsuchinsurancecancovermost basicpersonalhealth
servicesneeds. But insurancepoliciesare just as relevant
in other contexts-such as coveringdread diseases,pro-
vidingsupplementalbenefitsto individualscoveredunder
a public system, and coveringlong-term care and social
servicesthat are not coveredby standardhealth insurance
policiesor public systems.These are important issues in
middle-and high-incomecountries.Althougha detaileddis-
cussionof these issues is beyond the scope of this paper,
suchinsuranceneeds to be evaluatedusingthe same crite-
ria (economicefficiency,equity,and administrativesim-
plicity)definedearlier.
Furthermore,it is importantto analyzethe effectsthese
supplementalpolicies have on the insurance policies or
systemsthat they supplement. For example,if a supple-
mental policycoversthe cost-sharingrequirementsof the
primarypolicy,then individualswilluse more of the basic
servicescoveredby the primarypolicy Thus severalcoun-
tries' insurance regulationsforbid private insurers from
fillingin for cost sharingin publicprograms.Other issues
of concernincludehigh administrativecostsand deceptive
marketingpractices,whichareespeciallyproblematicwhen
policiesare marketed to the elderly,who tend to overin-
sure (Cholletand Lewisin this volume).
Much of the abovediscussionhas focusedon the con-
ceptual and operational bases for health insurance and
on policyprescriptionsfor industrialcountries.Developing
countries,however,face additionalchallenges.It is diffi-
cult to provide private insurance in the absence of well-
developedfinancialmarkets. Operatingprivateinsurance
is complex,requiring effectivemanagementinformation
and accounting systemsas well as mechanismsto deter
fraud and abuse. Reinsurancemechanismsand consumer
protectionsare needed. Privatehealth insuranceneeds to
be regulatedby government,yet suchregulationsare com-
plex and may be beyond the capacityof governmentsin
manydevelopingcountries.There are alsopotentiallyhigh
administrativecosts associatedwith insurance,potentially
divertingmoneyfrom health servicesinto health admin-
istration. All these factors must be given careful
consideration.
Charitablecontributions
Becausecharitable contributionscan be directed-to pub-
lic or private institutions, they can be a public or private
revenue source. In industrial countries such contribu-
tions are sometimesencouragedby the tax system (that
is, suchcontributionsare tax deductible),but theyare gen-
erallya smallportion of total health spending. Such con-
tributionsare generallybeyondthecontrol of policymakers,
and cannot be relied on as a stable and long-term source
of financing.
Moreover,there are virtuallyno data on the importance
of this source of financing for developing countries.
Charitablecontributionscan be domesticor foreign.Foreign
charitablecontributionsare the sameas foreigngrant assis-
tance (discussedabove)that has few or no policy condi-
tionalitiesattached.There are no efficiencyor equitycosts
in obtainingthese funds.Domesticcharitablecontributions
do havean opportunitycost-the alernative domesticuses
of the funds raisedby the charity.If such contributionsare
encouragedby the tax system,there willbe efficiencyand
perhaps equitycoststo the economy.
Privatefinancingimplicationsfor developing
countries
Althoughprivatefinancecannot addressequityand public
health issues,it is generallyessentialgiven governments'
limited abilityto marshal sufficientresources to finance
(directlyor throughtransfers)mostpersonalhealthservices.
Yetgiventhe market failuresaffectingboth the demandand
supplysidesof the market for health services,efficiencyin
healthservicesconsumptionand provisionis unlikelyto be
achievedthroughconsumersovereigntyand a laissez-faire
approach to the health servicesmarket. This is especially
true in developingcountries,wherethe bulk ofhealthspend-
ing, especiallyin the poorest countries,is private.Putting
aside issuesof equity,if heavyrelianceon private spending
is necessaryfor financialreasonsor desirableon efficiency
grounds,governmentsshould implementregulationsthat
offsethealth market failures.
Severalgovernmentinterventionsin the privatemarket
forhealth servicescan improvethe efficiencyand equityof
privatefinancing:
A CURMUDGEON'S GUIDE TO FINANCING HEALTH CARE IN DEVELOPING COUNTRIES
31
* Publicregulationof privatehealthinsuranceis essential
to ensure effectiverisk pooling,affordability,solvency,
informedchoice,and continuityof coverage.
In the absenceof a private insurancemarket, govern-
ments in developitg countries may want to consider
establishinga publiclyorganized(or quasi-public)but
privatelyfinancedinsurancesystem.
* Governmentsneedto providethe regulatoryframework
to ensure,quality,efficientpricing,and relevant infor-
mationconcerningboth healthinsuranceand healthser-
viceprovisionand consumption.
Reliance on private financing is a necessity in many
low-incomecountries: As a result governments'abilities
to deal with equityand redistributionobjectivesare some-
what limited. Still, for people who are able to pay, gov-
emient regulation of insurance and public provision of
privately financed health insurance can result in more
equitable and efficient insurance markets and in gains
from risk pooling. Moreover, effective oversight of the
entire deliverysystem (public and private) can result in
higher quality; greater efficiency,and better value for
money for both publicly and privatelyfinanced services
(insurance and out-of-pocket). The difficult questions
are: Given the limited institutional capabilitiesof many
developingcountries,is effectiveregulation of the insur-
ance industryand private providers possible?Can effec-
tive insurance markets be developed given the state of
these countries' financialmarkets? And can administra-
tive costs be kept withinreasonablelimits?These impor-
tant questions need much more attention from the
countries themselvesas well as from development agen-
cies and donor organizations.
RegionalPerspectiveson Health Financing
Reforrn
This sectionsummarizesthe main challengesconfronting
eachregion'shealthcare financingand deliverysystemsand
proposesessentialreforms.These observationsare based
on the epidemiologicalandeconomicinformationpresented
earlieraswellas on the characteristicsof the countriesand
health systemsin each region.Many of the health financ-
ing challengesfacingdevelopingregions can alsobe ana-
lyzedfrom an income perspective(box 5).
Europe and Central Asia
The formerly socialist economies of Europe and Central
Asia face major challenges as they move awayfrom centrally
planned health care systems toward systems based on social
insurance models (see World Bank 1996b; Goldstein and
others 1996; and Klugman and Schieber 1996). The region
contains about 9 percent of the world's population and
accounts for 3 percent of its income. Among developing
regions Europe and Central Asia has the highestpublic share
of health spending-more than 70 percent. Moreover,health
systems account for 7.2 percent of GDP, or $154 per capita,
the highest regional per capita expenditure after Latin
America and the Caribbean.
Yet although Europe and Central Asia is within the mid-
dle-income range, with a per capita GDP of about $1,800,
the transition to market-based economies and the breakup
of the Soviet Union have brought economic crisis to many
BOX 5
Trends in health care financing at different income levels
Low-income countries(per capitahealthexpenditureof less
than$80, withhealthsectorsaccounting for 3-4 percentofGDP).
Govemmentbudgetsfinancemostpublichealthcareexpenditures,
butare limitedby narrowtax basesand weakcollectioncapacity.
Althoughpublicspendingis sometimessupplementedby user
fees, these accountfor a smallpercentageof publicrevenues.
Externalassistancecontinuesto be a significant
sourceof revenue,
especiallyinSub-Saharan Africa,SouthAsia,andthe Pacific. Apart
fromthegovemmentbudget, optionsforformal insuranceschemes
(publicor private)arelimitedbythe smallsizeoftheformalemploy-
ment sector,limitedsavings,underdevelopedfinancialsectors,
andweakinstitutions. Instead,householdsrelyon informal arrange-
ments(extendedfamilies,traditional communitysupportsystems,
NGOs,charitable organizations,ruralcooperatives)to providepro-
tectioninthe eventofcatastrophic illnesses.
Keyissues:Developing
informalrisk-pooling mechanisms(suchas NGO or community
schemes)to expandcoverageto the poor and makingbetter use
of extemalassistance.
Middle-incomecountries(per capitahealth expenditureof
$80-400, with healthsectorsaccountingfor 5-6 percentof GDP).
Middle-income countriesrely on morefinancingsourcesthanlow-
incomecountries,includingsocialsecurityschemesfor civil ser-
vantsand other groupsof formallyemployedworkers.Although
somecountriesprefertoexpandcoverageusingsocialsecuritysys-
tems,others are movingtoward nationalhealthservicemodels
INNOVATIONS IN HEALTH CARE FINANCING
32
countriesin the region.As a result health outcomeshave
declined, and in some countries diseases such as diphtheria
and tuberculosis (both under control before the transition)
are on the rise. Adult health is also a serious problem.
Overnutrition, substance abuse (alcohol, tobacco), and vio-
lence are major public health concerns. Althoughunder cen-
tral planning Europe and Central Asia managed to achieve
universal access in all countries-the only region to do so-
these systems relied on large numbers of low-quality inputs.
Moreover, central planning provided few incentives for
efficiency. And clinical standards are dated, resulting in
inefficient and ineffective treatments for many conditions.
The challenge for the region is to maintain universal
access, increase efficiency, reduce adult mortality, and
improve quality. Over-resourced delivery systems need to
be reconfigured. Clinical practices must be updated. Health
promotion and disease preventionprogramsshould be imple-
mented. Sustainable and efficient financing arrangements
thatrelyon generalrevenues.Thereis usuallyrapidgrowthinpri-
vate spendingas modernprivatehealthservicesexpand;private
insurancebecomesmore common,althoughit (aswellas private
providers)remainlargelyunregulated. Expansion informalemploy-
ment,capitalmarkets,and financial sectors,improvedinstitutional
contextsfor formalinsurance,and urbanizationand other social
changeshelphouseholdsandenterprisesmakethe transitionfrom
informalto formalrisk-pooling mechanisms.Butcountriesinthis
groupstillfallshortof universalcoverage.Rural-urban disparities
and the persistenceof povertycontributeto these problems,and
the multiplicityoffinancingsources,
oftenwithoverlappingorincon-
sistentpolicies,adds to inefficiencies
and inequities.Keyissues:
Developing consistentsocialinsuranceschemes,expandingcover-
ageto ruraland informal sectors,and regulating
the privatesector.
High-incomecountries(per capitahealthexpenditureof more
than $400, withhealthsectorsaccountingfor 6-15 percent of
GDP).Exceptfor the UnitedStates,allindustrialcountrieshave
achieveduniversal
coverage,largely
throughpublicfinancing
(whether
publiclymanagedor publiclymandated). Alsowith the exception
of the UnitedStates,privateinsuranceis used mainlyto supple-
mentthe core servicescoveredby publicfinancing. The Republic
of Koreaanda numberof newlyindustrialized countriesalsohave
attained,or are closeto attaining,universalcoverage.Keyissues:
Containing costs,dealingwithaging
populations,andensuring qual-
ity of serviceandpatientsatisfaction.
must be put in place, and the adverse economicconse-
quences of high payrolltaxes to support social programs
(includinghealthinsurancefunds)mustbe mitigated.Service
provisionshouldbe separatedfromfinance,andincentive-
basedproviderpaymentsystemsshouldbe introduced.The
ongoingtransitionfrompublicfinanceand provisionto pub-
licand privatefinanceand provisionmustbe carefullymon-
itored. With real per capita GDP projectedto grow at 3.7
percent ayearoverthe nextten years,the need fornewgov-
ernance structuresand human development,rather than
financialconstraints,maybe the biggestchallengeto effec-
tive reforms.
LatinAmerica andthe Caribbean
Health systems in Latin America and the Caribbean are
moving toward universalaccess, with mixed public and
private provision and financing(see Burki 1995;World Bank
1996c;and Mediciand others in thisvolume).The region,
which contains 8 percent of the world's population and
accounts for 6 percent of its income, has a per capita income
of some $3,100-the highest among the six developing
regions. Health systems account for an average of 6 percent
of GDP, or $200 per capita, the highest per capita expen-
diture of any developing region.
At 49 percent, Latin America and the Caribbean ranks
fifth among developing regions in terms of the public sec-
tor's share of health spending. Health care is financed by
ministries of health, with social insurance funds playing a
prominent role, although out-of-pocket payments and pri-
vate insurance are also major sources of financing. Most
countries in the region need to work harder to ensure coor-
dination between ministries of health and social insurance
funds on the financing and provisionof services,and achieve
a balance in the financing and delivery of preventive and
curative services. Nutrition, reproductive health, and com-
municable diseases are still major problems in some of the
poorer countries. Still, injuries and noncommunicable dis-
eases account for most of the burden of illness-and receive
most of the attention of policymakers-in nearlyevery coun-
try. In the larger countries there are significant urban-rural
differentials in access and health outcomes.
Universal access, quality, and efficiency are major con-
cerns in most countries. Many countries are reforming
A CURMUDGEON'S GUIDE TO FINANCING
their financing systemsto increase access, generallyby
expandingsocialinsurancesystemsandby encouragingpri-
vate financing.In somecountriesthese effortsare leading
to two-tieredsystemsof care. Fiscalsustainabilityis alsoa
concern,especiallysincehealth and other socialinsurance
funds (suchaspensions)are often commingledwithoutspe-
cific earmarks. A number of countries have introduced
broadlybasedreformsto addressthese issues.Withrealper
capita GDP projected to grow at 2.2 percent a year over
the next decade, the fourth highestof the six developing
regions,manyslow-growth countriesin the regionwilllikely
face significantfinancialconstraintsin improvingaccessand
quality
Middle Eastand North Africa
TheMiddleEast andNorthAfricaregionis extremelydiverse
in terms of economic development, political orientation,
and socialconditions(seeWorldBank 1996d).The region
contains 5 percent of the world's population and accounts
for 2 percent of its income. At $2,700, the region's average
per capita income is second only to that in Latin America
and the Caribbean. The public sector accounts for half of
health expenditures, the fourth-highest share among the six
developing regions. Health expenditures account for 5.2
percent of GDP, or about $120 per capita, the third high-
est among developing regions.
Non-oil and oil-producing countries as well as socialist,
nonsocialist, and statist health systems have very different
approaches. There is considerable political instability in
the region as a result of interregional rivalries. High adult
illiteracy (particularly among women) and poor women's
health indicators are important concerns, especially in the
lower-income countries. Most countries in the region face
dual disease burdens-noncommunicable and communi-
cable. Given the region's high levels of fertility and popu-
lation growth, reproductive health continues to be a major
issue in many countries. The public-private mix in financ-
ing and provision is important in the lower- and middle-
income countries, whilethe wealthier oil-exportingcountries
tend to have national health service systems. Social insur-
ance is not widespread and private health insurance plays
a small, albeit increasing, role in some countries. The qual-
ity of care in the public sector is often a problem.
HEALTH CARE IN DEVELOPING COUNTRIES
33
 Like the region, countries' health care reform agendas
are diverse, though there are commonalties. The poorer
countries still face the unfinished public health agenda of
communicable diseases, with a focus on women's health
and education. All countries are experiencing substantial
increases in their burden of noncommunicable disease.
Efficiency and quality are problems in most countries.
Unfettered and uncoordinated growth in private health care
systems is increasing costs and resulting in two-tiered sys-
tems of care. Appropriately dealing with the public-private
mix in financing and delivery should be a priority. More
equitable risk pooling and financing of public systems is also
important in most of the region's lower-and middle-income
countries. With the end of the oil boom in the mid-1980s
and the globalization of the world economy, many Middle
Eastern and North African countries are facing serious
economic constraints. With real annual growth in per capita
GDP projected at just 0.4 percent over the next decade,
the lowest of the six developing regions, increasing effi-
ciency in financing and delivery will be essential to prevent
deterioration in access and quality.
East Asia and the Pacific
The economies in East Asia and the Pacific are extremely
diverse, accounting for some of the world's largest (China,
Indonesia, Philippines) and smallest economies. The region
contains 30 percent of the world's population and accounts
for 4 percent of its income; per capita income, at about
$1,200, is the fourth highest among developing regions. The
public sector accounts for just over half of health expendi-
tures, the third-highest share among developing regions.
Health expenditures account for 4.1 percent of GDP, or
almost $40 per capita, making it (along with Sub-Saharan
Africa) the second-lowest health expenditure region.
The region has experienced rapid and sustained eco-
nomic growth for manyyears, with most countries approach-
ing or having achieved middle-income status and&market
friendliness.There is confusion about the government's role
in the health sector, however, and largely unregulated pri-
vate financing and deliveryhave grown rapidly,often dupli-
cating public efforts. Many public health services are poorly
funded and inefficiently delivered. Decentralization is
needed, but administrative and managerial capacity at
INNOVATIONS IN HEALTH CARE FINANCING
34
regional and local levels are weak. Traditional medicine is
a significant part of the health system. Clinical skills are
often limited. Most countries face dual disease burdens
(communicable and noncommunicable) but have been slow
to react. Access, universal coverage, efficiency,quality, and
the public-private mix are major concerns in most countries.
Needed reforms in the region span virtuallyevery major
health services issue, including the unfinished public health
agenda,access,quality,costs, financing,and the public-private
mix. Rebuilding the shattered health systems in Southeast
Asia, completing the unfinished health agenda in the poorer
countries and rural areas, and implementing insurance
reforms and effective risk pooling are important issues for
the region.Given the emergingprivatemarket and the empha-
sis placed on individual and family responsibility for health
care in many countries, particular attention needs to be
paid to the public-private mix in delivery and financing and
to user charges and cost recovery. Given the region's prag-
matic governments and the high rates of annual growth in
real per capita GDP projected over the next decade-aver-
aging 6.8 percent, the highest among developing regions-
there are major opportunities for basic health reforms. It will
be particularly important to ensure that these systems absorb
the lessonslearned in other countries (such as inthe Republic
of Korea, with its serious cost-containment problem) as their
incomes rise and their systems are reformed. Ensuring uni-
versalaccesswith an appropriate public-privatemix in financ-
ing and deliveryin an expanding economy creates major risks
for cost escalation as well as for two-tiered systems of care.
Pooling risks through private and social insurance, medical
savings accounts, and other approaches to national health
serviceswill be essential as access imnproves.
South Asia
South Asia also contains some of the world's largest coun-
tries (Bangladesh, India, Pakistan), as well as some of the
poorest. But South Asia differsfrom East Asiaand the Pacific
in several important ways, including its much lower income
level and lower projected economnicgrowth. South Asia con-
tains 22 percent of the world's population and accounts
for 2 percent of its income. At $440, per capita income is
the lowest among developing regions. The public sector
accounts for just 39 percent of all health expenditures, the
lowestpublicshareamongdevelopingregions.Healthexpen-
ditures accountfor 3.7 percent of GDP,or $12 per capita,
alsothe lowestamongdevelopingregions.
Exceptin India,externalassistanceis an importantcom-
ponent of health spending,accountingfor more than 10
percent of expenditures.The region contains the largest
numberofpeoplelivingin povertyamongdevelopingregions,
andfaces a highburdenof diseaseand undernutritionasso-
ciated with poverty.Women'seducationand roles in soci-
ety are importantissues that affect their health. There is
significantinefficiencyin health care systems,both sys-
temwideand in individualfacilities.Publichealth services
are often poorlytargeted in terms of easingthe burden of
diseaseand ensuringcost-effectiveness. Formalinsurance,
both socialand private,is extremelylimited,althoughsome
householdexpendituresare collectivizedthroughinformal
arrangements.As in East Asia and the Pacific,decentral-
izationis an importantissue,but administrativeand man-
agementcapacityis weak.Qualityis poorin both the public
and private sectors,and traditionalmedicineis an impor-
tant componentof the health care system.Significantseg-
ments of the populationlack coverageand access.
Reformsin these countriesshouldfocus on basic pub-
lic health, nutrition, and women's health and education.
Targeteddiseaseprogramsare stillneeded. Given limited
resources,reformsshould focuson the poor, on financing
and deliveringservicesefficiently,and on financialsustain-
ability.Moreover,the privatesector's sizablerolein financ-
ing and deliverymakesthe public-privatemixan important
issue.Given the size of many of these countries, decen-
tralizationis essential, and intergovemmentalfiscalrela-
tionsmust be addressed.Managementcapacityat alllevels
of governmentmust be improved,and qualityin both the
publicand privatesectorsneedsto be assured.Trainingfor
healthcareworkersand managersis needed.Providingthe
regulatoryand policyframeworkfor effectiverisk pooling
for people who can afford private insurance would also
enhance socialwelfareand increasethe efficiencyof pri-
vate financing.Despite myriadbasic healthproblemsand
a low baselineexpenditurelevel,projectedreal per capita
GDP growthof 3.7 percent a year over the next ten years
willallowcountriesin the region to deal more effectively
with the poverty and health systemchallengesthey face.
Effectivemanagementand policymakingwillbe key,
A CURMUDGEON'S GUIDE TO FINANCING
Sub-Saharan Africa
Sub-SaharanAfricancountriesare generallysmalland het-
erogeneous,a resultof tribalismand migration(WorldBank
1994).The regioncontains10percentof the world'spopu-
lation and accountsfor 1 percentof its income.Witha per
capitaincomeof about $775,the regionhas the secondlow-
est income among developingregions.The public sector
accountsfor 54 percentof healthexpenditures,the second-
highestshare amongdevelopingregions.Health expendi-
turesaccountfor4 percentofGDP or $38per capita,ranking
the region(alongwith EastAsiaand the Pacific)as the sec-
ond lowestin percapitahealthexpendituresafterSouthAsia.
The povertyof individualsand the financialsolvencyof
governmentsare major concerns,The smallsize of most
countriesmakes it difficultto achieveeconomiesof scale
in the provisionof certainservices.Externalassistanceis a
major componentof health systemsupport in the region,
accountingfor more than 15percentof health spendingon
averageand morethan50 percentin somecountries.Formal
publicandprivateinsuranceis limited,andriskpoolinggen-
erallytakes the form ofextendedfamiliesandinformalrural
risk pooling (suchas Bamako-typeinitiatives).Basic pub-
lichealthinterventionsfor diseasecontroland prevention,
including the spread of HIV/AIDS, continue to be the
region'smainconcern.The communicablediseaseburden
is the highestin the world and will continue to be for at
least twenty-fiveyearsdespite a shift towardnoncommu-
nicablediseases.Affordingand delivering basicpublichealth
and nutrition servicesare major challenges.There is little
coordinationbetween the public and financiallydominant
private sector. Servicequalityis poor, and the knowledge
base is problematic.Over the next ten yearsrealper capita
GDP is expectedto increaseby just 0.9 percent a year,the
lowest growth rate among developingregions after the
MiddleEast and North Africa.
Giventhis poor economicprognosisand low levelsof
incomeandhealthspending,reformin Africamustbe viewed
over the long term-say, twentyyears.External assistance
willcontinueto be essentialto reform.Althoughsectorwide
reformsare needed,effortsshouldfocuson improvingtar-
geting,ensuringcost-effectivedeliveryofbasicpublichealth
services,and promotingwomen's education and health.
Improvingphysicaland financialaccessto care is a neces-
HEALTII CARE IN DEVELOPING COUNTRIES
35
saryconcomitantto suchefforts.Governancecapacityneeds
to be strengthened,and coordinationbetween the public
and private sectors needs to improve.Althoughmuch of
the policy focus needs to be on communicablediseases
(includingHIV/AIDS),countriesshouldstartaddressingthe
impending noncommunicable disease burden through
focusedpreventionand healthpromotionprograms.Better
risk sharingfor personalhealth servicesthrough Bamako-
typeinitiativesandthroughpublicly(whereaffordable)and
privatelyfinancedinsurancearrangementsshouldbe a priority.
Conclusion
Althoughthis paper has focused on health care financing
in termsof publicandprivaterevenuesources,nationalpoli-
cies must also be concerned with the management and
uses of such funds. Still,strictlyfrom the financingside, a
number of health policyprescriptionshave emergedthat
applyto all countriesand regions:
* Revenue-raising effortsinvolvetradeoffsbetweenequity
and efficiency.
* Administrativecapacityisan essentialcomponentof rev-
enue-raisingefforts.
X Sinceraisingrevenueimposessignificant economiccosts,
governmentsmust maximizethe returns on the uses of
such funds.
, abitytoraisereveueincreasssigni
Agove.ment's
icantlyas incomeincreases.
* The structural characteristicsof labor and industryin
low-incomecountrieslimitthe instrumentsavailableto
these governmentsto raise significantrevenues.
* For reasonsof equity and economicefficiency,broadly
based taxes with fewloopholesshouldbe the preferred
developing
tax approachcountries.
in
taxapproachin developingcountnes.
* Userchargesthat are properlydesignedand focusedon
serviceswith tangiblebenefitsto consumersand service
facilitiescan lead to increasedfinancingand better ser-
vices,althoughuser chargesare notequitableunlessthey
are appliedto income-elasticgoodsand services.
* Publichealthservicesshouldbe publiclyfinancedor sub-
sidized.
* The advantagesof healthinsuranceas a means of pool-
ing risks and the random nature and potentiallyhigh
costsof treatingmanyillnessesmake publicand private
INNOVATIONS IN HEALTH
36
insurancethe preferredvehiclesfor financingpersonal
health services.
* Given these advantages,their limited revenue-raising
capabilities,and the importance of private financing,
governmentsin developingcountries should increase
their institutionalcapacityto ensure the availabilityof
efficientlyrun and privatelyfinancedhealth insurance
to supplementgovernmentefforts.
* Informalrisk-poolingschemes,oftenvoluntaryand spon-
soredby local governments,appear to be viablemech-
anismsfor poolingrisksin poor rural areas.
* Recentinnovationsin managinghealthrevenues-includ-
ingmanagedcompetitionandmedicalsavingsaccounts-
mayhelppromote the efficientuse of resourcesand can
offsetsomeof the problemsinherentin insurancemar-
ketsin countrieswith sufficientadministrativecapacity
and developedfinancialmarkets.
* Priorityneeds to be givento collectinginformationon
publicand private sourcesof health care revenuesand
expendituresin all regions. National health accounts
are essentialfor effectivepolicymaking.
Notes
1.Forexample, thePanAmerican HealthOrganization (PAHO)
iscurrently updatinghealthexpenditure dataforcountries in Latin
Americaandthe Caribbean.Theauthorsaregratefulto Ruben
Suarez,PAHO,forsharingthe earlyresultsofthiseffort.
2. Therearesomedefinitional problemsinthecaseofpublicly
mandatedbut privatelymanagedfunds.Althoughthe national
healthaccountsfor OECDcountriescategorizethesetypesof
fundsunderthepublicdomain,in manydeveloping (mainlymid-
dle-income) countriesthatdistinction
isnotalways clear.Forexam-
ple,privateinsurancein Chileis a privately managedfundthatis
partof a publicly mandatedhealthfinancing system,andis des-
ignatedasbeingprivateinthePanAmerican HealthOrganization's
regionaldatabase(Suarez1997).
3.Foran evaluation framework ofhealthfinancingreformssee
WHO 1995.
4. Othermarketfailuresthat mayjustifygovernment involve-
ment,suchasentrybarriersanddecreasing productioncosts,are
beyondthe scopeof thisdiscussion; seeJonssonandMusgrove
in thisvolumeandHsiao1995.
5. In the caseof certainpurelypublicgoodsfromwhichno
individual canbe excluded fromthebenefit(vectorcontrol,national
defense), individuals
maybeunwillingto purchaseanyofthe good,
sincetheycannotbe excludedfromconsuming it.
CARE FINANCING
 6. In this context insurance simplymeans pooling risks through
a public or private entity that is the recipient of a premium, tax,
or other financial contribution. A national health service pools
risks through collective contributions to finance health services
even though an insurance fund is not specificallyestablished. In
addition, a national health servicedirectly controls the uses of such
funds through direct provision of services, although in many devel-
oping countries social insurance funds also have their own deliv-
ery systems.In most OECD countries with socialinsurance funds,
such funds purchase services from public and private providers
that are not owned by the fund. Similarly,many of the OECD
countries with national health service systems are promoting
facility autonomy and separation of finance from provision.
7. For an in-depth discussion of private health insurance see
Hall 1994 and Chollet and Lewis in this volume. For an analysis
of insurance market failures and potential remedies see Hsiao
1995, pp. 130-34.
8. For a theoretical treatment of insurance see Laffont 1989,
chapter 8.
9. This occurs because pooling a large number of similarrisks
reduces the variance in the probability distribution of the adverse
event occurring. The reduction in variance occurs through the law
of large numbers, which states that the variance for a given risk is
reduced the more times the risk occurs (see Hall 1994, p. 7).
10.People payinsurance premiums both to be relievedof uncer-
tainty about the loss and to be compensated should the loss occur.
Thus there is a value to the individual even if the loss does not
occur (Dorfman 1982, p. 5).
11. For a risk to be insurable it must be important, accidental,
and calculable, with definite losses that are not catastrophic rela-
tive to the size of the pool (Bickelhaupt 1983, pp. 13-14).
12. If individuals do choose their employers for insurance
purposes, then there is clearly an adverse selection problem within
employee groups as well (CBO 1994).
13. Indemnity insurance (in which the individual or provider is
reimbursed on a fee-for-servicebasis by the insurer for the medical
expenses incurred) produces more moral hazard thanmanaged care,
but also has weaker incentives to risk select (Newhouse 1996).
14. Arrow (1963) argues that "competitive insurance markets
willyield optimum allocation when the events insured are not con-
trolled by individual behavior" and "non-market controls, whether
internalized as moral principle or externally imposed, are to some
extent essential for efficiency" (pp. 537-38).
15. See Newhouse (1996) for an interesting conceptual and
policy-relevant discussion of tradeoffs between production effi-
ciency and selection by health plans and providers.
16. Some of these problems have been exacerbated by tax sub-
sidies for employer-provided health insurance; see Hall 1994.
17. Community rating differs from experience rating, in which
premiums are based on the experience of a specific,generallymore
homogeneous, group (such as an employment group).
18. In practice the situation is more complicated because in
many countries general revenues are used to subsidize social insur-
A CURMUDGEON'S GUIDE TO FINANCING
ance funds. Moreover, user charges are also prevalent under both
arrangements.
19. The International Monetary Fund's (IMF) Government
FinanceStatistics often do not include revenue data for regional
and local governments.Althoughwe obtained such data from other
sources where they were absent for large countries, it is possible
that missing data for some small countries biases the ratio down-
ward. This would be a more serious problem in Latin America
and the Caribbean given the region's higher income and govern-
ments' greater ability to raise revenues.
20. Grant assistance and borrowing can be considered private
sources of financing when the entity receiving the aid or loan is a
private entity (nongovernmental organizations, private providers,
private citizens, and so on). Health services can also be financed
by cutting other public expenditures. Moreover, another method
of financing, not addressed in this paper, is to increase efficiency
in the consumption and production of services. Scarce public tax
dollars should be used to purchase services that provide the max-
imum social benefit, not just in terms of health status but also in
terms of economic efficiencyin the production and consumption
of such services.
21. Two other criteria for evaluatingtaxes are flexibility:the tax
systemshould be ableto respond easily(in somecases automatically)
to changed economic circumstances; and political responsibility:
the tax systemshould be designed so that individualscan ascertain
what they are paying so that the politicalsystem can more accurately
reflect the preferences of individuals (see Stiglitz 1988, p. 390).
22. Revenue generation is also sometimes included as a sepa-
rate criterion. The amount of revenues that will be raised depends
on the tax base, tax rates, exclusions, deductions, exemptions,
tax avoidance, and tax evasion. The more recent taxation litera-
ture takes the position that since there are efficiency costs to rais-
ing revenues, for a givenlevel of expenditures governments should
choose the sets of taxes, rates, and so on that maximize that
country's socialwelfare objectives in terms of economic efficiency,
equity,and administrative feasibility.
23. In a partial equilibrium setting (that is, where the tax on
this segment of the economy will not affect any other segment),
the excess burden (B)can be defined using the followingformula:
B =1/2*p*Q*T2
1/TI+ l/e
where P is price, Q is quantity, T is the tax rate, Tl is the elasticity
of demand, and E is the elasticity of supply (Rosen 1995, p. 314).
24. Assuming that there are two commodities, X and Y, that
are taxed at rates Tx and Ty, that the elasticities of supply are
infinite, and that lTxand TIyrepresent the elasticities of demand
for commodities X and Y, the excess burden is minimized when:
Tx/Ty = TIx/TIy
See Rosen 1995, p. 332.
HEALTH CARE IN DEVELOPING COUNTRIES
37
 25. Thereare alsolikelytobe tradeoffsbetween equityandgrowth,
especiallyregardingtaxes on capitalincome.Such taxes maybe prob-
lematic since capital is highly mobile in the global economy
26. For a conceptual and policy-oriented debate on mandates
in the United States see HealthAffairs 2 (spring) 1994, pp. 7-107.
27. Official development assistance is defined as grants and
loans made on concessional terms (that is, having a grant element
of at least 25 percent). See World Bank 1996a, pp. 44-45, and
Feyzioglu, Swaroop, and Zhu 1996, p. 1.
28. Some countries have attempted to finance such borrow-
ing out of the most destructive tax of all, inflation. See Musgrave
and Musgrave 1976, chapters 33-34.
29. Each of the fifty U.S. states is responsible for regulating
privatehealth insurance. However, employer-providedhealth insur-
ance is often exempt from state regulation because of a federal
preemption in the Employee Retirement Insurance Security Act
of 1974 (ERISA), which provides for federal oversight of certain
employer-providedbenefits, including health insurance. Such dual
responsibilityhas precluded states from effectively dealing with a
large element of the market segmentation problem. See Chirba-
Martin and Brennan (1994).
References
Aaron, Henry. 1991. Serious and Unstable Condition: Financing
America's Health Care. Washington, D.C.: The Brookings
Institution.
Abel-Smith, Brian. 1967. An International Study of Health
Expenditures and Its Relevancefor Health Planning. Geneva:
World Health Organization.
Allan, Charles. 1971. The Theory of Taxation.Middlesex, England:
Penguin.
American Academy of Actuaries. 1994. An Analysis of Mandated
Community Rating. Washington, D.C.
Arnett, Grace-Marie. 1996. "Health Care: The Time Bomb Is
Ticking." Paper presented at the Galen Institute Conference
on A Fresh Approach to Health Care Reform, March 25,
Washington, D.C.
Arnott, Richard, and Joseph Stiglitz. 1988. "The Basic Analytics
of Moral Hazard." NBER Working Paper 2484. National
Bureau of Economic Research, Cambridge, Mass.
.1990. "The WelfareEconomics of Moral Hazard." NBER
Working Paper 3316. National Bureau of Economic Research,
Cambridge, Mass.
Arrow,Kenneth. 1963. "Uncertainty and the Welfare Economics
of Health Care." American Economic Review 53: 537-38.
Bickelhaupt, David. 1983. General Insurance. Homewood, Il.:
R.D. Irwin.
Burki, Shahid. 1995. "The Health Gap and the Challenges for
Reform in Latin America and the Caribbean." World Bank,
Latin American and the Caribbean Region, Washington, D.C.
INNOVATIONS IN HEALTH CARE FINANCING
38
CBO (Congressional Budget Office). 1994. The Tax Treatmentof
Employment-Based' Insurance. Washington, D.C.: U.S.
Government Printing Office.
Chirba-Martin,MaiyAnn, andTroyen Brennan. 1994. "The Critical
Roleof ERISAin State Health Reform."HealthAffairs2(spring):
142-56.
Committee on Energy and Commerce, House of Representatives
and Special Committee onAging, U.S. Senate. 1989. Costsand
Effects of Extending Health Insurance Coverage.Washington,
D.C.: U.S. Government Printing Office.
Creese, Andrew, and Joseph Kutzin. 1995. "Lessons From Cost-
Recovery in Health." Discussion Paper 2. World Health
Organization, Geneva. WHO/SHS/NHP/95.5.
Cutler,David, and BrigitteMadrian.1996. "LaborMarket Responses
to RisingHealth Insurance Costs." NBERWorking Paper 5225.
National Bureau of Economic Research, Cambridge, Mass.
Dorfman, Mark. 1982.Introductionto Insurance.Englewood Cliffs,
N.J.: Prentice Hall.
Feyzioglu,Tarhan, VinayaSwaroop, and Min Zhu. 1996. "Foreign
Aid's Impact on Public Spending." Policy Research Working
Paper 1610. World Bank, Washington, D.C.
GAO (U.S. General Accounting Office). 1996. "Health Insurance
Regulation, Varying State Requirements Affect Cost of
Insurance." GAO/HEHS-96-161. Washington, D.C.
Gertler, Paul, and Jacques van der Gaag. 1990. The Willingnessto
Payfor Medical Care.Washington, D.C.: World Bank.
Goldstein, Ellen, Alexander Preker, Olusoji Adeyi, and Gnanaraj
Chellaraj. 1996. Trendsin Health Status,Services,and Finance:
The Transition in Central and Eastern Europe. World Bank
Technical Paper 341. Washington, D.C.
Gruber, Jonathan. 1994. "The Incidence of Mandated Maternity
Benefits." The American Economic Review (June): 622-41.
Hall, Mark. 1994. ReformingPrivateHealth Insurance.Washington,
D.C.: American Enterprise Institute.
Helms,David, Anne Gauthier, and Dan Campion. 1992. "Mending
the Flaws in the Small Group Market." Health Affairs (sum-
mer): 7-27.
Hoff,John. 1996. "Improvingthe System for Delivering Subsidies:
Cap or Scrap the Exclusion?" Paper presented at the Galen
Institute Conference on A Fresh Approach to Health Care
Reform, March 25, Washington, D.C.
Hsiao, William. 1995. "AbnormalEconomics in the Health Sector."
Health Policy 32: 125-39.
IMF (International Monetary Fund). Various issues. Government
FinanceStatistics.Washington, D.C.
Institute for Health PolicySolutions. 1995. StateExperiencesWith
Community Rating and Related Reforms.Washington, D.C.
Khalilzadeh-Shirazi,Javad, andAnwar Shah, eds. 1991. TaxPolicy
in Developing Countries.Washington, D.C.: World Bank.
Klugman, Jeni, and George Schieber. 1996. A Survey of Health
Reform in Central Asia. World Bank Technical Paper 344.
Washington, D.C.
La Forgia,Gerard,and CharlesGriffin.1993.ExtendingCoverage
andBenefitsofSocialFinancing SystemsinDevelopingCountries,
Washington,D.C.: The Urban Institute.
Laffont,Jean-Jacques.1989. The Economicsof Uncertaintyand
Information.Cambridge,Mass.:MIT Press.
Lewis,Maureen.1993."UserFeesinPublicHospitals:Comparison
of Three CountryCase Studies."EconomicDevelopmentand
CulturalChange41 (April):513-32.
Michaud,Catherine,and ChristopherMurray.1994. 'Aid Flows
to the Health Sector in DevelopingCountries."Background
paper prepared for WorldDevelopmentReport1993. World
Bank,Washington,D.C.
- 1996.'Annex5." In Reportof the Ad Hoc Committeeof
WHO: Investingin HealthResearch and Development.Geneva:
World Health Organization.
Moffet,Robert.1993. "Overdosingon Management:Reforming
the Health Care Systemthrough Managed Competition."
HeritageLectures441.HeritageFoundation,Washington,D.C.
Musgrave,Richard.1987."TaxReformin DevelopingCountries."
In DavidNewbery,and NicholasStem,eds.,TheTheoryofTaxation
for Developing Countries.
NewYork:OxfordUniversityPress.
Musgrave,Richard,and PeggyMusgrave.1976.PublicFinancein
TheoryandPractice.NewYork:McGraw-Hill.
Musgrove,Philip.1996.PublicandPrivateRolesin'Health.World
Bank DiscussionPaper 339. Washington,D.C.
Newbery,David,and NicholasStern, eds. 1987. The Theoryof
TaxationforDeveloping Countries.NewYork:OxfordUniversity
Press.
Newhouse,Joseph.1996."ReimbursingHealth Plansand Health
Providers:Efficiencyin ProductionVersusSelection."Journal
of EconomicLiterature34 (September):1236-63.
Nichols,Len. 1996. Personalcommunication.Urban Institute,
Washington,D.C.
OECD (Organization for Economic Cooperation: and
Development).1977.PublicExpenditures on Health.Paris.
1985.Measuring HealthCare,1960-83.Paris.
1992.TheReformof HealthCare:A Comparative Analysis
of SevenOECDCountries.Paris.
- 1993. OECD Health CareSystems:Factsand Trends,
1960-91.Paris.
- 1994. The Reformof HealthCareSystems:A Reviewof
SeventeenOECDCountries.Paris.
1995.NewDirections in HealthCarePolicy.Health Policy
Studies7. Paris.
PAHO(PanAmericanHealthOrganization). 1996.NationalHealth
ExpenditureandFinancing in LatinAmericaandthe Caribbean:
Challengesfor the 1990s.Washington,D.C.
Pauly,Mark.1986."Taxation, HealthInsurance,andMarketFailure
in the MedicalEconomy"Journalof EconomicLiterature24
(June):639-40.
1996. 'An Efficientand EquitableApproachto Health
Reform."Paper presentedat the Galen InstituteConference
A CURMUDGEON'S
GUIDETOFINANCING
HEALTHCAREIN DEVELOPING
39
on A Fresh Approach to Health Care Reform, March 25,
Washington,D.C.
Poullier,Jean-Pierre. 1992. "AdministrativeCosts in Selected
IndustrialCountries."HealthCareFinancingReview (summer):
167-72.
Ray,Anandarup.1975. "Cost RecoveryPoliciesin Public Sector
Projects."PolicyPaper206. WorldBank,Washington,D.C.
Rosen,Harvey.1995.PublicFinance.4th ed. Chicago:Irwin.
Saltman,Richard,and Joseph Figueras.1996.EuropeanHealth
CareReforms: Analysisof CurrentStrategies.
Copenhagen:World
Health Organization.
Stiglitz,Joseph. 1988. Economicsof the PublicSector.2nd ed.
NewYork:WW.Norton.
Suarez,Ruben. 1997.Personal communication.Pan American
Health Organization,Washington,D.C.
Tanzi,Vito.1987."QuantitativeCharacteristicsofthe TaxSystems
of DevelopingCountries."In David Newbery and Nicholas
Stern, eds., The Theoryof Taxationfor DevelopingCountries.
NewYork:OxfordUniversityPress.
Thirsk,Wayne.Forthcoming."The Substanceand Processof Tax
Reform in Nine Developing Countries." In World Bank,
ReformingTaxesin DevelopingCountries.Washington,D.C.
Tibouti,Abdelmajid."TheBamakoInitiative:Experienceto Date
with CommunityFinancing in Poor Rural Areas." United
Nations Children'sFund, NewYork.
Van Doorslaer,Eddy, and AdamWagstaff.1995."Equityin the
Financeand Deliveryof Health Care in Europe."Paperspro-
ducedfor the EuropeanUnion'sBiomedProgramme.Erasmus
University,Rotterdam.
Van Doorslaer,Eddy,AdamWagstaff,and F.F.H.Rutten. 1993.
EquityintheFinance andDeliveryofHealthCare:AnInternational
Perspective.Oxford: Oxford UniversityPress.
WHO (WorldHealth Organization).1995. "Health Financing
Reform: A Framework for Evaluation." RevisedWorking
Document WHO/SHS/NHP/96.2. Division of Analysis,
Research,and Assessment,Geneva.
. 1996a.Equityin HealthandHealthCare:A WHO/SIDA
Initiative.Geneva.
. 1996b. Investingin Health Researchand Development.
Geneva.
WorldBank.1987.Financing HealthServices
inDevelopingCountries:
An Agendafor Reform.Washington,D.C.
. 1988.WorldDevelopment Report1988:Opportunitiesand
Risks in Managingthe WorldEconomy.New York: Oxford
UniversityPress.
. 1991.Lessonsof Tax-Reform. Washington,D.C.
- 1993.WorldDevelopment Report1993:InvestinginHealth.
NewYork:OxfordUniversityPress.
.. 1994. Better Health in Africa:Experienceand Lessons
Learned.A Developmentin Practicebook. Washington,D.C.
. 1996a.Global EconomicProspects and the Developing
Countries1996.Washington,D.C.
COUNTRIES
 . 1996b.'AHealthSectorAssistance
StrategyfortheEurope
and Central AsiaRegion."Europe and Central AsiaRegion,
Washington,D.C.
- 1996c."Investingin Human Developmentin the Latin
America and Caribbean Region." Latin America and the
CaribbeanRegion,Washington,D.C.
1996d. "MiddleEast and North AfricaRegionHealth,
INNOVATIONS
IN HEALTHCAREFINANCING
40
Population,and NutritionSectorAssistanceStrategy."Middle
East and North AfricaRegion,Washington,D.C.
. 1997.WorldDevelopmentIndicators
1997.Washington,D.C.
. Forthcoming.ReformingTaxesin DevelopingCountries.
Washington,D.C.
Zelman,Walter.1994."The RationaleBehindthe ClintonHealth
ReformPlan."HealthAffairs1 (spring):9-29.
Government Financing
of Health Care
BengtJonssonand PhilipMusgrove
JIn a marketeconomypricesservethreefunctions:they
guide the allocation of resources,they ration scarce
goods and services,and they finance the paymentof
compensationto the factorsof production.Whenthe mar-
ket providesthese functions,the question of how to pay
for a particulargood or servicedoes not arise:consumers
buy the product in amounts that are determined by their
wants and capacity to pay, and producers deliver those
amounts,with prices equilibratingdemand and supply.
There are severalreasonsdirect paymentby consumers
for health care is inefficientand unequitable,creatingthe
needforgovernmentinterventionsandforalternativemech-
anismsto allocateresourcesand financing.Thispaper con-
siders the financing mechanism-that is, the role of
governmentfinancingof health care. Other issuesrelated
to governmentintervention,such asthe optimalallocation
of resourcesin the provisionof health care, are discussed
onlyif they directlyrelate to governmentfinancing.
In consideringthe alternativesfor governmentfinanc-
ing of health care, we focus on three questions: Should
governmentpay for health care?Does it matter which of
two generalmodels-social insuranceand directfinancing
from generalrevenues-is used forpublic finance?Which
taxesshouldbe usedto financehealthcare,anddo the types
andlevelsoftaxes matterforcoverage,benefits,andexpen-
ditures?Nearlyallthe availableempiricalevidencerelevant
to these issues is for OECD countries; thus the conclu-
sions do not necessarilyapply to developingcountries.
Whereverpossible,the situationin developingcountriesis
discussedseparately.
BengtJ6nssonis professorof healtheconomicsat the StockholmSchoolof Economics.PhilipMusgroveis principaleconomistin the
Human DevelopmentDepartmentat the WorldBank.
41
Relationsbetween Financeand Provision
of Health Care
There is no necessaryconnection between the way that
health care is paid for and the way that it is delivered;in
particular,public financedoes not implypublic provision.
The argumentsfor and againstpublic provisionof health
servicesare verydifferentfromthose relatedto publicand
privatefinancing(J6nsson 1996).It is commonto distin-
guishthreerelationsbetweenfundersandprovidersofhealth
care: the reimbursement, contract,andintegratedapproaches
(OECD 1995).Theserelationsare largelyindependentof
specifictaxes and other sourcesof funds.
Under the reimbursementapproach,providersreceive
retroactivepaymentsfor servicessupplied.Thesepayments
maybe billed directlyto insurersor to patients,who may
be partly or entirely reimbursed by insurers. The reim-
bursementapproach,often coupledwithfee-for-service pay-
ment arrangements,can be found in systemswith multiple
private and public insurers and multiple(usuallyprivate)
suppliers,asin the UnitedStates.In low-and middle-income
countriesit is rare for the reimbursementmodelto be com-
bined with public finance.Chileis an exception,with part
of government financing reimbursingprivate providers
retroactively.
The contract approachinvolvesan agreementbetween
third-partypayers(insurers)andhealthcareprovidersaimed
at greater control over total funding and its distribution.
This approachtends to be found in socialinsurancesystems
with predominantly private (nonprofit) providers.
Prospectivebudgetsare combinedwithper diem,casemix
(diagnosticrelatedgroup,or DRG),andfee-for-service pay-
ments. A variant of this system is used in Brazil, where
budgets are set by the state or municipalityand providers
arepaidunderaDRGtariff(Lewis1994).Preferredprovider
organizationsin the United States alsouse the contractual
approach.
In integratedhealth systemsthe same agencycontrols
boththe fundingandthe provisionof healthservices.Medical
personnelare generallypaid salaries,and budgets are the
main instrumentfor allocatingresources.Integrated pub-
licsystemsareusedin the Nordiccountriesanduntilrecently
were the model in the United Kingdom,and are the com-
mon organizationalform for ministriesof healthin devel-
oping countries. In many such countries the integrated
approachis alsoused for socialsecuritysystems,whichhave
their ownhospitalsandclinics,althoughthere are oftenalso
contractualrelationswith privateproviders.Health main-
tenance organizations(HMOs) in the United States are
examplesof integratedprivate systems.
Most health care systemsincludeelementsof all three
systems,justasmosthavea mixofmodelsforpublicfinance.
There also havebeen significantchangesover time. Italy
and Spain'spublic healthcare systemshavemovedfrom a
contract approachto an integrated system,while those of
New Zealandand the Uriited Kingdomhave movedfrom
an integratedsystemto a contract approach.
Theremaybea trendtowardtwotypesof relationbetween
fundersand providers (J6nsson1996;van de Ven,Schut,
and Rutten 1994). The first type involvesa (near) public
monopolyin health care funding, through taxes or com-
pulsorysocialinsurancecontributions,andcompetitivecon-
tractswithprivateandpublicproviders.Thusfinancingand
provisionmaybe separated,in what is sometimesreferred
to asa purchaser-provider
split.Thesecondtypeisan inte-
gratedmodelwithcompetitionbetweendifferent
integrated
systems(HMOs).
In the first type consumersusuallyhaveno (or limited)
choiceof insurer,but do have a choiceof provider.In the
secondtypethere is a choiceofinsurer,but oncethis choice
is made the consumeris tied to the providerslinkedwith
that insurer.No healthcare systemin the worldoffersa freeMoney
choice of both insurer and provider to everyone in the
population.
INNOVATIONS
IN HEALTHCAREFINANCING
42
The role of governmentdiffersbetweenthese two orga-
nizationalmodels. In the first type the governmentmust
raisemostof the necessaryfunds,andthen contractwiththe
providers.In the second type the governmentmust regu-
latecompetitionbetweeninsurers,orthird-partypayers,and
distributethe publicsubsidiesneeded to guaranteeuniver-
sal accessto a certainlevelof healthcare.Determiningthe
criteriaandmechanismsforthe distributionof risk-adjusted
subsidiesis amajorproblem(Newhouse1994;vanVlietand
van de Ven 1992).Suchsubsidiesare usuallydifferentiated
by the consumer'sage and sex and by region,taking into
accountthe incomeof the populationserved(whichaffects
the demand for services)and the costs of providingcare.
This approachhas beenusedin Chileforprimaryhealthcare
(whichis a municipalresponsibility)and is being imple-
mented in Argentinaand Colombia.
Insurance is central to any discussionof health care
finance. And while there are markets for many kinds of
insurance,healthcare insuranceis peculiarbecauseof the
nature of the asset being protected-human healthrather
than nonhumancapital(Musgrove1996).The introduction
of insuranceforhealthcare,whethervoluntaryand private
or publiclyfinanced, has consequencesnot only for the
distributionof paymentsfor health care, but also for the
allocationof resourcesto andwithinthe healthcaresystem.
Byintroducinga third partythat collectsrevenueand pays
providers,healthcareinsurancechangesthe relationbetween
consumersand providersof health care (unless,as in the
integratedmodel, insuranceand provisionare combined
in a singleagency;figure1).Thecrucialimportanceof insur-
FIGURE
I
Economic relationsinthe fiance anddelivery
of healthcare
Money(directpayments)
ofcare
Consumers ofcare
Providers
(patients) Healthservices (docors,hospitals)
Insurance Claims
coverage
Money
pes (government) budgets)
anceandthe problemspeculiarto it provideoneof the ratio-
nalesforpublicfinancingof care,andraisemostof the issues
discussedhere about how best to payfor it.
Shouldthe Government Pay for Health
Care?
One majorreasonfor publicfinancingof healthcare is the
provision of public goods, such as programsfor medical
research, healthpromotion, vector control,and food and
water safety.Such public goods provide benefits that are
sharedby manypeople,regardlessof whether theypayfor
them. Thus entirelyprivatemarketswould yieldan ineffi-
cient allocationof resources, and governmentfinancing
(or some other nonmarketalternative)is needed to opti-
mize allocation.The situation is essentiallythe same for
goodswith externalities(suchas vaccinations),whichcan
be produced and consumed privatelybut whose conse-
quences-good or bad-affect other consumersregard-
less of whether they chooseto consumethose goods. For
example,immunizationof part of the populationalsopro-
tects the unimmunized.
However,most health care interventionsproduce pri-
vate goods,with benefitslimitedto individualconsumers.
Argumentsfavoringpublic finance of purelyprivate per-
sonalhealth care expendituresdiffer from thosefor public
goods,fortworeasons.The firstisthe needto financehealth
care for the poor-that is, peoplewho cannot affordwhat
societyconsidersan adequateamountof healthcare either
out of pocket or by buyinginsurance.The second affects
the entire population,and derivesfrom imperfectionsin
insurancemarketsthat preventthemfromprovidingan effi-
cient and equitableallocationof health care resources.
The relativeimportanceof these two domains-subsi-
dies for the poor and insurancefor people who can help
financeit-depends on a country'sleveland distribution
of income,and explainsmuch of the differencein health
care outcomes between rich and poor countries. Many
governmentshave become the main insurer for personal
health care,particularlyin high-incomecountries,and this
involvementbecomes the quantitativelymost important
reason for governmentsto raise money for health care.
Because of the complexityof the insurancemarket, and
the fact that both efficiencyand equityare involved,such
GOVERNMENT FINANCING
43
interventionraisesthe questionsmentionedearlier-whether
to establish public insurance,how to organizeit, which
sourcesto use to financeit, and with what decisionsand
consequencesfor coverage,benefits, and spending.
Financing
of individuol(personal)
healthcoreexpendi-
tures
There are threewaysto financeindividualhealthcare: pri-
vate individualpayments,private collectivepayments,and
public finance.
Privateindividualpayments.These payments are also
called out-of-pocketcosts. The problem with direct indi-
vidualpaymentsforhealthcareis that medicalexpensesare
sometimesso high that even peoplewith higherthan aver-
age incomes cannot afford them. This problem is aggra-
vatedbythe factthat increasedhealthcarecostsmaycoincide
with reduced incomedue to the health problems.In such
a situationpersonalsavingsmaybe inadequate and oppor-
tunities to borrow for investmentsin health are limited.
Moreover,peoplewith the lowestincomes,whooften need
care the most, will be excluded from much care if it is
financedby directpayments.
This problem does not precludedirect individualpay-
ments for health care-which usuallyaccount for 10-50
percent of all payments-but it does call for protection
againsthigh costs.That meanspayingfor health care col-
lectively,and sharingthe financialrisk. (There are no gen-
eral reasons for governmentto finance smallamounts of
medicalexpenditures.The socialbenefitsare small-all but
the verypoor can affordsome medicalexpenditure-and
the socialcosts maybe high.)
Privatecollectivepayments(insurance).Because many
illnessesoccurrarelyand seeminglyat random,healthcare
expendituresare uncertainaswellaspossiblyhigh.Bypool-
ing a large number of people,insurancereduces the vari-
abilityof their incomesnet of medicalexpenditure.Health
expendituresmaybe highlyvariablefor a givenmemberof
the pool, but averageoutlayscan be predictedfairlywell.
Thus insurancereduces financialrisk for consumerswho
are risk averse (that is, who have a diminishingmarginal
utilityof wealth or income) and lowershealth risks since
OF HEALTII CARE
care is more accessible. Financial risk is usually not elimi-
nated because coinsuranceand deductibles are used to make
the insured person share the costs (see Chollet and Lewis
and Creese andBennettinthisvolume).Apolicymayrequire
that the insured pay the first $200 of healh care costs out
of pocket each year (deductible) and then pay 20 percent
of all charges (coinsurance).
This cost sharing is one way to control moral hazard-
the increased use of services and reduced precaution in tak-
ing care of one's health that results when risk pooling leads
to reduced marginal costs for services. Moral hazard can
manifest itself in two ways,one static and the other dynamic.
People with health insurance tend to see doctors more often
and to use costly treatments even if the benefits are small
(Pauly 1968; Zeckhauser 1970). Doctors also may change
their behavior, particularly in fee-for-service systems. Since
costs are not borne by the patient, it is easier for doctors to
suggest more expensive treatments. The dynamic effect of
moral hazard is the incentives it creates to introduce new
medical technology for which there would be no market in
the absence of insurance (Weisbrod 1991). Both problems
derive from the inability of the insurer to monitor service
providers and the insured.
Insurance firms incur costs for doing business such as
processing claims and marketing. These are called loading
costs,and they generallymake competitive private insurance
more costly to administer than uniform public insurance.
Many of these costs arisebecause insurance companies have
an incentive to exclude high-risk consumers or to at least
identify them so that they can be charged more, but have
trouble identifying which risk class people belong to. In
the short run this situation of asymmetric information-
consumers who know their risks better than the insurer
does-may benefit high-risk people who, if they know they
are likely to need medical care, will be eager to buy insur-
ance. This tendency of the highest risks wanting the most
insurance is called adverseselection.If insurance companies
compensate for it by raising premiums, some low-risk per-
sons may decide not to buy insurance. This can lead to a
vicious circle in which only high-risk people remain.
There are severalwaysto reduce moral hazard and adverse
selection, although there is no complete solution for com-
petitive insurers in a situation where information is asym-
metric and imperfect (Pauly 1974;Zeckhauser 1970; Spence
INNOVATIONS IN HEALTEI CARE FINANCING
44
and Zeckhauser 1971; and Mirrlees 1971). The optimal
insurance contract is a second-best, nonlinear solution
with a mix of risk spreading and incentives such as a mod-
erately high deductible and a diminishing coinsurance rate
(Blomqvistforthcoming). An alternative is to include 'bonus
options" (Zweifel 1992) or rebates in the event that the
insured does not submit any claims during the year, with
the rebate increasing in subsequent years without a claim
until a maximum is reached. This approach provides first-
dollar coveragebut still provides incentives to reduce moral
hazard. Few insurance contracts have these features, but
many include suboptimal provisions such as an annual ceil-
ing on copayments. Integrating insurance with service pro-
vision is another alternative, and removes the incentive for
providers t6 overtreat since they then bear the financial risk.
Public finance: governmentas insurer Moral hazard is a
problem in any insurance system, but adverse selection
and the attempts of insurers to counter it by excluding poten-
tial consumers and adjusting premiums are peculiar to pri-
vate insurance. This is perhaps the main argument in favor
of public insurance, which can more easily be made uni-
versal and in effect force everyone to share the risks. Public
insurance is also often justified by some related problems-
of free riders, of excluded population groups, and of col-
lective risks that are largelyindependent of individual risks.
In a voluntary insurance system people can choose not
to insure. This is not a problem if the uninsured can be
ignored when they need medical care but cannot pay for it.
If they are taken care of anyway-that is, allowed to "ride
free"-the incentive to have insurance is reduced. It is dif-
ficult to judge how important this problem is. In Switzerland
most families have insurance despite the fact that it is not
compulsory in all cantons. Cultural tradition probably plays
a large role; and if most people make an effort to take care
of themselves, there is room for generosity to those who do
not. It is often suggested that private insurance be made
compulsory in order to avoid the free-rider problem. But
doing so would raise another problem-which sanction to
use for people who do not comply-and in any case is infea-
sible for poor populations.
Exclusionis the opposite of the free-riderproblem. People
may want insurance but cannot buy it because of lowincome
or high risk. One solution may be to give them a voucher
(subsidy)so they can buy privateinsurance.The practical and in particularhowpublic financeforhealth care affects
problemsof calculatingand administeringsuch subsidies the privatemarket.
maybe considerable,makingpublic insurancesimpler-
again,particularlyin countrieswith largepoor populations. Publicfinance:government subsidizesbut doesnot insure.
The elderly,with highrisksfor illnessand related expendi- It is also possiblefor governmentsto finance health care
tures, pose a particularproblem,and some countrieshave without actingas insurers,by subsidizingprivateinsurance
created public insurancejust for them. The need to pre- throughthe tax system.Employersoften pay a significant
pare for health expensesin old age can be partly solved portionofworkers'healthinsurancepremiums:in the United
through a funded systemin whicheachinsuredpaysinto a States,for example,about 80 percent of the premiumsfor
fund that coversfuture needs (in contrastto a pay-as-you- privatehealthinsurancearepaidbyemployers(Phelps1986).
go system,where each person's contributiongoes toward If employersare allowedto deduct these costs from the
the current expensesof allmembers). income on whichthey pay corporate taxes but employees
This approachalsohas its problems.Knowledgeabout are not taxed on the value of the premium-that is, the
the future incidenceand prevalenceof illnessand poten- cost of the insuranceis not treated as incometo eitherthe
tial treatmentsis limited.Thus it is difficult,if not impos- companyor theworker-then the insuranceispartlyfinanced
sible, to calculatethe premiumsthat 20-year-oldsshould by a subsidyor "taxexpenditure"equivalentto the tax that
payfor healthcare that theywillreceivein fiftyyears.One the governmentdoesnot collect.The same situationoccurs
solutionto such collectiverisksis for the governmentstep if individuals'private health insurancepremiumsare tax
in as a re-insurer.The problemwith life-longinsurancecan deductible, or if employerspay directly for health care
be seen in countrieswhereprivateinsurancefundsgobank- (self-insurance).This kind of public subsidyis not used
rupt when their membersbecome older, and have to be much outsidethe UnitedStates,andthereare few estimates
mergedinto funds with youngermembers.(This problem of its cost. But in Brazilin the early 1980s it appears to
alsoaffectspay-as-you-go publicinsurance,evenin middle- haveaccountedfor$1billionin healthcare spending,about
income countries,as the populationages.)An alternative a quarter of what the governmentspent (Lewis1994).
is for voluntaryinsuranceto be restrictedto a certainage Payingforhealthcarewithemployment-related taxdeduc-
group, for examplebelow65. The governrmentmust then tionscan solvesome of the problemsof privateinsurance,
finance care for people over that age.But with increasing in that adverse selectionis limitedby contractingin large
life expectancyand the concentrationof costlyillnessat groupsrather than one personat a time, loweringadminis-
advancedages, this approachmeans that the government trativecosts.However,this approachintroducestwo other
willend up payingfor the bulk of health care. problems.First, employeesmaynot recognizethat theyare
It is important to distinguishbetween actuariallyfair payingfortheirinsurance,atleastpartly,throughlowerwages.
insurance,as providedthrough risk pooling,and govern- Thisleadsto higherthanoptimalinsurancecoverage(overin-
ment socialinsuranceprograms.Actuariallyfair insurance surance)and thushigherhealthcareexpenditures.The con-
is providedthroughmarketsin whichbuyersvoluntarilypay sequencesfortotalhealthcareexpenditurescanbe substantial.
forprotectionagainstinfrequenthighmedicalexpenditures Phelps (1986)estimatesthat employergroup health insur-
whoseprobabilitiescan be statisticallydetermined,withpre- ancepremiumsin the UnitedStateswouldbe about45 per-
miumsadjusted accordingly(seeChollet and Lewisin this cent lowerif the tax subsidywere not in effect,eventhough
volume).Sodal insuranceprogramsare providedby gov- marginaltax ratesare only25-35 percent.There arealsowel-
ernment,often involvean income transferbetween popu- farelossesdueto employmentchoicesandwagelevels,which
lationgroupsforreasonsunrelatedto health,havea defined areaffectedbythe subsidizedexcessinsurance(Feldmanand
set of eligibilityrules, and are partly or whollyfinanced Dowd 1991; Feldstein1973; Manningand others 1987).
throughtaxesor compulsoryinsurancepremiumsthat need Second,the sizeof the subsidyincreaseswith the marginal
not be actuariallyfair.These differencesraise questions tax rate; if taxes are progressive,the highersubsidiesgo to
about the best combinationofprivateand publicinsurance, peoplewith higherincomes,whichis inequitable.

GOVERNMENT FINANCING OF HEALTH CARE

45
Totalhealthexpenditures
andgovemment
shares
Totalandpublicfinancingofhealthcarein differentregions
are shownin table 1.The publicshareof healthcareexpen-
ditures is at least 50 percent in everyregion exceptAsia.
The shareis highestin rich countries,whichalsohave the
highesttotal expenditures.Privatefinancingdominatesin
low-incomecountries,and direct out-of-pocketpayments
are more important than privateinsuranceas a sourceof
revenue.
A similar,though less clear,picture emergesfor OECD
countries(table2). The countrieswith the lowestincomes
alsohave the lowestshares of public finance.The United
States is an exception,having the lowestshare of public
finance of all OECD countries.By contrast, the govern-
ment providesnearlyall healh care resourcesin Iceland,
Norway,Sweden,and the UnitedKingdom.
Sinceannualpublicspendingon healthtypicallyreaches
$1,000-2,000 per capita in rich countries,the questions
of how to organizeand pay for care are highlyrelevant.In
low-income countries the amounts at stake are much
smaller,both absolutely-$10 per capita per yearin many
Africancountries,and less than $100in most of Asiaand
Latin America-and as a share of income,but it is harder
to raise revenue and a larger share of the population is
too poorto affordanysignificantamountof privatehealth
care.A better wayto comparerich and poor countriesmay
be to estimate subsistenceincomeand comparetotal and
publichealthspendingwith the remainingnonsubsistence
or discretionaryincome. On this basis public efforts to
financehealthcare in poor countrieswould probablylook
larger than in OECD countries,makingissues of organi-
zation and sources of revenue just as important, if not
more so.
Which Model for PublicFinance?
As noted earlier,there is a basic distinctionbetween tax-
based (directlyfinanced)andinsurance-basedpublichealth
care systems.In tax-basedsystemsgeneral revenuetaxes
are the main sourceof finance,and the governmentusu-
allyactsasthe mainproviderofhealthcare.Insurance-based
systemsare financedmainlythrough payrolltaxes,up to a
ceilingonwagesatwhichpointthe marginaltaxrate becomes
zero.The number of insuranceagenciesvariesby country,
from one in most Latin American systems,to several in
Europe, to more than 200 in Argentina.Directlyfinanced
systemsalsoprovidethe protectionthat characterizesinsur-
ance, but the insuranceis implicit,and individualsneed
not be explicitlyaffiliatedto receivebenefits.
There is also considerablevariationin the connection
between finance and provision.Social securityinstitutes
operate their own clinics and hospitals in most Latin
Americancountries,but contractingwith privateproviders
is the norm in Argentina,Brazil, and most of Europe. In
most cases these institutes overseenonprofit institutions
and independentphysicians.Different systemsuse differ-
ent methods to reimburse providers for services;some-

TABLEI
Globalhealthcareexpenditures
byregion,1990
Totalhealth Health Publichealth
Shareof expenditure expenditureas expenditureas Shareof GNP Per capitahealth
world population (billionsof percentageof percentageof spent on health expenditure
Region (percent) U.S.dollars) world total regionaltotal (percent) (U.S.dollars)
OECD countries 15 1,483 87 60 9.2 1,860
Transition
economiesof Europe 7 49 3 71 3.6 142
Developing countries 78 170 10 50 4.7 41
LabnAmericaandthe Caribbean 8 47 3 60 4.0 105
MiddleEastandNorth Africa 10 39 2 58 4.1 77
OtherAsiaand islands 13 42 2 39 4.5 61
India 16 18 1 22 6.0 21
China 22 13 1 59 3.5 11
Sub-SaharanAfrica 10 12 1 55 4.5 24
Worid 100 1,702 100 60 8.0 329
Source:WorldBank1993.

INNOVATIONS IN HEALTH CARE FINANCING

46
TABLE2 In practice, the distinctionbetween the two modelsis
Publicandper capitahealthcarespending,
OECD not so sharp as this discussionsuggests.A tax-based sys-
countries tem can obtain part of its financingfrom the socialinsur-
Publicfinancing Total health ancesystem(asin Sweden),and part of the servicescanbe
as percentage spending GDP providedby privatecontractors(for example,generalprac-
Country of total percapitae percapitaa titionersin Denmarkandthe United Kingdom).
Australia 69 1,606 18,970 Moreover, social insurance systems often receive com-
Austia 63 1,965 20,216
Belgium 88 1,653 20,184 plementary financing from general revenues (as in
Canada 72 2,010 20,608 Switzerland), and the insurance systemcan be supplemented
Denmark 83 1,362 20,784 with other services del paid operated the
Finland 75 1,357 16,362 o1rectl for and by
France 78 1,866 19,169 government (for example, the Veterans Administration in
Germany 74 1,869 19,720 the United States). In extreme cases such as Brazil and
Iceland 84 1,577 19,402
Ireland 76 1,201 15,202 Costa Rica, the distinction between the two systems ceases
Italy 71 1,561 18,698 to exist because coverage is extended to (nearly) all non-
JMexico 58 14395 27,484 contributors, and wage taxes become just one of the sources
Netherlands 78 1,641 18,570 of revenue. This extension of coverage, whichwas large and
New Zealand 77 1,226 16,424 rapid in both countries during the 1980s, effectivelymerges
Norway 95 1,604 21,980
Portugal 56 938 12,313 social insurance with the existing tax-financed system.
Spain 79 1,005 13,572 Changesin other countries have also blurred the dis-
Sweden 83 1,348 17,435
Switzerland .- 72 2,294 23,961 Tax-based systemshave introduced a split between
tinction.
Turkey 58 223 5,271 purchaser and provider, as in the United Kingdom, and
UnitedKingdom 84 1,211 17,560 insurance-based systems have introduced global budgets
UnitedStates 44 3,516 24,629 in order to control costs, as in France and Germany. In the
Note:Dataarefor 1994or latestavailable
year.
a. Expressed
inpurchasingpowerparitydollars. Czech Republic a purely tax-based system has been sup-
Source:
OECDHealth Database 1996.
-plemented with social insurance funds that are supposed
times it is fee for service, but contracts with negotiated to compete for clients, with a general fund acting as the
prices are more common. Another difference between the insurer of last resort. But the use of general revenue taxes
two models is that while tax-based systems are alwaysadmin- has not disappeared, because the government now buys
istered directly by the financing agency (usually a ministry insurance for people who do not contribute through what
of health), the administration of social insurance is some- is effectivelya wagetax. Because this change was introduced
times taken care of by independent government-regulated together with a move from salaried providers to largelyfee-
bodies (such as Germany's Krankenkassen and France's for-service medicine with inadequate regulation, one of
mutuelles). the results has been an explosion of costs. That outcome,
Dependence on wage taxes means that social insur- however, is not specific to a social insurance or a tax-based
anceis paid forby workers(and possiblyemployers)rather system.The same logic,of competingbut regulatedinsur-
than the general public. Tax-based or direct systems, by ers financed partly by a uniform wage tax and partly by
contrast, are in principle universal. From a strictly financ- subsidies for the poor and the unemployed from general
ing perspective, the same problem has to be solved whether taxation, characterizes current reforms in Argentina and
the financing mechanism is taxes or compulsoryinsur- Colombia.
ance contributions: the state obtains involuntary payments Even when the two models remain distinct, any country
from members of society, and these payments are unre- can use both at once. This is the typical Latin American
lated to individual benefits. In terms of coverage, effi- model, with both a ministry of health and a social security
ciency, and equity, however, the two systems can differ institute, where in principle the institute serves only its
substantially. contributors but the ministry offers services to the entire

GOVERNMENT FINANCING OF HEALTH CARE

47
population.In practice the users of ministryservicesare
determinedby the qualityof servicesand by the existence
of privateinsurance,as wellas by coststo consumers,and
the same people may draw on a varietyof providers and
sourcesoffinance-payingout of pocketfor drugsand inex-
pensiveconsultations,using public facilitiesfor hospital-
ization, and even covering some services with private
insurance.This complexityalsocharacterizesSouthAfrica,
wherethe tax-basedsystemfinancescareforthe poormajor-
ity,socialinsurancecoversmiddle-and high-incomework-
ers, and there is alsoprivate insurance.Providersin this
systemtend to workwith onlyone financingsystemand so
are almostas sharplysegregatedas the sourcesof funding.
Suchoverlappingpatternsmakeit difficultto describe"cov-
erage" consistently,and the coexistenceof two or more
financingsystemsmakesit almostimpossibleto judgewhich
(if any)of them worksbest.
Publicspending
pottemsunderthe two models
Asnoted, the public share of healthspendingrisesand the
private share falls as country income grows, and (except
for the United States) rich countriesare more alike than
poor countries in this respect. The situation is much less
homogeneouswhen the comparisonis between countries
withtax-basedandsocialinsurancesystems.Andlow-income
countriesare more similarthan middle- and high-income
countriesin sharesof spendingon the two models.
The shareof pubic healthcare spendingthat isfinanced
directlyor by taxes (that is, not through social insurance)
in low- and middle-incomecountries for whichestimates
are availableis shown in table 3. In many countries all
publicspendingis tax-basedand socialinsurancedoesnot
exist; nearly all these countries are quite poor. This is a
natural consequenceof a small formal employmentsec-

TABLE3
Publichealth care financingin variouscountries, circa 1990
(percent)
Shareof public
healthspending Publicfinancing Totalhealth
financeddirectly as percentage Income spendingas
Region/country or bytaxes of total percapitaa percentageofGDP
Africa
Sub-Saharan
Cameroon 68 38 2,400 3
Ethiopia 100 61 370 4
Kenya 56 63 1,350 4
Madagascar 76 50 710 3
Malawi 62 58 800 5
Mozambique 100 75 600 6
Nigeria 100 44 1,360 3
SierreLeone 52 71 800 2
Tanzania 100 68 570 5
Zimbabwe 65 52 2,160 6
Asia
Bangladesh 100 44 1,160 3
Bhutan 100 620 2
China 0 60 1,680 4
India 21 22 1,150 6
Indonesia 67 35 2,730 2
Korea,Rep.of 12 41 8,32.0 7
Malaysia 100 43 7,400 3
Myanmar 100 3
Nepal 44 49 1,130 5
Pakistan 100 53 1,970 3
PapuaNewGuinea 100 64 1,830 4
Philippines 80 50 2,440 2
SriLanka 83 49 2,650 4
Thailand 90 22 5,270 5

INNOVATIONS
IN HEALTHCAREFINANCING
48
tor, which makes it difficult to raise significantrevenue
from wages and salaries.Even where such taxes can be
collected, they are often pooled with other tax revenues
rather than used separately to finance social insurance.
(Socialinsurance schemeslimited to civil servants are a
more commonexception,andthese existevenin somepoor
countries.)There arefewcountriesin whichallpublichealth
spendinggoesthrough socialinsurance;evenin countries
that relyheavilyon insurancethere is usuallysomedirect,
tax-based expenditure-if only for those public goods
that cannot be associatedwith individualsand therefore
cannot be insured.
Thecountriesthat relymoston socialinsurance-Bolivia,
China,the CzechRepublic,Denmark,Israel,the Republic
of Korea, Mexico,the Netherlands,Sweden-are a het-
erogeneous group. In some, social insurance dominates
becausealmosteveryonecan paywagetaxes and it is pub-
lic policyto use that model to financehealth care. In oth-
ers, socialinsurancecoverageis lower but those covered
havehigher than averageincomes,and tax-basedfinance
may cover as much or more of the population but cost
verylittleper person.In general,sharesof financeare unre-
lated to shares of coverageor utilizationof services; on
the contrary,whenboth systemsoperate in a country,social
insuranceusuallyspends more per person. This outcome
simplyreflectsthe fact that formalemploymentpayswages
well above the average income, at least in low-income
countries.
At middle and high incomes there is great variationin
howpublicspendingis divided.In somecountriesonemodel
or the other dominates,whilein others substantialpublic
resourcesflowthroughboth the directand the socialinsur-
ance channel.Both systemsare used in Australia,Latin
America,somecountriesin EasternEuropeand the Middle
East,andthe UnitedStates.There aretworeasonsto believe
that thisis an inefficientwayto organizegovernmentfinance

TABLE3
Publichealthcarefinancinginvariouscountries,
circa1990(continued)
(percent)

Shareof public
healthspending Publicfinancing Totalhealth
financeddirectly aspercentage Income spendingas
Region/country or by taxes of total per capital percentageof GDP
EasternEurope.
MiddleEast,and NorthAfrica
CzechRepublic 5 85 6,280 6
Hurgary 84 83 6,080 6
Turkey 58 38 4,840 4
Tunisia 63 67 4,690 5
LatinAmnenco
ondtheCaribbean
Argentina 45 61 5,120 10
Bolivia 17 29 2,170 6
Brazil 50 43 5,240 6
Colombia 43 57 5,460 5
CostaRica 19 82 5,100 9
DominicanRepublic 77 34 3,080 6
Ecuador 62 63 4,140 4
ElSalvador 32 30 2,110 6
Guatemala 61 33 3,180 5
Jamaica 100 35 3,670 9
Mexico 15 56 7,170 5
Nicaragua 100 62 2,550 8
Panama 41 60 4,910 9
Paraguay 31 25 3,420 4
Peru 59 34 3,130 3
Uruguay 81 76 6,670 8
Venezuela 75 47 8,120 4
a. Expressed
inpurchasing
powerparty
dollars.
Source:
Musgrove1996.

GOVERNMENT
FINANCINGOF HEALTHCARE
49
of health care, and that it is advantageous to have a single At least in high-income countries, coverage by publicly
government program. First, it is easier to control the total financed care is unrelated to whether one or both models
flow of resources if there is only one channel (Reinhardt are used or in what proportions, because coverage is usu-
1992). Second, having one channel reduces the risk of ally almost universal. Among OECD countries the only
suboptimization through incomplete coordination between exceptions are Mexico (whichhas much lower income than
the different systems; for example, Sweden's combination the other members), the Netherlands (where high-income
of an open-ended reimbursement system for prescription people do not pay the contributions but buy private insur-
drugs and globalbudgets for hospital care promotes a trans- ance), and the United States (which has two substantial
fer of costs from inpatient to outpatient care. To avoid public programs, one of each type, limited to the elderly
such problems, governments must take a comprehensive and the poor). Coverage is harder to estimate in poorer
view of their role in financing health care. countries, and there may be some relation between bene-
The share of public financing in total financing differs ficiaries of public finance and which model or combination
among different kinds of health expenditures. The shares of models is followed; data are insufficient to support any
of drugs and hospital care in total health spending in OECD conclusion on this point.
countries are shown in table 4. Such detailed data are scarce
in developing countries, but the public share in hospital Features and failingsof the two systems
financing is generally high and the public share in paying
for drugs is rather low.The lower is the share of public expen- The great variation among countries, and even within them,
diture in the total, the more it is likely to be concentrated strongly suggests that neither the tax-based nor the social
on hospital care, independent of whether the financing is insurance model is a systematically superior way of paying
tax-based or socialinsurance. A large share of out-of-pocket publidy for health care. It is easyto list the theoretical virtues
spending typically goes for drugs, particularly where pri- of either system: for example, in principle direct, tax-based
vate insurance is nonexistent or covers only a small share finance is easier to extend to everyone, and social insur-
of the population. Differences in reimbursement for dif- ance makes people (at least those with formal employment)
ferent health care services can encourage the use of ser- contribute proportionally to their ability to pay. But these
vices-such as hospital care-that are not necessarily the potential advantages are inconclusive without an examina-
most cost-effective. tion of the specific taxes used to finance direct payment,
TABLE4 or the coverage and benefits under one scheme or the other.
Publicexpenditures
for drugsandhospitalcare, And the theoreticalbenefits of one model are lost when
selectedOECD countries, 1983 and 1993 the two forms of financing are used together, which is com-
(percentage of total spent on drugs and hospitalcare) mon. This is whythere is no discerniblepattern to public
Drugs Hospital
care health care financing in high-income countries. The greater
Country 1983 1993 1983 1993 homogeneityamonglow-incomecountriesthat dependon
direct finance results from the difficult task of raising suf-
Austria 29 63 42 36
Belgium 54 60 66 68 ficient revenue from wages and salaries to finance care for
Canada 7 27 88 86 anybut a smallshare of the population.
Denmark 44 49 100 100 One complexity of mixed systems is that there are many
France n.a. 62 n.a. 91
Germany 68 61 85 85 partial public health insurance programs in various coun-
iceland 64 69 100 100 tries that are limited to different groups or different treat-
Italy 72 49 85 85
Netherlands 62 94 84 84 ments. Such programs illustrate several issues in the choice
NewZealand 80 66 95 n.a. of publicfinancingmechanismsand other characteristicsof
Sweden 72 69 na, n.a. public payment for health care. The best-known subsystems
Unted Kingdom 65 63 n.a. n.a.
UnitedStates 7 12 53 57 are Medicare and Medicaid in the United States. Medicare
Source:
OECDHealthDatabase
1995. is a uniform federal program that provides compulsory

INNOVATIONS IN HEALTH CARE FINANCING

50
hospitalinsuranceto the elderly(partA) aswellas optional
supplementarymedicalcoverageto whichnearlyallelderly
subscribe(part B). Medicaidis a programoperatedby the
states, with matchingfederal dollarsbut different criteria
and benefitsfrom one stateto another,that financeshealth
care forpoor householdsthat alreadyreceivebenefitsfrom
either or both of twowelfareprogramsunrelatedto health.
Someother countriesalsolinkpreferentialhealthinsurance
for the poor to an existingnon-healthwelfareprogram,to
avoid the need for another targetingmechanism(Grosh
1992).
Contributions. MedicarepartA is financedthrougha 1.45
percentadditionto eachworker'ssocialsecuritytax forboth
employersandemployees.Thusit isclearlysocialinsurance,
differingfrom most suchprogramsin other countriesonly
in its limitationto the elderly.PartB is voluntaryandfinanced
by premiums,just as with private insurance,except that
about 75 percent of the premiumis subsidizedwith gen-
eral taxation-which makesit a directlyfinancedsystem.
Users pay a deductibleand coinsurance,whichin part
A increaseswith increased expenditures.This is contrary
to the recommendationderivedfrom the theoryof optimal
insurance,of a largedeductibleand decreasingcopayments
to provide protection against catastrophic financial risk
rather than "first-dollar"coverage(Arrow1963).In part B
copaymentsare constant,which alsoleavesconsumersat
severefinancialrisk.This lackof catastrophiccoveragehas
created a market for "Medi-gap"privateinsurance,so the
system operates with all three forms of collectivefinanc-
ing,whichis inefficient(Blomqvistand Johansson1996).
Medicaidisfinancedby stategeneraltaxationwithmatch-
ing federal grants. It has no deductibles or proportional
coinsurance,but some nominalcopayments.There is cat-
astrophicprotection,but no incentivesto controlmoralhaz-
ard or costs.
Benefits.Eligibilityfor Medicaredependson age or, for
somenon-elderly,on healthstatus (suchas permanentdis-
ability).In additionto beneficiariesof the two welfarepro-
grams, Medicaid enrollment can be extended to other
medicallyneedygroups.Thuseligibilitydependson income
and wealth.As a result patients may become eligiblefor
Medicaidbyspendingdowntheir assets,and sohavea dis-
GOVERNMENT FINANCING OF HEALTH CARE
51
incentive to save and an incentive to dispose of wealth
throughgifts. Medicaidcoversnursinghome costs,which
are not coveredby Medicare. People who need nursing
homecare,andwhodo not haveprivatefundsor insurance,
must shift to Medicaidfor coverage.
Providerreimbursement. Medicare and Medicaid were
enacted withina systemof privateinsurance,and initially
copied the way private insurancecompaniesreimbursed
providers.However,unexpectedlargecost increaseshave
led to changes in the way providersare paid, by making
them share part of the financialrisk through diagnostic
relatedgroup-basedpaymentforhospitalization andthrough
capitation(HMO enrollment).Proposalshavebeen made
to introduce global budgets to control doctors' fees. In
Germanysuch budgetshavebeen used since 1976for hos-
pital care, ambulatorycare, and drugs, and in the United
Kingdomthe systemof indicativedrug budgetsis based on
the sameidea.
Asthis summaryindicates,non-universalpublichealth
finance may respond to important needs and attempt to
combineequitywith control over expendituresas part of
overallefforts to resolvehow govemmentfinanceshealth
care.In everycountrythe elderlyand the poor are likelyto
be the neediestgroups,althoughoften in differentsenses
of "need"-the elderlyneedmorecare,independentof their
financialstatus, and the poor need more help payingfor
care,independentof theirhealthstatus.It ispartlythe desire
to satisfydifferentkinds of need that makesthe choiceof
revenuesources and financingmechanismsso important
and complicated.
Do the Type and Level of Taxes Matter?
The two mainissues concerningthe combinationof taxes
(and compulsorysocial insurance contributions)used to
finance health care are equity and efficiency.For equity,
the importantdistinctionis betweenfinancingand the use
of services.For efficiency,there is similarlya difference
betweenhowrevenueis raisedandhowit isused.The choice
of taxes maycreate inefficiencyoutsidethe health system,
throughwelfarelossesassociatedwith obtainingthe funds;
or inside the system,through the effects of taxes on the
behaviorof providersand consumers.
Generalprinciples
onhealth care andfiscalpolicy
Since health care is just one of many uses for public rev-
enue,it isimportantto considerwhetherpublicfinancethe-
oryoffersanyprinciplesabouthowto raiseand spendmoney
forhealth.Thisquestionis applicablewhetherhealthis paid
for from the same budget as all other uses, or there are
sub-budgetsdefined by specifictaxes that are earmarked
or assignedto specificexpenditures.
The benefitprinciple.The benefit principlestates that,
unlesstaxationis explicitlyused for redistribution,the cost
of anypublic serviceshouldbe borne by those who use it
(Musgrave1959).This principleis similarto what happens
in a market-people payin proportionto their consump-
tion-and can be appliedto allindividualor privategoods
in healthcare.It is harderto applyto publicgoods,forwhich
individualconsumptionis poorlydefined and in any case
difficultto measure. If it is assumedthat everybodycon-
sumes the same amount, then people should pay accord-
ingto theirmarginalvaluationof the publicgood.However,
it is highlyquestionablewhether evenpublichealth activi-
tiesand emergencyservicescan be consideredpublicgoods
of whicheverybodyconsumes the same amount. If con-
sumersare heterogeneous,everybodyshould theoretically
payan individualtax. The informationneeded for suchan
approachsimplyis not availableat any cost, however,so
the benefitprinciplecannotbe followedexactlywherepub-
licgoods are concerned.
Since the incidenceof diseaseis to a large extent ran-
dom and everyoneis at risk of needing health care, the
benefitprinciplesuggeststhat the tax base for healh care
financingshouldbe as broad as possibleto ensure that no
oneescapescontributing.Thisis an argumentagainstfinanc-
ing health care with, say,tobacco or alcoholtaxes, which
havea rather narrowbase.The fact that there maybe exter-
nalitiesassociatedwithbehaviorssuchas drinkingor smok-
ing is an argument for taxes to correct the behavior,but
thesetaxesmaybeusedto financeothergovernmentexpen-
ditures.
Theability-to-payprinciple.Atthe oppositeextremefrom
the benefit principle(payingin proportionto one's use of
a service)is the notion that people should pay in propor-
INNOVATIONS IN HEALTH CARE FINANCING
52
tion to their ability,whichin practicelargelymeansin pro-
portion to their income.To the extent that the need for
and use of healthcareare not highlycorrelatedwithincome,
this principle favors subsidiesfrom the rich to the poor,
and probablycorrespondsto the most widelysupported
notionof howto financehealthcare(vanDoorslaer,Wagstaff,
and Rutten 1993). The notion of financingaccordingto
abilityto pay can be interpreted in terms of both vertical
equity-individuals with a greater abilityto pay ought to
paymoreregardlessof theirhealth-and horizontalequity-
individualswith the same abilityto pay ought to pay the
sameregardlessof theirhealth.Becauseincomedifferences
canbe large,verticalequityandprogressivityin healthcare
financinghave been studied more thoroughly.The pro-
gressivityof a health care financingsystem refers to the
extentto whichpeople's paymentsfor healthcare rise as a
share of their incomewhen their incomerises.
Regardlessof how it is financed,any insurancesystem
transfersresourcesfromthe healthyto the sick,but it does
not necessarilymakethe richsubsidizethe poor-depend-
ing on the sourcesof revenueand patterns of health care
utilization,the oppositecan happen.It is difficultto design
healthfinancesystems,partlybecauseboth types of subsi-
dies usuallyseem desirable for equity reasons but they
need not be consistent.And neither is consistentwith the
benefit principle(payingfor what one consumes).
Earmarking, or linkingspecificrevenuesto specificexpen-
ditures.For the most part there is no reasonto relate a spe-
cifictax to a specificarea of publicexpenditure.It is more
rationalto decide the optimal levelof governmentspend-
ing and to decidehow best to raise the necessaryrevenue
withoutanyexpenditurebeingtied to anysource(Musgrave
1959).There is nothingunique to health about the choice
of financingmechanism:the same considerationsapplyto
all areas of governmentspending. The choice of the tax
base-which taxes or contributions,at what levels and in
whatproportions-is a problemofpublicfinance,not health
economics.
However,it is commonfor socialinsurancesystemsto
be financedprimarilyby earmarked taxes on wages and
salaries,wherethe contributionis thought of as payingthe
insurancepremiumeventhoughthe premiumdiffersaccord-
ing to income (rather than representing a market price).
Moreover,in recent yearsthere has been renewedinterest Statesin 1980,5 percent of the populationaccounted for
in earmarked (hypothecated)taxes for healthcare (Jones 55 percent of healthcare expenditures(table5). In a sim-
and Duncan 1995).This interesthas been sparkedby con- ilar studyfor Francein 1980-81,5 percent of the popula-
cernsabout underfundingof health systems,as a result of tionconsumedtwo-thirdsof healthcare expenditures,while
the generalpublicfinanceproblemsthat mostgovernments 25 percent had no healthcare consumptionat all (OECD
experience. The immediate questions are whether ear- 1987).A Swedishstudy on the consumptionof prescrip-
markingin fact makesmore funds available,and whether tion drugsfound that 3.8 percent of consumersaccounted
it has other effects on the equity and efficiencyof health for half of all costs,and that one-third of the population
care spending. had no costsfor prescriptiondrugs.
These large differencesin per capita expendituresare
Equityaspectsofpubliclyfinancedhealthcore relatedto ageandthe risk ofmortality.In Sweden,forexam-
ple, annualper capita health care expendituresare closely
Definingequityis not easy Appliedresearch on equity in linked to whether a patient died during the year or sur-
the deliveryof health care often focuses on horizontal vived(table6).This trend is indicativeof twohealthspend-
equity-that is,whetherpeoplewith equalneeds (in terms ing characteristicsin high-incomecountries:the tendency
of morbidity)are treated the same (in terms of utilization) to consumemore health care with advancingage (at least
regardlessof income (Wagstaffand van Doorslaer 1993). after infancy)and the tendencyto have verylarge expen-
A major argumentfor public financingof healthcare is to dituresin the finalyear of life.
guaranteeeveryone,includingpeoplewho are too poor to In retrospectthisdistributionof spendinglooksextremely
buy private health insurance or to pay out-of-pocketfor inefficient,because some people die despite having con-
health care, access to a minimumpackageof health ser- sumedlarge amounts of care. But it is uncertaintyabout
vices.Peoplewith the lowestincomesand wealthare gen- outcomes,not simplywastefulness,that leadsto this result.
erallyin the worst health and thus have a greater need for Whetherthe distributionis equitableis hard to say,but the
services.One reason is that both health and income are increasingexpenditurewithagedoesreflectincreasingneed
often low amongthe elderly.Thus equityaspectsare cen- as health deteriorates.In poor countriesneither tendency
tral to discussionsof health care financingfor individual is likelyto be so marked,becausefewerresourcesare avail-
services. ableandpeopleare morelikelyto die withoutreceivingsub-
The equity argumentmaybe less importantfor financ- stantialcare.
ingofpublicgoods.Interventionsto improvepublichealth, The distributionof health care expendituresis neither
suchas sanitationandvaccinationprograms,generallyben- evidenceof inequitynor a measureof servicebenefits.If it
efit the entirepopulation.Dependingon the program,the werepossibleto examineconsumptionoveralifetimeinstead
benefitsmaybe more or lessimportantfor somegroups.It of withina singleyear,the variationbetweenpeoplewould
can be arguedthat the poor generallygainthe most, since
people with more moneycan compensateby substituting TABLES
private for publicexpenditures.On the other hand, health States, 970c1977 anda1980
promotionactivitiesmay be of particularbenefit for peo- (percentage
of totalexpenditures)
ple withmoreeducationandgreatercapacityto benefitfrom
such activities.Such groups are often the first to change Population
expense
group 1970 1977 1980
theirbehaviorto improvetheirhealth,aswithreducedsmok- TopI percent 26 27 29
ing and increasedphysicalexercise. Top2 percent 50 38 39
Top 10percent 66 70 70
Distributionof healthcareexpenditures
in thepopulation. Top30 percent 88 90 90
Top50 percent 96 97 96
The distributionof healthcare expendituresin a givenyear Bottom50 percent 4 3 4
is highlyconcentratedamonga fewconsumers.In the United Source:
Berk,Monheit.
andHagan
1988.

GOVERNMENTFINANCINGOF HEALTHCARE
53
be much smaller,at least so long as they were of the same
age.Whatwereallyneedto knowto answerquestionsabout
equityand efficiencyis the lifetimeconsumptionof health
care and the resultinghealth status at differentages, and
howthisdiffersaccordingto incomeand other factors.But
improvementin health status over time and rapid techno-
logicalchangein healthcaremakesuchcomparisonsmean-
inglessfor people of differentages.
Income-relatedequityin the use of services.Anotherway
of lookingat the distributionof healthcare expendituresis
to lookat consumptionin differentincomebrackets.Astudy
of the United Kingdomand the United Statesfound that
about one-tenth of health care spendingoccurs in each
income decile in both countries (table 7). This is rather
surprisinggiventhe differencesin healthcarefinancingand
provisionbetweenthe two countries:both incomedistrib-
ution andaccessto healthinsurancecoverageare lessequal
in the United States.
In other words, in rich countrieshealth care consump-
tion is almostindependentof income,thanksto near-uni-
versal coverage of insurance and especially of public
insurance.In low-and middle-incomecountrieswith a less
evendistributionof income,the lowestdecilesare unable
to affordmuchhealth care, and insurancecoverageis typ-
icallymuchless complete.For both reasonsthere are usu-
allysizabledifferencesin healthcarespendingbetweenhigh-
and low-incomeeamers.The poor are naturallymore sen-
sitiveto prices than the rich,whichgreatlylimitstheir use
of full-costprivate services(Gertler and van der Gaag 1990)
and makes them more dependent on public subsidies.
Althoughsuch subsidies are seldom enough to equalize
health care consumption across income levels, they can
transfer substantialamountsof real incometo the lower-
incomedecileswhen coverageby publicprogramsis high.
Van Doorslaerand Wagstaff(1993)used two methods
to measureincome-related inequityin ten (mainlyEuropean)
countries.The first method involvedranking individuals
byincomeandcomparingthe cumulativehealthcareexpen-
diture (standardizedfor differencesin morbidity)across
income groups.The second method controlled for mor-
bidityby using regressionanalysisto test for significant
income effects on healh care received.They found that
income-relatedinequityexistsin most countries,and that
it usuallyarisesfrom the effectof incomeon the amountof
care receivedby peoplewho use at least somehealth care.
In fact, the effect of income on the probabilityof seeking
care was significantin only one country (Denmark).This
result almost surely does not apply to developingcoun-
tries,whereincomeis a majordeterminantof whetherpeo-
ple obtain care-not only becauseof the priceof care, but
becausepoor people often livefar from the nearest doctor
or clinic (particularlyin rural areas),and travel time and
cost are major deterrentsto seekingcare (Gertler and van
der Gaag 1990).
Data on the distributionof healthcare expendituresin
the populationare importantto understandingthe oppor-
tunities and limitationsfor financinghealth care through
directpayments,and forassessingthe need forgovernment
TABLE7

TABLE
6 Distribution
of healthcare expenditures
Annualhealthcare expenditures
per capita in (consumption)by incomedecile,UnitedKingdom
differentageandmortalitygroups,Sweden,1994 and UnitedStates
(kronor) (percent)

Died during Incomedecile United Kingdom United States

Agegroup Survivors the year I 9.1 8.9
2 11.0 10.3
0-4 9,900 410,000 3 11.2 8.8
5-14 4,000 133,000 4 10.4 9.6
15-44 7,000 186,000 10.3 9.5
45-64 9,300 174,000 6 9.5 9.2
65-74 16,000 141,000 6 9.5 9.3
75-84 22,000 93,000 8 8.8 10.8
85 andabove 23,000 61,000 9.7 11.7
Total 9,200 106,000 10 10.4 11.9
SOUdata.
Source: Source:
Gottschalk,
WoHfe,
andHaveman
1989.

IN HEALTHCAREFINANCING
INNOVATIONS
54
support to health insurance, to compensate for differences tions; such contributions are negligiblein Denmark, Ireland,
in age-related needs and in incomes. They are less infor- Portugal, and Switzerland. The balance between direct
mative about the appropriate taxes to use to finance pub- and indirect taxation is less varied.
lic health care, although the fact that expenditures usually The distinction among direct taxes, indirect taxes, and
rise with age carries some important implications for how social insurance contributions is important, but further
social insurance is financed, particularly as to whether it is distinctions are possible. Direct taxes can be applied at the
funded or pay-as-you-go.Systemswith pay-as-you-gofinanc- central, regional, and local levels. Indirect taxes can be
ing are vulnerable to changes in age composition as the pop- general, such as a value added tax (VAT),or on particular
ulation ages; funded systems escape that that problem but goods, such as an excise tax. Social insurance contribu-
are vulnerable to cost-raising technical change in medi- tions can be paid by employers or employees and be pro-
cine. To have a better idea of which taxes are most appro- portional or have an upper limit (and thus be regressive).
priate, we need to examine the taxes that are used and who Different taxes and their design have different consequences
pays them. and must be assessed against the objectives of health care
arndfiscal policy.
Distribution of income and tax payments:progressivityof Kakwani progressivityindexes for thirteen countries are
financing. The way health care is financed affects people in shown in table 9. Direct taxes are the most progressive,
a number of ways, and the distributional consequences are and indirect taxes are regressive, in all thirteen countries,
difficult to assess.An increasein direct payments mayreduce and are especially regressive in Spain and the United
the use of health care, which probably has a greater effect Kingdom.Socialinsurance-financingisregressivein Germany
on the poor, who are sick more often. Higher direct pay- and the Netherlands (whichexplainswhy total public finance
ments also affect the distribution of disposable income, and is slightlyregressivein those countries) but is otherwise pro-
thus other types of consumption. Depending on how tax- gressive, although much less so than direct taxes. Private
payers are affected, increased public financing can also have direct (out-of-pocket) payments are strongly regressive, as
several consequences. is to be expected since medical care needs are largelyinde-
In the absence of consensus on how much more the pendent of ability to pay. Because it spreads risk across
better-off should pay than the worse-off, health care sys- income groups, private insurance is less regressive or even
tems can be judged by the progressivity of the taxes used progressive, so total private payments are more equitably
to finance them. A variety of indexes have been pro- distributed than those paid out of pocket.
posed to measure progressivity (Lambert 1989), the most
common of which is Kakwani's index (1977). This index Mix of publicfundingin OECDhealthcaresystems
measures the extent to which a tax system departs from (percentage
of total publicfunds)
proportionality. The cumulative proportion of the popu-
lation, ranked according to pretax income, is plotted Public
financing
against the cumulative proportion of tax payments to asa share
obtain the tax concentration curve. A zero index means Direct Indirect Social Total of total
the tax is exactly proportional to income; positive values County taxes taxes insurance public (percent)
indicate progressivity (the rich pay a larger share of taxes Denmark 58 42 0 100 83
France' 0 3 97 100 78
than their share of income) and negative values, regres- Ireland 38 52 10 100 76
sivity. Italy 25 28 47 100 71
Detailed data on the mix of funding sources are avai- Pbrtugal 239 64 91 100 576
able for only a small number of countries (table 8). There Spain 10 8 82 100 79
are large differences among countries in the composition Switzerland 78 18 4 100 72
Unted Kingdom 44 36 20 100 84
of the taxes used to finance health care. France, the United States 52 15 33 100 44
Netherlands, and Spain rely on social insurance contribu- Source:
vanDoorslaer,
WagstafandRutten1993.

GOVERNMENT FINANCING OF HEALTH CARE

55
TABLE
9
Progressivityof componentsof health care financing(Kakwani indexes),selectedcountries

Direct Indirect General Social Total Private Direct Total Total

Country,year taxes taxes taxes insurance public insurance payments private payments
Denmark,1987 0.062 -0.113 0.038 0.000 0.038 0.031 -0.265 -0.241 -0.003
Finland,1990 0.128 -0.097 0.056 0.090 0.060 0.000 -0.246 -0.246 0.017
France,1989 0.000 0.000 0.000 0.094 0.094 -0.186 -0.228 -0.218 0.012
Germany,1988 0.251 -0.092 0.112 -0.081 -0.040 0.093 -0.103 -0.022 -0.037
Ireland,1987 0.267 n.a. n.a. 0.126 n.a. -0.021 -0.147 -0.096 n.a.
Italy,1991 0.161 -0.112 0.038 0.112 0.075 0.177 -0.077 -0.057 0.045
Netherlands,1992 0.200 -0.089 0.071 -0.129 -0.100 0.083 -0.038 0.043 -0.070
Portugal,1990 0.218 -0.035 0.060 0.185 0.072 0.137 -0.242 -0.229 -0.045
Spain,1990 0.214 -0.152 0.048 0.050 0.049 -0.012 -0.212 -0.190 -0.003
Sweden,1990 0.053 -0.083 0.036 0.010 0.030 -0.240 0.027 0.029
Switzerland,1992 0.172 -0.072 0.131 0.038 0.113 -0.270 -0.403 -0.319 -0.165
UnitedKingdom,1992 0.284 -0.152 0.046 0.187 0.079 0.077 -0.223 -0.092 0.052
UnitedStates,1987 0.192 -0.065 0.124 0.019 0.090 -0.175 -0.461 -0.288 -0.131
n.a.isnotavailable.
Source:Wagstaff,
vanDoorslaer,
andothers1996.

Overall, the way health care is paid for in these coun-
tries is nearly proportional to incomes. Since consumption
of care is relatively independent of income, there is usually
a substantial net transfer from the rich to the poor when
benefits are compared with payments. The exception is
Switzerland,where a large share of care is financed privately
and regressively. Otherwise, the results do not seem to
depend much on the exact combination of private finance
and taxes.
These findings probably do not all carry over to develop-
ing countries. Out-of-pocket spending in these countries is
undoubtedlyregressive,as inhigh-income countries-in fact,
it is generally more so, since it may be the only form of
spending for the very poor, who do not have access to pub-
lic subsidies. Private insurance spending is highly progres-
sive because only the rich buy it, directly or through their
employers. Indirect taxes, which the poor do not escape, are
slightlyregressive.Direct taxes are not used much in devel-
opingcountries because of the ease of evading and difficulty
of collectingthem; thus they may be less progressive than in
industrial countries. Social security payments can be regres-
sive or progressive, since both coverage and the incomes to
which they apply varygreatlybetween countries. Thus wage
taxes are probably progressivewhen coverage is low (10-20
percent), even though the rich do not pay them, because only
high-income workers are covered. They become less pro-
gressiveor even regressive as coverageis extended, although
even then the very poor do not pay them.
Efficiencyaspectsof publiclyfinancedhealth care
As with equity, efficiencyis a concern for how revenues are
raised and how they are used. Put another way, the type
and level of taxation used to finance health care may have
effects both inside and outside the sector.
Excessburden of taxation. Since people prefer not to pay
taxes, and since they usually have choices about employ-
ment and consumption that affect how much tax they pay,
taxes affect economic behavior and hence the allocation of
resources. This impact is referred to as the excessburden of
taxation. The implication is that public financing comes at
a price. All taxes other than a lump-sum tax are associated
with a welfare loss. Thus there is a tradeoff between effi-
ciency-which generally calls for a small tax burden-and
equity-which calls for progressive subsidies and thus for
a larger share of income taken in taxes.
The optimal tax structure is one that maximizes society's
welfare, where the balance between deadweight loss and
equity reflects attitudes toward the competing goals of
efficiencyand equity. It is possibleto estimate onlythe dead-
weight loss; there is no theoretical basis for deciding how
much a particular improvement in equity is worth paying
for. Although the theory of optimal taxation offers a few
simpleinsights, in practice there maybe disagreements about
values and about the empirical question of what the trade-
offs are. Thus it is impossible to make recommendations

IN HEALTHCAREFINANCING
INNOVATIONS
56
about the best mechanismsfor financinghealthcare from
theory alone. It is also importantto considerthe problem
of second best-that is, the designof governmentpolicies
in situationwherethere are importantdistortionsthat can-
not be removed.
A study for Swedenconcludedthat the excessburden
is lower for a payrolltax or value added tax than for an
income tax, because an income tax is easier to evade
(Hansson 1984).Raising1 Swedishkrona (SEK)through
a payrollor valueadded tax costs SEK2.30 if the revenue
is used for transfersand SEK 1.70 if the revenueis used
for public consumption,making the excessburden SEK
1.30for transfersand SEK 0.70 for public consumption.
For an incometax the excessburden is SEK3.00 fortrans-
fersand SEK2.30 forpublicconsumption.Theseare much
higher estimatesthan earlier ones for the United States,
which suggestedthat a tax on labor income,instead of a
lump-sumtax,wouldcostonly2.5percentofrevenueraised
in deadweightloss (Harberger 1964).They are, however,
close to recent estimates on the deadweight loss of the
incometax in the United States,whichincludethe effects
of tax avoidancethroughchangesin the form of compen-
sation (suchas employer-paidhealth insuranceor housing
that is not be counted as income).Feldstein(1995) esti-
mates that a proportionalrise in all personal income tax
rates involvesa deadweightlossof $2 per incremental$1
in revenue.
Comparableestimatesof the costof raisingtax revenue
are scarceor nonexistentin developingcountries,but the
possibilities
for evadingtaxesandchangingeconomicbehav-
ior to escape taxation are at least as great as in industrial
countries.One particularlyimportantescapeis self-employ-
ment in the autonomousor informalsector,where taxes
on incomeare nearlyimpossibleto collect.The difficulties
of raisingrevenueexplainboth whygovernmentspayfor a
smallershare of health care in most developingcountries
andwhytheyrelyon indirecttaxes (salesandexcisetaxes),
import and export duties, and social security contribu-
tions.Thesedifficultiesalsoexplainwhy,whena socialinsur-
ance schemeexists,it is rare for coverageto be universal,
andwhythere is often a struggleoverusingthose resources
to subsidizehealth care for noncontributors.
Because of the problemsof financingadequate public
insurance,increasingattention has been paid to the design
GOVERNMENT FINANCING
57
of mixedsystemsin whichthe publicsectorprovidessome
type of compulsoryand universalplan but the private sec-
tor is allowedto offer complementarycoverage.Such an
arrangementmightallowpublic funds to be concentrated
on the poor without havingto raisemore resourcesgener-
ally,but there is no consensusabout the overalleffect on
efficiency.This is true evenwhen the argument over the
introduction of government catastrophic insurance in a
systemof privateinsurance(and howsuch a mixedsystem
would comparewith a purely private competitiveinsur-
ancemarket or an optimallydesignedgovernmentmonop-
olyplan) is limitedto the issueof moral hazardand excess
consumptionof health care (Besley1989; Selden 1993;
BlomqvistandJohansson 1996).
Effectsonoverallfunding.
Governmentfinancemayinflu-
ence total spendingon health care in differentways.The
traditional argument has been that the introduction of
public insurance,as with private insurance,willincrease
spendingbecauseof moralhazard-particularlywhengov-
ernmentfinancescarethroughtax subsidieson healthinsur-
ancepremiumsand stimulatesoverinsurance.The optimal
expenditure on health would include insurance,but the
costsof that insuranceshouldbe recognizedby the people
payingfor it, andtheyshouldmakerationaldecisionsabout
how much of it to buy.
Anotherargumentis that in healthcare systemsfunded
through direct taxation, there is a risk that expenditures
may be too low.This could happen becausepublic goods
andexternalitiesare undervaluedby the public,or because
people are unwillingto pay the optimal level of taxes if
theydo not think the resourcesare beingused to buy iden-
tifiableinsuranceforthemselves.Thispossibilityhasrevived
discussionof earmarked taxes, which dropped from the
mainstreamof publicfinancemanyyearsago.A number of
proposalshave recentlybeen put forward, particularlyin
the UnitedKingdom,forintroducinga specifictaxto finance
health care (Jonesand Duncan 1995).
One reasonfor these proposalsis surveysshowingthat
peoplefavorincreasesin the scaleof publicfinancing(and
provision)of healthcare.Atthe same time, somecountries
havecut publicspending.Thusthere seemsto be littlecor-
relationbetween public expendituresand the preferences
of voters.However,surveysmayask questionsin suchaway
OF HEALTH CARE
that the respondentsdo not think of a realtradeoffbetween
differenttypes of spending.If the questions were formu-
lated correctly,the discrepancywould be much smaller
(Eckerlundand others 1995).
A secondreasonearmarkedtaxes mightbe attractiveis
that peoplewould accept tax increasesfor health, but not
for other areas.Earmarkingprovidesgreater transparency
and responsivenessto the preferencesof voters and thus
would help ensure that the resourcesare used for health
and nothingelse.However,it is difficultto test the hypoth-
esisthat people are morewillingto paytaxes forhealthser-
vices if they know the proceeds are earmarked for that
purpose. Does earmarking make the government more
responsiveto public preferences,or are the benefitsreal-
ized throughbehavioralresponsesby taxpayers(whocon-
sumemore of the taxedgoods,declaretheir incomesmore
honestly,and so on)?
An earmarkedtax willnot by itselfdeterminethe opti-
mal levelof spending.In fact, it willhaveno effecton total
health care spendingif revenuesare insufficientto payfor
everything,and other tax contributionsare loweredto off-
set the earmarking.A designatedtax is usuallysufficientto
payfora healthcareprogramonlyin the caseof socialinsur-
ance.Even so, these contributionsmaynot be strictlyear-
marked,sincetheyoftenfinanceboth healthcareandtransfer
payments such as pension or unemploymentinsurance.
Usingthe same taxes to financehealth care and pensions
has causedsevereproblemsin LatinAmericabecausepen-
sionsusuallyget priority,causingseverefinancialcrisesfor
health spendingwhen revenues are inadequate for both
(McGreevey1990).Earmarkinga tax for two uses is of lit-
tle use unless the proportionsare specified.
A third reasonis that in somehealth care systemstaxes
can be described as earmarked for health, although the
correspondenceis not necessarilyone-to-one. Sweden's
countycouncilslevya regionalincometax that is usedmainly
to financehealthcare.Butthe councilsalsospendthe money
on other areas, and part of their revenuecomesas grants
from the centralgovernment.In suchcases wherethe tax
is almostcompletelyearmarked,offsettingadjustmentsin
other revenue sourcesare likelyto be small, and the link
between sources and uses is effective.Earmarked taxes
are commonin LatinAmerica,especially"sintaxes"onalco-
hol and tobacco, and part of the proceeds of lotteriesare
INNOVATIONS IN HEALTH
58
usuallyearmarkedfor health. There is little evidencethat
this approachmakes more resources availablefor health,
sincecentralbudgetscan adjustthe contributionsfromother
taxes.
From a theoretical perspectivethere are strong argu-
mentsagainstusingdifferentialcommoditytaxesfor financ-
ing.First,if thereis awell-designedincometax,differential
commoditytaxationis likelyto add little,if anything,to the
abilityto redistribute income. Of course, where income
taxes are poorlydesignedand hard to collect,there maybe
a gainfrom excisetaxes; this argumentappliesto develop-
ingcountries.Second,differentialtaxationmaybe admin-
istratively complex. Third, such taxes can be used to
discriminateagainstcertaingroups.The argumentsagainst
differentialcommoditytaxesarerelevantfordistortivetaxes
only
For correctivetaxes, suchas a tax on productswith neg-
ativeexternalities,the argumentis different.However,the
theory of correctivetaxes does not tell us anythingabout
howthe revenuefromthese taxesshouldbe used.One con-
sequenceofcorrectivetaxesisthat ifthe behavioris affected
in the desiredway,the revenue for health serviceswillbe
smaller.And the more people smoke and drink, the more
moneywillbe availablefor health care.Thus governments
maybe ambivalentabout reducingthe behaviorthat brings
in revenue,especiallyifanygainsfromreducedhealthexpen-
diture willonly materializein the future. And healthgains
do not necessarilymean reduced lifetimeexpenditureon
healthcare.
Effectsonthe allocationof resourceswithin thehealthcare
sector.It is often advocated,or hoped, that a changein how
health care is financedwill increaseefficiencyin the allo-
cationof health care resources.Some of the proposalsfor
earmarkedtaxes can be interpretedin that way.But in the-
orythere isno specificlinkbetweenthe waymoneyis raised
and the wayit is spent. It is possibleto combinedifferent
mechanismsfor allocatingresourcesin the healthcare sys-
tem and different ways of providinggovernment funds.
Moreover,the level of spendingis not in any significant
waydeterminedbythe levelor compositionof publicfinance
(OECD 1987,annexA). Thisobservationdoesnot ruleout
that changesin health care financingmechanismscan be
an importantelementof health care reformin a particular
CARE FINANCING
country at a particular point in time. However, general
conclusionsabouttherelationbetweenefficiencyandfinanc-
ing mechanismscannot be supportedby theoryor empiri-
cal evidence.
An exceptionto this generalizationoccurswhen spe-
cific sources or amounts of revenue are allocatednot to
overallhealthcarespendingbut to specificprogramsor cat-
egoriesof spending, and the earmarkingis not offset by
changesin other sourcesof funds.This approachaffectsthe
allocationof resources between individual medicalcare
(whichis highlyvisibleand thereforepoliticallyattractive)
and publicgoods whosebenefitsmay not be perceivedor
appreciatedby voters;it is an extensionof the argument
that directfinancingmaybe suboptimalfor such goods.It
is alsoan exampleof the issuesraisedbypublic choicethe-
ory (Buchanan1963),whichtakesself-interestinto account.
It does not followthat intrasectoral earmarkingwill be
optimal,but it mayimprove on the politicalallocationof
resourcesthat wouldotherwiseresult(particularlyinadecen-
tralizedsystem,whereone politicallevelwouldchoosedif-
ferentlythan another). Colombia'scurrent health system
reformis an exampleof tryingto use earmarkingto improve
the overallallocationof resourceswhile leavingdecisions
about individualcare to providersand insurers.
Conclusionsand Recommendations
Direct out-of-pocketpaymentscan solvepart of the financ-
ing problem in health care,and are the natural wayto pay
for inexpensivegoodsand services.But health care can be
socostlyas to makedirectpaymentinfeasible,makingrisk
sharingnecessary.Private,competitive,voluntaryinsurance
existsfor thisreasonand can financea largeshareof health
care. Yetsuch insuranceis unaffordableby the poor, dis-
criminatesagainstthose most in need, and can be expen-
siveto administer.For allthese reasons,thereremainsarole
for governmentin health care financing.
The issuesrelatingto how much and what form of gov-
emnmentfinanceare best affectboth equity and efficiency
and include poverty,high-riskgroups,and the difficulties
of predictingfuture needs and costs.We end with a brief
summaryof howhealthcareis financedin the worldtoday;
considerwhat conclusionscan be drawn about the proper
roleofgovernment,the appropriatemodelforpublicfinance,
GOVERNMENT FINANCING OF HEALTH CARE
59
and the sourcesof revenue;and examinesomeof the prob-
lemsaffectingdevelopingcountries.
The currentsituotion
Public finance is the main source of revenue for health
care systemsin most parts of the world. In high-income
countriespublicfinance accountsfor about 75 percent of
the total. The main exceptionis the United States,where
public financecoversonly about 50 percent of healthcare
financing.Takinginto accountthe high health care expen-
dituresin the United States,however,the share of public
financefor healthcare relativeto GDP is similarto that in
other industrialcountries.In developingcountriespublic
financeis less important,sometimesas little as 20 percent,
and out-of-pocketpaymentsare a larger shareof the total.
The compositionof public financevariesconsiderably,
particularly in high-incomecountries.France,Germany,and
the Netherlands rely mainly on social insurance, while
Canada, the Nordic countries,and the United Kingdom
relyon generaltaxation.(TheUnitedStatesis againunusual,
in that it operatesboth kindsof publiclyfinancedsystems.)
Socialinsuranceis rare in developingcountries,and it is
commonto find 100percentdirectfinancinginmanyAfrican
and Asiancountries.At intermediateincomes-as in most
LatinAmericanand Caribbeanaswellas someAsiancoun-
tries-the shares are variable,with some countries using
both modelsandsomerelyingprimarilyon socialinsurance.
Except for a few countries (Switzerland,the United
States),privatehealthinsuranceis aminorsourceof finance.
In developingcountriesthis is becausemost of the popu-
lation cannot affordprivate insurance;in industrialcoun-
triesit isbecausethe statehas assumedmostofthe insurance
function. Private insurance usuallycomplementspublic
insurance,and a countrymayfinancehealth care in differ-
ent ways accordingto the income,employment,age, and
locationof the insured and accordingto the setvicescov-
ered and their cost.
Ro/eof government
financing
Governmenthealthcare financingservesseveralpurposes.
One is to ensurethe provisionof publicgoods.Most pub-
lic financing,however,is for private goods in the form of
individualhealthcare,particularlywhentheyare costly.The
onlyreasonforgovernmentsto financeinexpensivehealth
careis to makeavailableto the poor whatthe nonpoor can
buy out of pocket.
Moregenerally, the mainargumentforgovernmentinter-
ventionis the desire to achieveuniversalaccessto health
care.Thisargumentis basedboth on efficiency(arisingfrom
market failuresin the provisionof private insurance)and
equityconsiderations(the financingof health care should
be accordingto abilityto pay, and the distributionof ser-
vicesaccordingto need).The need to providepublicgoods
and to alleviatepovertyare importantreasonsfor govern-
ments to pay for health care in developingcountries;it is
public insurancefor high-costcare that accounts for the
higher share of public spending in industrial countries.
Variationsin the generalpattern of financingreflectdiffer-
encesin the resourcesavailableto governments,the func-
tioningof markets,people'svalues,and the responsiveness
of the politicalsystemto publicpreferences.There is little
consensusor empiricalevidenceas to the "right"level of
financingfor health care, in total or by the state.
Governmentfinancingcan be combinedwith manydif-
ferent modelsfor deliveringhealth care,by private as well
as publicproviders.In particular,public financedoes not
implypublic provision.As with private insurance,exten-
sivegovernmentfinancecan create moralhazardand inef-
ficiencyin the allocationof resourcesto healthcare as well
as withinthe health care sector.However,there is no sim-
ple relationbetweenthe sourceof financingand efficiency
in healthcare provisionand delivery.
Modelsforgovemment
finonce
The factthat somehigh-incomecountriesfollowthe directly
or tax-financedmodel while others rely on social insur-
ance, and that in many countriesboth systemsoperate at
once, suggeststhat neither model is better than the other.
There is no empiricalevidencethat the theoreticalvirtues
of eithermodel (universalityand a less distortingtax base
in tax-based systems, transparencyand ease of revenue
collectionin socialinsurance)lead to systematically better
health, lowercosts, or any other sign of superiority.The
factthat low-andmiddle-incomecountriestend to use only
direct finance and to have little or no socialinsurance-
INNOVATIONS IN HEALTII CARE FINANCING
60
just as they havelittle or no privateinsurance-is not evi-
dence to the contrary.It simplyreflects the difficultyof
financinginsurancefor a mostlypoorpopulation,and espe-
ciallythe problemof collectingthe taxes on laborthat usu-
ally finance socialinsurance. Explicitsocial insurance is
harder to organizeand depends on more favorableeco-
nomicconditions,but that doesnot makeit better or worse.
For manycountriesthe questionis not whichmodel to
follow so much as how they should interact when both
alreadyexist-social insuranceforurban,formalwagelabor
and civilservants,and directfinanceforthe poor andinfor-
mal workers.Shouldcontributorsto socialinsurancehelp
subsidizethe tax-financedbeneficiaries,who usuallyhave
lowerincomes?Should different taxes be pooled so that
socialinsurancecan be more extensiveandnot dependonly
on labor taxes?And howshouldeither modelrelate to pri-
vate insurance?Severalcountrieshaveexperimentedwith
hybridmodels that distinguishcontributionsfrom subsi-
dies,in an attemptto meet the two governmentrolesof pro-
tectingthe poor andguaranteeingandregulatinginsurance
forthosewhocan affordit. Conclusionsabout howwellany
systemworks cannot be derived from theory or general-
izationsbut dependon the specificarrangementsfor financ-
ing, coverage,and benefits.
Sourcesof revenue
Accordingto the benefit principle,assumingthat the pur-
poseof governmentinterventionis to achieveuniversalcov-
erage,the tax base shouldbe asbroad as possible.Specific
taxes (such as exciseduties) that only part of the popula-
tion willpay should not be used. Such taxes maybe justi-
fiedto curb the use of productswith negativeexternalities,
but there is no reasonthe revenuesshouldbe devoted to
health care.
One wayto achievea broad tax base is to create a fund
that receivesrevenuesfrom severaltaxes.Thisis what gen-
eralrevenuefinancingdoes,but the fundcouldcombinetaxes
in differentproportionsspecifically to increasethe stability
of revenuesandprotect againstcyclicalvariation.Theseout-
comeswould be usefulbecausethe need for health care,
particularlyforpublic subsidies,is likelyto be countercycli-
cal, increasingunder adverseeconomicconditions.It is also
importantthat the revenuebase growat least as fast as the
generaleconomy,giventhepressureto increasehealthexpen-
dituresasincomegrows.Income,valueadded,andsalestaxes
are goodcandidates.Indirecttaxes,employercontributions
to socialinsurance,corporationtaxes,and othersourcesmake
it less clearto the publicthat there is a link betweenwhat
they pay and the servicesthey receive,whichis what the
benefitprincipleis about.Particularlywithemployercontri-
butions there is a risk for misunderstandingabout the inci-
dence of taxes,and consequentlya tendencyto overinsure.
Toachieveequity,governmentfinancingshouldbe based
on abilityto pay, whichmakes a case for proportionalor
progressiveincome taxes.Taxeson labor incomeare sec-
ond-bestin this respect, since capitalis not taxed and the
contributionis usuallyproportionalto earningsonlyup to
someceiling.An incometax alsohas the advantageof being
transparent.The empiricalevidenceis that directtaxes are
generallymore progressivethan socialinsurancecontribu-
tions,whichare moreprogressivethan indirecttaxes.In this
respect,the same taxes that can be recommendedfor cre-
atinga broad base for fundingare alsothose most likelyto
collect from people in proportion to their abilityto pay.
Excisetaxes on luxuryconsumptionare an exception:they
are progressivebut narrowlybased.
Governmentfinanceby anycombinationof taxes and
contributionsis almostcertainto be more progressivethan
out-of-pocketpayments.This findingimpliesthat govern-
ments shouldfinanceas muchhealth carespendingas effi-
ciencyconsiderationspermitin developingcountries,where
a large share of finance is direct paymentsby consumers.
Publicfinance maybe less progressivethan privateinsur-
ance financing,but that is not an argumentin favorof pri-
vate insuranceif the progressivityis simplydue to the fact
that onlythe rich buyprivatecoverage.Publicsubsidiesfor
insurancepurchasebythe poor (forexample,throughvouch-
ers) can in principlerespectabilityto paywhilegivingpeo-
ple a choiceof insurers. However,subsidiesthrough the
tax system-deducting the price of insurancefrom income
for tax purposes-are regressiveand should be avoided.
Theyare alsoinefficient,stimulatingthe purchaseof insur-
ance withoutcontrol over costs,coverage,or benefits.
From an efficiencyperspective,taxes shouldminimize
distortionsto economicactivityand encouragean optimal
overalllevelof spendingand an efficientprovisionof health
care.Theoretically,lump-sumtaxes are the idealsourceof
GOVERNMENT FINANCING
61
publicfinance,sincetheydo not createanydistortions.Such
taxesare difficultin practice,however,andchoosingamong
differenttax alternativesin terms of excessburden is diffi-
cult. The fact that the burden associated with any tax
increaseswith the marginaltax ratereinforcesthe argument
for a broad base of revenuesthat mayinclude many dif-
ferent taxes at lowerrates rather than a few taxes at higher
rates. Unfortunately,there are few estimatesof marginalor
total deadweightlossesdue to particulartaxes in develop-
ing countries,and the ranking of differenttaxes in indus-
trialcountriesmaynot applyto conditionsof much greater
poverty,informalemployment,and easiertax evasion.
The deadweightlosses associatedwith taxes limit the
amountthat can be raised for health care spendingwith-
out makingan economyworse off. If people are willingto
pay more for health care but not for public spendingin
general,then earmarkedtaxes are an attractivealternative
for achievinga higher and more optimal allocation of
resources to and between different health care services.
Earmarkedtaxes are often used to paypart of government
healthcarecosts,but the argumentsfor suchtaxesare weak,
and there is no evidencethat they lead to more or better
expenditurebecausethe amountsraised are usuallysmali
and allocationsfrom other taxes can be reduced to offset
them. Socialinsurancecontributionsare usuallythe only
earmarkedtaxes capableof payingfor a health system.
Of course,anotherbroad-basedtax couldbe earmarked
for healthcare,and if the revenueswere largeenoughand
the tax were used to buy insurancewith subsidiesto the
poor,it mightbe possibleboth to guaranteeadequatefinanc-
ing and to fulfillequityand efficiencycriteria.For thispur-
pose, as wellas for a generalrankingof taxes with respect
to their distortingeffects,a value added tax is more effi-
cient than a payrolltax (sinceit does not tax labor while
ignoringcapital),and relativeto an incometax encourages
savingand discouragesconsumption.A valueadded tax is
alsoharder to evade and, when taken in conjunctionwith
the healthbenefit it provides,is clearlyprogressive(Fuchs
1996).
Structureofinsurance
In additionto decidinghowmuch insuranceto payfor and
forwhom,andwithwhichtaxes,governmentscanalsodeter-
OF HEALTH CARE
minehow to share costswith patients and what economic
incentivesto build into insurance.The same principlesof
optimalinsuranceapplyto publicfinancingandprivatecov-
erage.Thus,forexample,deductiblesandcoinsurance should
be introduced so that the insuranceprotects againstthe
highestfinancialrisks rather than leavingpatients unpro-
tected after some limit. Public insurance generallydoes
not respectthese principles,and user chargesare far from
optimal.Especiallyin developingcountries,this situation
reflectsthe difficultyof discriminatingaccordingto ability
to payas wellas the informationrequirementsfor efficient
protection.
Another conclusionis that socialinsurancethat is lirn-
ited to specialgroups(suchas the elderlyor the poor), dis-
eases,or treatmentsleavespart of the populationwithout
coverageandincreasesthe riskfor suboptimization.Basing
eligibilityon income or wealth also creates incentivesfor
inefficientbehaviorin the intertemporalallocationof funds.
It is probablybetter to establisha singleinsurancecover-
age and then reduce or waivepaymentsby the poor.
of developing
Problems countries
Althoughpublic and private levels of spendingon health
carediffersubstantially,
the problemsrelatedto publicspend-
ing are similarin economiesat differentincomelevels.In
most countriesgeneralbudget constraintsmakeit difficult
to increasepublicspendingon healthcare,whetherfinanced
by debt, taxes, or reallocationfrom other sectors. High-
incomecountrieshavenearlyallslowedthe rate of increase
in health spendingin recent years,sometimesstabilizing
spendingas a share of GNP In verylow-incomecountries,
however,expendituresare stillsolowthat eveniffundswere
spent as cost-effectivelyas possible,theywould meet only
the most criticalhealth needs. And many middle-income
countriesfacesimultaneously risingexpendituresandunmet
health care demands (WorldBank 1993).
Whatcanbe done?Moreattentionhasbeenpaidto what
developingcountries should buy with their public health
careresourcesthanto howthoseresourcesshouldbe raised.
These countriesshould concentratespendingon services
that benefit societyas a whole, particularlycost-effective
publicgoodssuch as immunization,sanitation,healthedu-
cation, and control of vector-borne disease. Remaining
INNOVATIONS IN HEALTH CARE FINANCING
62
resources should be spent on private curative care, with
somemixof considerationforcosts(to providecatastrophic
protection),effectiveness(to ensurereal healthgains),and
responseto needsperceivedbythe public(WorldBank1987
and 1993;Musgrove1996). For very poor countriesit is
possibleto design a basic packagefor whichgovernment
financewouldbe justified;forlesspoor countriesthe choice
of what servicesto financebecomesmore complex.
What implicationsdoes an appropriate strategyhave
forpublicfinancein developingcountries?Thegeneralprin-
ciplesdiscussedaboveallapply,but threequestionsacquire
particularurgencywhengovernments'capacitiesto finance
a reasonablelevelof universalcare are limited:
* How caninsurancecoveragebe increasedwithoutincur-
ring perversesubsidiesor overinsurance?
* How shouldusers be chargedfor servicesthat are pub-
licly subsidized(see Gertler and Hammer in this vol-
ume)?
* How can governmentsensure that benefitsare concen-
trated on the poor?
If there is little alternativeto chargingusers of govern-
mentfacilitiesand privateprovidersthat publicfinancesub-
sidizes,then differentialfeesto protect the poor are crucial.
In countrieswhere high-incomeconsumerscarry private
insurancebut use public facilitiesfor free, collectingthe
cost of servicesis urgent for both equity and efficiency.
And the taxes used to financewhat is not chargedto users
need to be as progressiveand free of distortions as eco-
nomic conditionsallow.Tax expendituresin the form of
tax exemptionsare evenlessadvisablethan in richer coun-
tries.
Perhaps the most difficult question is how to extend
catastrophicinsurancecoverageto peoplewhocannotafford
adequate,unsubsidizedprivateinsurance.Thegovernment
can encouragethe developmentof an efficientinsurance
marketbyimposingappropriateregulationon privateinsur-
ance, which often operates with little control. To avoid
subsidizingindividualhealthcarefor the better-off,the gov-
ernment shouldavoiddirect and indirectsubsidiesfor pri-
vate health insurance. And to avoid cost escalation,
compulsoryinsuranceplansshouldincludeadeductibleand
coinsuranceup to a ceiling.When insuranceis subsidized
for the poor,both the deductibleand the coinsurancepay-
ments must be lower, or the insurance cannot be used
when it is most needed. The problem is the same as for Feldstein,Martin.1973. "TheWelfareLossof ExcessHealth
user fees generally. Insurance." Journal of Political
Economy 81:251-80.
Concentratingspendingon certainservices,as suggested . 1995."TaxAvoidance andthe Deadweight Lossof the
Cbonentratinghift
spmendingurces
ono IncomeTax."NBERWorkingPaper5055. NationalBureau
above,willshiftsomeresourcesfrom the rich to the poor, of Economic Research,Cambridge, Mass.
becausethe poor suffersomewhatmore from diseaseswith Fuchs,VictorR. 1996."Economics,Values,and HealthCare
cost-effectivepreventiveor treatment measures.But the Reform." American Economic Review86:1-24.
best wayto ensureequityin governmentfinanceis through Gertler,Paul,andJacquesvanderGaag.1990.TheWi'llingness to
the financingmechanismsthemselves-the combinationof PayforMedicalCare: Evidence fromTwoDeveloping Countries.
taxes, contributions,and feesthat payfor servicesthat the Baltimore, Md.:JohnsHopkinsUniversity PressfortheWorld
taxles
, Bank.
wholepopulationmayneed. Gottschalk,P.,B. Wolfe,and R. Haveman.1989."HealthCare
Improvingpublicfinancingof health care willnot solve Financing in theU.S.,U.K.andtheNetherlands: Distributional
allthe problemsof the healh sector.Politicaldecisionsare Consequences." InA.ChiaconeandK Messere,eds.,Changes
neededto ensurethat fundsare not onlyraisedas efficiently in RevenueStructures.Detroit: WayneState University
and equitablyas possible,but also are used for the poor Press/International Instituteof PublicFinance.
and forserviceswith publicbenefits.Nor are the suggested Grosh,Margaret.1992.Administering Targeted
SocialProgramsin
LatinAmerica:FromPlatitudes to Practice.A WorldBank
reformssimpleto ilmntnRegionalandSectoralStudy.Washington, D.C.:WorldBank.
to require much more informationabout insurancebene- HanssonI. 1984."Marginal Costof PublicFundsforDifferent
ficiaries,patients,providers,and servicesthan governments TaxInstruments andGovernment Expenditures."Scandinavian
are accustomedto dealingwith.Current reformsin several Journal of Economics 86: 115-30.
mniddle-income countriesshowjust howimportantand dif- Harberger, ArnoldC. 1964."Taxation, ResourceAllocation,
and
ficultthis processis. Still,there is scopefor improvinghow Welfare." In J. Due,ed.,TheRoleof DirectandIndirectTaxes
publicresourcesforhealthcareareobtainedandin
theFederal System.Princeton,NJ.: PrincetonUniversity
public resources for health care are obtained and used, press
and the potential benefitsare verylarge. Jones,Andrew,andAlanDuncan.1995. 'Hypothecated Health
Taxes-AnEvaluation ofRecentProposals."
OfficeofHealth
References Economics,
UnitedKingdom.
J6nsson, Bengt. 1989. "What Can AmericansLearn from
Arrow,
KennethJ.1963."Uncertainty
andtheWelfareEconomics Europeans?"HealthCareFinancingReview(annualsupple-
of MedicalCare."AmericanEconomicReview53(5): 537-38. ment):79-93.
Berk,M.L.,AC.Monheit,andM.M.Hagan.1988."HowtheU.S. -. 1996."Comprendela r6formedes systemesde sant6
SpentItsHealthCareDollar:192940."HealthAffairs
7(4):46-60. (Makingsenseofhealthcarereform)."In OECD,LaRe'forme
Besley,
Timothy.
1989."Publicly ProvidedDisasterInsurancefor desSystemesde Sante'-LaVolonte'
de ChangementEtudesde
Healthandthe Controlof MoralHazard."Journalof Public Politique
deSante(OECDHealthPolicyStudies)8. Paris.
Economics39: 141-56. J6nsson, Bengt, and U. Gerdtham. 1995. "Cost Sharing for
Blomqvist,A. Forthcoming."Optimal Non-linear Health Pharmaceuticals:
TheSwedishReimbursement System."
InN.
Insurance."
Journalof HealthEconomics. Mattison,ed.,Sharingthe Costsof Health:A Multi-country
A.,andP-OJohansson.1996."EconomicEfficiency
Blomqvist, Perspective.Basel: PharmaceuticalPartners for Better
and MixedPublic/PrivateInsurance."Workingpaper 10. Healthcare.
Stockholm Schoolof Economics,Sweden. Kakwani,N.C. 1977. "Measurement of TaxProgressivity:
An
Buchanan,J.M. 1963."TheEconomicsof EarmarkedTaxes." International
Comparison."Economic Journal87:71-80.
Journalof Political
Economy 71: 457-69. Lambert,L. 1989."HorizontalEquity:Measuresin Searchof a
Eckerlund,
I., M.Johannesson,andP-OJohansson,M.Tambour, Principle."
NationalTaxJournal
42: 139-54.
and N. Zethraeus.1995."ValueforMoney?A Contingent Lewis,Maureen.1994.TheOrganization,DeliveryandFinancing
ValuationStudyof the OptimalSizeof the SwedishHealth of HealthCarein Brazil:Agendaforthe 90s."WorldBank,
CareBudget."HealthPolicy34:135-43. LatinAmericaand the CaribbeanDepartment1, Human
Feldman,R.,andB.Dowd.1991."ANewEstimateoftheWelfare Resources Division,
Washington, D.C.
Lossof ExcessInsurance."AmericanEconomic Review81: Manning,WG.,JosephNewhouse,N. Duan,andothers.1987.
297-301. "HealthInsuranceandtheDemandforMedicalCare:Evidence

GOVERNMENT FINANCING oFHEALTH CARE

63
 from a Randomized Experiment."American EconomicReview
77: 251-73.
McGreevey, William.1990. Social Securityin Latin America:Issues
and Optionsfor the Bank. World Bank Discussion Paper 110.
Washington, D.C.
Mesa-Lago,Carmnelo.1992.HealthCareforthePoorin LatinAmerica
and the Caribbean.Washington, D.C.: Pan-American Health
Organization and Inter-American Foundation.
Mirrlees, James A. 1971. "An Exploration in the Theory of
Optimum Income Taxation." Review of Economic Studies 38:
175-208.
Musgrave, Richard A. 1959. The Theory of Public Finance.New
York: McGraw-Hill.
Musgrove, Philip. 1996. Public and Private Roles in Ilealtb Care.
World Bank Discussion Paper 339. Washington, D.C.
Newhouse, J. P 1994. "Patients at Risk: Health Reform and Risk
Adjustment." Health Affairs 13(1): 132-46.
O'Donnell, Propper, and Upward. 1993. "United Kingdom." In
Eddy van Doorslaer, Adam Wagstaff, and Frans Rutten, eds.,
Equity and FinanceintheDeliveryofHealth Care:An International
Perspective.Oxford: Oxford University Press.
OECD (Organization for Economic Cooperation and
Development). 1987. Financingand DeliveringHealth Care:A
ComparativeAnalysis of the OECD Countries.Paris: OECD.
- . 1995. "New Directions in Health Policy."OECD Policy
Studies 7. Paris.
Various years. Health Database. Paris.
V Weisbrod, Burton A. 1991. 'The Health Care Quadrilemma: An
Pauly,Mark V 1968. "The Economics of Moral Hazard."American
Economic Review 58(3): 531-37.
1974. "Overinsurance and Public Provision of Insurance:
The Roles of Moral Hazard and Adverse Selection." Quarterly
Journalof Economics 88: 44-62.
Phelps, C. E. 1986. "Large-Scale Tax Reform: The Example of
Employer-Paid Health Insurance Premiums." University of
Rochester Working Paper 35. Rochester, N.Y
- 1992. Health Economics.New York: Harper Collins.
Reinhardt, Uwe E. 1992. "Organizing the Health-Care Sector:
Regulatoryversus Market Approaches. " Princeton University,
Woodrow Wilson School, Princeton, NJ.
Selden,T M. 1993. "Should the Government Provide Catastrophic
Insurance?" Journal of Public Economics 51: 241-47.
INNOVATIONS IN HEALTH CARE FINANCING
64
Spence, M., and Richard Zeckhauser. 1971. "Insurance,
Information and IndividualAction."AmericanEconomicReview
Papersand Proceedings61: 380-87.
van de Ven, W.P, FT. Schut, and Frans Rutten. 1994. "Forming
and Reformiingthe Market for Third Party Purchasing of Health
Care." Social Science and Medicine 39: 1405-12.
van Doorslaer, Eddy, and Adam Wagstaff. 1993. "Equity in the
Finance of Health Care." In Eddy van Doorslaer, Adam
Wagstaff, and Frans Rutten, eds., Equity in the Finance and
Delivery of Health Care:An International Perspective.Oxford:
Oxford University Press.
van Doorslaer, Eddy, Adam Wagstaff, and Frans Rutten. 1993.
Equity in theFinanceand DeliveryofHealth Care:An Intemational
Perspective.Oxford: Oxford University Press.
VanVliet, R. C., and W.P van de Ven. 1992. "Toward a Capitation
Formulafor CompetingHealth Insurers: An EmpiricalAnalysis."
Social Scienceand Medicine 34(9): 1035-48.
Wagstaff, Adam, and Eddy van Doorslaer. 1993. "Equity in the
Delivery of Health Care: Methods and Findings." In Eddy
van Doorslaer, Adam Wagstaff, and Frans Rutten, eds., Equity
in the Finance and Delivery of Health Care:An International
Perspective.Oxford: Oxford University Press.
Wagstaff, Adam, Eddy van Doorslaer, and others. 1996.
"Preliminary Results on the Redistributive Effect of Health
Care Financing in TwelveCountries." Paper presented at the
Equity Project workshop, April, Berne.
Essayon TechnologicalChange,Insurance, Quality of Care, and
Cost Containment."JournalofEconomicLiterature 29(2):523-52.
WorldBank. 1987.FinancingHealth SetvicesinDevelopingCountries:
AnAgendaforReform. AWorld Bank Policy Study.Washington,
D.C.
. 1993. World DevelopmentReport 1993:Investingin Health.
New York:Oxford University Press.
. 1995. "Brazil: A PovertyAssessment." Report 14323-BR.
Washington, D.C.
Zeckhauser, Richard. 1970. "Medical Insurance: A Case Study of
the Tradeoff between Risk Spreading and Appropriate
Incentives." Journalof Economic Theory 2: 10-26.
Zweifel,P 1992.BonusOptionsin Health Insurance.Norwell, Mass.:
Iluwer Academic Publishers.
From Beveridge to Bismarck: Health
Finance in the Russian Federation
IgorSheiman
F or decadesthe Soviethealthsectordevelopedusing
the Beveridgemodel-that is, a tax-financedand
highlyorganizedsystemwith an emphasison uni-
versal accessto comprehensivecare. In the early 1990sa
countrywidereform of the health care systembegan, the
coreofwhichwasa transitionto a payrolltax-basedmanda-
tory health insurance approach known as the Bismarck
model. This shift is not limited to the method of raising
health revenue, and entailsprofound changesin the way
the systemis managedand financed.This new approachis
a reactionto the negativefeaturesof the formercommand
andcontrolsystem,the mostimportantofwhichwerechronic
underfunding,governmentdominanceof healthcarefinance
and provision,top-downnoncontractualresourcealloca-
tion,no consumerchoice,and input-basedfundingof health
care providers.These characteristicsled to considerable
inefficienciesand irrationalstructuresin healthcare provi-
sion.Moreover,comprehensiveanduniversalcoveragewere
underminedby inadequatefundingand inefficientutiliza-
tion of healthresources.
The maingoalsof reformare to:
* Raise additionalfunds
* Increase the intemal and allocativeefficiencyof health
careprovisionbymovingtowarda systembasedon con-
tractualrelationshipsbetweenhealthcareconsumersand
providersusingnew paymentincentives
* Enhancethe qualityof care and ensure consumerpro-
tectionthrough third-partypayers
* Maintainsocial solidarityand equity while increasing
consumerchoice.
IgorSheimanisan assistantprofessor
at theMoscow
MedicalAcademy
andseniorhealtheconomist
in Moscow.
65
The conceptualbasisof the reformis regulatedcompe-
tition,whichcombinesmarket incentivesand regulationin
the purchase and provision of health care. The Health
InsuranceAct, passedin June 1991and amendedin early
1993,is the legalbasisforhealthsectorreform.TheActman-
dates universalcoveragefor all citizens,includinga com-
prehensivepackageofmedicalbenefitsdefinedbythe basic
(national)and territorialprogramsof mandatoryhealthinsur-
ance.The systemis financedby an earmarkedpayrolltax
andgeneralbudgetrevenue.Employersmakeincome-based
contributionsto the newlycreated territorial mandatory
healthinsurancefunds (3.2percent of payroll)and the fed-
eralmandatoryhealthinsurancefund (0.4percent)to cover
their employees.Localgovernmentsmakecontributionsto
the territorial funds for the nonworkingpopulation and
directlyfinancea numberof healthprogramsandproviders.
The system is highly decentralized. Each of the
Federation'seighty-eightregions(oblasts),with populations
rangingfrom500,000to severalmillionresidents,is respon-
sible for its mandatory health insurance system. This
approachreflectsthe country'sgeneralmovetoward polit-
ical and economicdecentralization.
Each territorialfund poolsthe premiumsand allocates
themto insurersbasedon aweightedcapitationformula.Some
oblasts have developeda pluralisticsystemof health care
purchasingwith a few competinginsurers;othersuse a one-
purchasermodelactingthroughthe localbranchesofthe ter-
ritorialfund.Sincecommunityratingrestrictsrisk selection,
competitionamonginsurersis focusedon increasingmarket
share.Insurers,actingasthird-partypurchasersofhealthcare,
at KaiserPermanente
International
contracthealthprovidersandpaythemaccordingtothemethod (1996,p.20) use 1990asa benchmarkforRussiaandCentral
determinedby eachoblast'smandatoryhealthinsurancereg- Asia.Here 1992is used-the last yearbefore reform and
ulation.Ratesare setby multilateraltariffagreementwiththe a startingpoint for big structural changesin the Russian
involvement ofthe oblasthealthcommnittees, mandatoryhealth economy.
insurancefund, and medicalassociations(for detailson the Considerablechangesin healthspendingoccurreddur-
designsee Sheiman1994and Klugmanand Schieber1996). ing 1992-95, with two opposingtrends (table 1). During
Three and a half years of reform implementationpro- 1992-94 the share of public health spendingin the econ-
vide some indicativeoutcomesthat are the subjectof hot omyincreasedsubstantially, from2.6to 4.1percentof GDP
debatein Russia.Evaluationsofthe reformrangefrom deep In 1994real public health expenditureswere 24 percent
frustrationto high enthusiasmand often depend on the higherthan in 1992.This increaseis overstated,however,
vestedinterests of the evaluators. becausethe GDP deflator seemsto understate growthin
The internationalcommunityis increasinglyinterested the pricesof health goods and services,particularlyphar-
in the processand outcomesof Russia'stransitionto manda- maceuticals.Still,evenwhena morereasonablepricedefla-
tory health insurance.Although recent reports from the tor for medical goods and services is used, real health
WorldBank,the U.S.AgencyforInternationalDevelopment spendingin 1994was 20 percent higherthan in 1992.
(USAID),and other donors evaluate health reforms in a Suchgrowthwasa rarepositivetrend in adecliningecon-
number of countries in Central and Eastern Europe and omy,however.Mandatoryhealh insurancestarted at a time
Central Asia (Ensor 1993; Klugmanand Schieber 1996; of relativelygoodfundingand high hopes.But by 1995the
Langenbrunner and others 1996; Goldstein and others oppositetrend had started, and is still under way.Health
1996),their multicountryapproachmakes generalizations spendinghas fallenin both absoluteand relativeterms.The
difficultand sometimesmissesessentialcharacteristicsof share of health spendingin GDP dropped to 3.3 percent
healthreformin individualcountries. in 1995 and, according to preliminaryestimates, to 3.1
This paper reviewsrecent developmentsin the Russian percent in the firstnine monthsof 1996.Realhealthexpen-
health sector and appraisestheir impactusing traditional ditures fell to pre-reform levels.Thus the drop in health
criteria. Thus the emphasisis on the reform's successes spendingwas more substantialthan was the declinein the
andfailuresin securingadditionalfunding,achievinghigher country'soveralleconomy.As a resultthe overalltrend for
equity,providingefficient and qualitycare, and overcom- the entire period is positivein terms of the relativeweight
ingstructuraldistortions.The mainproblemsof the reform of public health expendituresbut negativein terms of the
and potential waysof resolvingthem are then discussed. absolute volume of real health expenditures (with a 10
The paper concludeswith health policyimplicationsfor percent drop).
transitioneconomies. Expectationsformandatoryhealthinsurancewerebased
on twoassumptions.First,that localgovernments, afterbeing
The Reform's Impact on Health Finance liberatedfrom centralizedresourceallocation,wouldreori-
ent their budgets toward health and other socialservices.
Currentdebatesonhealthcarereformfocuson howto raise Second,that economicdeclinewould not last long. Both
additionalrevenuesto strengthenthe health sector.This assumptionswerewrong.Regionaland localgovernments
section assessesthe new revenue-raisingmodel's effects
on health revenuesand identifiesthe relativecontribution TABLEI
of differentsourcesof finance. Publichealthspending inthe Russian Federation,
1992-95
Publichealthexpendituretrends 1992 1993 1994 1995
Publicspending/GDP 2.60 3.77 4.08 3.29
It is important to choosethe rightbenchmarkwhenesti- Real
healthexpenditure
index 100 135 124 90
matingtrendsin healthexpenditures.KlugmanandSchieber Source:
Estimate
basedonKorchagin
1996and Shishkin
1996.

INNOVATIONS IN HEALTH CARE FINANCING

66
faceincreasingfiscaldifficultiesandoften cannotaffordtheir Netherlands,and78 percentin Germany.Evenin Belgium,
newhealth care responsibilities.They alsotend to support wherestate subsidiesare more important,the payrolltax is
agriculture,new road construction,and other more tangi- stillthe main sourceof revenue-58 percent.
ble andvisibleservicesand projects.The economyhas been The reason the payrolltax accounts for such a small
decliningforsixyears,and onlyin recentmonthshavethere percentageof publichealthrevenuein Russiais that, accord-
been signsof improvement.As a result much less money ingto the legislation,
workersare not supposedto contribute
has beeninfusedinto the healthsectorthanwasanticipated. to mandatoryhealth insurance.Thus,contraryto declara-
Still,the declinein realpublichealthexpendituresshould tions about the transitionto a mandatoryhealth insurance
be seen in the context of changes in other sectors. A 10 model,the health systemis still financedmainlyby taxes.
percent drop in real spendingis somewhattolerablegiven Moreover, local governments are reluctant to allocate
that educationhas lost 27 percent of its funding and cul- resourcesto mandatoryhealthinsurancefunds.Their con-
ture has lost 31 percent.In relativeterms the health sector tributionscoverjust 24 percentof mandatoryhealthinsur-
is doingbetter than other socialservices,allof whichhave ance enrollees,although they are supposedto pay for 56
been fuhded for decadesusingthe residualmethod-that percentof the population.In twenty-threeregionslocalgov-
is,afterpriorityallocationsto defense,agriculture,andother emmentsmake no contributionsfor nonworkingpopula-
major sectors havebeen made. The new payrolltax pro- tions (FederalFund of MandatoryHealth Insurance1995;
tected the health sector from a more dramaticfall. To be Kravchenko1996)-a clearviolationof the legislation,which
moreprecise,substantiallydecreasingpublicresourceshave states that such contributionsare mandatoryand muchbe
been redistributedtoward the health sectorwithvery little equalto per capitacontributionsfor employees.The tran-
value added for healthcare providersand their patients. sition to mandatoryhealth insurancehas givenlocal gov-
Comparisonswith other transition economiessupport ernmentsan excuseto underfund the health system,with
the viewthat the new payrolltax has helped stabilizethe the shareof healthexpenditurecomingfrom localbudgets
healthcare system.For example,during1990-94allCentral droppingfrom 18 percent in 1993to 12percent in 1996.
Asiancountriesexperienceda much more substantialdrop
in healthcare funding,in both absoluteand relativeterms, Total healthexpenditurestructure
than did Russia(Klugmanand Schieber1996,pp. 17,20).
Similarly, the shift to nationalhealthinsurancesystemshas One of the goalsof healthreform in Russiawas to enable
had a positiveimpacton public healthfundingin Croatia, citizenswith high livingstandardsto purchasehealth care
the Czech Republic, Hungary,and the Slovak Republic usingvoluntaryhealth insuranceand direct out-of-pocket
(Goldsteinand others 1996,p. 34). payments.This goalcan be achievedby explicitlyspecify-
ingthe packageof medicalbenefitsundermandatoryhealth
Publichealthexpenditurestructure insurance-that is, by determining a clear-cut border
betweenpublic and private finance.
The relativecontributionof the twosourcesof publichealth
funding is shownin table 2. The contributionof the pay- TABLE2
roll tax is increasing,but not to the extent that mighthave Publichealthrevenueinthe Russian
Federation
been expected.In 1995it providedjust 26 percent of pub- bysource,1992-95
lic health revenue; generalbudget revenue (mostlyfrom (percent)
localbudgets)is stillthe mainsourceof healthfinance. Source 1992 1993 1994 1995
This outcomeis contraryto the designof health revenue General revenue
budget 100.0 88.9 76.5 73.9
underthe Bismarckmodel, in which a payrolltax is sup- budget
Federal 17.8 9.5 9.1 6.3
posed to be the main sourceof revenue,with the govern- Regional
andlocalbudgets 82.2 79.4 67.4 67.6
Payrolltax - 11.1 23.5 26.1
ment providingsmallsubsidies.A payrolltax provides90 Total 100.0 100.0 100.0 100.0
percent of public health revenue in France and the Source:
Estmate
based
onKorchagin
1996.

FROM BEVERIDGE TO BISMARCK: HEALTH FINANCE IN THE RUSSIAN FEDERATION

67 I I
 The main sourceof privatespending-direct payments
by enterprises that run their own medicalfacilities-has
fallendramatically(by half overthe past four years)as the
result of the economiccrisis.This dedine in direct health
provisionby employershas not been compensatedby the
growthof voluntaryhealthinsurance.Moreover,directsocial
servicesfrom employershavedecreasedsubstantially
Morethan 500 privateinsurancecompaniesunderwrite
voluntaryinsuranceplansthat provideenrolleeswith addi-
tional servicesand accessto the best hospitals and poly-
clinics.But coverageis limited to 3-5 percent of health
care expenditures.Voluntaryplans are popularin Moscow,
St. Petersburg,and other big industrialcenterswith well-
equippedteachinghospitalsand high livingstandards,but
are less developedin the rest of the country.Most medical
facilitiesare not involvedin these plans becauseinsurers
prefer to work with the medicalelite.
The role of formal private out-of-pockethealth fund-
ing is negligible-no more than 1.5 percent of total health
spending.Most privatemoneycomesto the health system
throughthe blackmarketas under-the-tablepayments,par-
ticularlyin bighospitalsin urbanareas.Patientsare increas-
inglychargedfor drugs,materials,andsurgery.In addition,
most pharmaceuticalsfor outpatientuse are purchasedout
of pocket,althoughsomegroupsare exemptedfrom direct
payment.Thusprivatespendingis substantially understated
becauseit is impossibleto measure informalprivate (out-
of-pocket)paymentsfor items that are not fullycovered
by national health finance statistics. There has been no
attempt to conduct a large-scalehousehold expenditure
survey to assess the scale of out-of-pocket health
expenditures.
TABLE3
Total health revenue in the RussianFederation by
source, 1992-95moeofudriigThBeeiganBsacko-
(percent)
Source 1992 1993 1994
Public 73.5 84.5 88.3
Tovtealb 1200 100.0 100.0
Totalrevenue/GDP 3.7 4.5 4.8
a.Sumof generalbudgetrevenue(federal,regional,and local)andmandatoryhealth
insurancecontnbutions
of employers.
b. Sumofconsumerhealthexpenditureanddirecthealthexpenditureof enterprises.
Source:Estimatebasedon Korchagin1996.
INNOVATIONS IN HEALTH CARE FINANCING
The private sector's rolein total healthexpenditurehas
been declining(table 3). The health system still lacks an
effectivemechanismto raiseprivatemoney.The mainrea-
son is that public commitmentsto the coverage,eligibility,
and comprehensivenessof health care under mandatory
health insuranceare too declarativeand are not based on
an actuarialapproach.The lackof specificationforthe pack-
age of medicalbenefitsunder mandatoryhealth insurance
alsolimitsthe developmentof formalprivate finance.
In summary,the expectationsof additionalfundinghave
not been met. The payrolltax has brought insignificant
growthto health sectorresources.There has been no shift
in the formalpublic-privatemixof health finance.Private
finance and provisionare not structured as a component
of national health policy,and have been developingin an
unregulated,chaoticmanner.
Impact on Equity
Thetransitionto mandatoryhealthinsuranceraisesthe issue
of its effect on the equity of health care provisionfor dif-
ferent groupsand on the redistributionof income.A pre-
vailingattitudein the literatureis that the generalrevenue
approachensuresa higherdegree of incomeredistribution
and socialsolidaritythan do socialinsurancemodels.Two
argumentsare usuallypresented. First, an income tax is
more progressivethan a payrolltax. Second,the Beveridge
modeltendsto poolrisksmore equitablythanthe Bismarck
model,whichhas troublepoolingrisksamongmultiplesick-
nessfunds(Ensor1993;Klugmanand Schieber1996;WHO
1996).
However,the effect a health care financingmodel has
on equityis more dependent on the designof the system,
particularlythe way resources are allocated,than on the
mode of fund raising. The Beveridge and Bismarck mod-
elsmayormaynot be designedto poolrisksby equalizing
financialresourcesacrossgroupsand areas. Health insur-
1995 ance systemsin Germany,the Netherlands,and elsewhere
83.5 used risk-adjustedcapitationformulasto ensureequitable
16050 allocation of resources (VanVliet and Vande Ven 1992; Files
3.9 and Murray1995).Yetsome countrieswith the Beveridge
modelhavenotmademuchprogress
in equalzingresources,
and forseventyyearsthe Russianhealthcare system(using
the Beveridgemodel)failedto achieveequitablehealthcare
68
provisionfor differentgroups.The eliteof communistsoci-
ety had accessto high-qualitymedicalfacilities,employees
of big enterprisesenjoyedhigh standardsof healthcare in
the facilitiesowned by their enterprises,and high-income
groupshad better accessto the best medicalfacilities.
The effect a transition to a payrolltax has on income
redistributionalsodependson a country'sstructureof tax-
ation. If the earmarked tax replacesa highlyprogressive
incometax, the impacton equity is clearlynegative.But
the replacement of highlyregressiveindirect taxes may
increasesocialsolidarity.In Russiathe structureof tax rev-
enue is skewedtoward indirecttaxes.The valueadded tax
provides44.1percenrof consolidatedbudget revenueand
excisetaxes provide 19.8percent,whilethe profit tax sup-
plies 18.3percent and personalincometaxes just 2.6 per-
cent (Izvestia1996).Giventhistaxationstructure,the new
payrolltaxes mayhave a neutral or even slightlyprogres-
siveeffect.
In addition,it is easierto collectpayrolltaxesthan profit
taxesintransitioneconomieswithweak tax collectioncapac-
ity,because it is harder for entrepreneursto hide payroll
than profits.In Russiamandatoryhealth insurancetax col-
lection provides about 90 percent of expected revenue
(FederalFund of MandatoryHealthInsurance1995),while
the rate of incometax collectionis so low that in late 1996
apresidentialdecreeon emergencymeasuresto collecttaxes
was required.However,the amount of payrolltaxes actu-
allycollectedis reducedby a number of factorsinherentto
transitioneconomies,such asnonpaymentof formalwages
and underreportingof actualwages,particularlyin small
businesses.
One feature unique to Russiais that the health payroll
tax has not increasedthe tax burden of employers.Instead,
itreplacedaportionofthe old-agepayrolltaxwhenit became
clear in 1993 that the pension fund had surplus funds.
Althoughabasicpackageofmedicalbenefitswasestimated
to require 7.2percent of payroll,the healthsectorreceives
only3.6 percent;the rest is coveredbygeneralbudget allo-
cationsforhealth. The shareof socialcostsin payrollis 41
percent-lower than in some Western European coun-
tries,but.high for a transitioneconomythat is starvingfor
investment resources. High labor costs have narrowed
resourcesfor investmentand growth.Moreover,this bur-
den discouragesemployersfrom creatingadditionaljobs,
FROM BEVERIDGE TO BISMARCK: HEALTH
69
thus forcingjobs into the black market or increasingthe
numberof unemployed.Althoughit is unrealisticto expect
an increasein the tax burden, tax revenuesourcescouldbe
restructured,with a shift from a generalto an earmarked
healthtax.
The negativeeffect reformhas had on equityin health
care provisionis evident from data on the geographicdis-
tributionof healthfunds.Therehas beena sharpturn from
highlycentralized,top-downresource allocationto exces-
sivedecentralizationof health finance and management.
The Ministryof Health does not subsidize the oblasts.
Moreover,the subsidies provided to the oblasts by the
Ministryof Financeare not earmarkedfor health, and are
usuallyallocatedby oblast governmentsaccordingto tra-
ditional budget priorities. Given the great differences
between regions' industrial bases and taxation revenues,
the gapbetweenrich andpoorregionsis growing.Per capita
healthexpendituresrangefrom 720,000rublesin Moscow
to 470,000rublesin FarEast oblaststo 130,000in Northem
Caucusesoblasts(Kravchenko1996,p. 54).
Tosomeextentthese inequitiesare offsetby the federal
mandatoryhealthinsurancefund,whichprovidessmallsub-
sidies to seventy of the eighty-eightoblasts. But these
resourcesare insufficientto fund the equalizationprogram.
At the oblast level,territorial mandatoryhealth insur-
ance funds provide some equalizationof health funding
acrossareasusingcapitationformulasto allocateresources
to insurancecarriers.However,the equalizationcapacity
of these funds is limited because they control less than
one-thirdof healthrevenue.The remainingfundingis allo-
cated by local governments,with little or no risk pooling
between local areas. Few oblasts collect enough taxes in
the mandatoryhealth insurance fund to implement the
equalizationprogram.
The lack of a uniform oblast fundingpolicyand man-
agement strategyhas limitedaccessto healthcare in rural
areas and small towns. Major city governmentshave no
interest in providingspecialtycare to residentsfrom areas
wheresuch capacitydoes not exist. They maketheir bud-
getswithno regardto the needsof outsideresidentswithin
the oblast.
Growinginequityin the Russianhealth systemcannot
be attributed to the transitionto mandatoryhealth insur-
ance. It is the result of the excessivedecentralizationand
FINANCE IN THE RUSSIAN FEDERATION
fragmentationof health systemsthat started beforeimple-
mentationof mandatoryhealthinsurance.Federaland ter-
ritorialmandatoryhealth insurancefunds are lookingfor
waysto equalizefinancialresources,but theycannotoffset
the isolationiststrategiesadoptedby local governments.
Mandatory Health Insurance, Contracting
Arrangements, and the Efficiency and
Quality of Care
The recentinnovationsin the Russianhealthsector reflect
the shift to contractualinteractionsbetween health fund-
ing authoritiesand health care providers.Actingas third-
party purchasers of health care, insurers change the
performanceof providers.Althoughcontractualarrange-
ments are possibleunder the Beveridgemodel, the scope
for contractingis higher under the Bismarckmodel.First,
anindependentinsurercaninteractwithproviders(orgroups
of providers)onlythrougha contractualmechanism.Second,
a third-partypayeris more flexiblein its purchasingpolicy,
sinceit has no obligationsto state-ownedmedicalfacilities
and mayreject the servicesof inefficientproviders.
In theory,the divisionbetweenfinanceand provisionis
clear-cut.Moreover,contractinghas the potentialto improve
health sectorperformanceby decentralizingmanagement,
improving health care planning and management, and
increasinglocalchoiceof healthproviders.In practice,the
effectsof contractingdepend on many preconditions,the
mostimportantofwhicharethe roleof purchasersin encour-
agingcompetitionamongproviders,the designof contracts
andmethodsof payment,the adequacyandstabilityof fund-
ing, and the skills to manage the contracts (Savas and
Sheimanforthcoming).
In Russiacontractualrelationshipsare a growingpart of
the healthsystem.The main outcomeof this transitionhas
beenincreasedoperationalautonomyofproviders.Providers
are stillownedby the statebut are nowself-governing enti-
ties that can sell their services to different purchasers.
Hospitalsand polyclinicscan keep surplusfunds,hire and
firemedicalpersonnel,reducebed capacity,deploynewunits
(likeoutpatientclinics),anddetermineemployees'paywith-
out authorizationof the health administration.Afterseven
decadesof the Soviethealth system,health managersnow
havean intefestin coll&tingmanagementinformation,pric-
INNOVATIONS IN HEALTH CARE FINANCING
70
ing services,and increasingtheir workload-efforts that
are essentialto fulfillingcontractualobligations,ensuringa
surplus,and increasingthe salaryof medicalpersonnel.
The problem is that the scopeof contractingis stillrel-
ativelynarrow,withbigdifferencesacrossregions.For exam-
ple, in Samara oblast around 80 percent of health
expendituresare contracted;in most oblastsno more than
30 percentare.Manyhealth authoritiesare reluctantto sur-
render control over resources and tend to allocatefinan-
cialresourceson a noncontractualbasis,makingcontracting
inconsistent.Contracting'seffecton efficiencyand quality
can be evaluatedby examiningnew paymentmethods,the
newroleof managementinformation,mechanismsforqual-
ity control and consumerprotection, and administrative
costs.
New payment methods
Contractinghas encouragedthe developmentof perfor-
mance-related payment methods. Of hospitals working
under mandatoryhealthinsurance,53 percent are paid by
insurersbasedonratesforeachinpatientdiagnosis.Another
7 percent are paid a flat rate accordingto the averagerate
by specialty.These two methods are used for 6.2 million
cases-64 percentof inpatientcases.Thisapproachhas cre-
ated incentivesto increasethe occupancyof hospitalsand
reduce the averagelength of stay.Length of stayin these
hospitalsis lowerfor eighteencost categoriesthan in hos-
pitals that are still paid using traditional methods
(Langenbrunnerand others 1996,p. 164).
In outpatientcare, 19.4 percent of polyclinicsare paid
accordingto a capitation method, 14.0 percent for each
episode of outpatient care, 12.2 percent on a fee-for-ser-
vicebasis,and 10.0percentusingsomecombinationof these
(Sheiman,Shevski,andZelkovitch1996).Polyclinics work-
ing under capitationtend to shift to provisionof primary
care andpreventiveservices,whilefee-for-service payments
encouragethem to increasethe number of services.Fee-
for-servicepaymentswere used in Moscowfor three years,
resultedin overutilizationof someservices,and gavewayto
capitation.Still,the negativeexperiencewashelpfulforboth
insurersand polyclinics.Fee-for-servicepaymentsencour-
age the developmentof informationsystemsthat are used
to monitorpaymentand management.Moreover,the shift
from input- to output-basedindicatorsis helpingto reduce
unnecessarybed capacityand to create outpatient clinics
and other alternativesto costlyinpatientcare.
The effect of performance-relatedpayment methods
on allocativeefficiencyand the structure of health care is
much more controversial.Decadesof bureaucraticcontrol
over health care systemscreatedsubstantialdistortionsin
the structure of health care provisionthat are unknownto
Western countries. Estimates of inappropriate inpatient
casesrangefrom20 to 35percent.Inpatientcarestaysaver-
age3.7days;in the UnitedKingdomthe averageis 2.0 days,
in the UnitedStatesit is 1.2days,and in the mostsuccessful
managed care settings it is 0.3 days (OECD 1993;
Kongstvedt1993).
The main cause of long staysis the excessivenumber of
physiciansand excesscapacityin medicalfacilities.Such
capacityis nearlyimpossibleto maintainin a financialcri-
sis. Mandatoryhealthinsurancehas not yet contributedto
the reductionof excesscapacityand dismissalof unneeded
personnel.The failureto lowerexcesscapacitycanbe attrib-
uted to a lack of skillsin planningand utilizationmanage-
ment. The prevailingmode of interactionbetween health
authoritiesand healthprovidersis still input-basedalloca-
tion of financialresources.Insurers,as purchasersof care,
contract providerspredominantlyon a cost-per-casebasis
withoutproperlyanalyzingutilizationand planningvolumes
of care. Contractingis noncompetitiveand lackscost con-
tainmentmechanismssuch as globalbudgeting,analysisof
appropriatenessof inpatient cases, and so on. Cost and
volumecontracts,whichimplyplanningforvolumesof care
and linkingthat informationto availableresources,are not
used. As a result limitedresources are spread across too
manymedicalfacilities.
This conservativepolicyis aggravatedby the peculiari-
ties of rate setting.The prevailingapproachis to exclude
utilitiesand some other fixed costswhen calculatingrates
of payment.These expendituresare directlypaid by local
governments.In other words, most providersdo not pay
anythingforheatingand electricity.Thustheyare not inter-
ested in closingfacilities,even if there is excesscapacity.
The main reasonfor the conservativepolicyis a deeply
rooted strategyto build the health systemby establishing
new facilitiesrather than to improve the performanceof
existingfacilities.Despite financialcrisis,this approachhas
FROM BEVERIDGE TO BISMARCK: HEALTH
71
not changed.In theirreports,federalpolicyymakers
stillpraise
themselvesfor"maintainingandstrengtheningthe network
of medicalfacilities."Accordingto a recent statement by
the Ministryof Health, an additional34,500hospitalbeds
were put in placeduring 1993-95 (Tzaregorodtsev1996).
There is evidencethat some purchasersare interested
in modern costcontainmentmechanisms.For example,in
Kemerovooblastratesare regulatedin order to avoidinap-
propriateadmissionsandstrengthenprimarycare.As a result
the share of inpatient care spending in total spending
dropped from 64 to 60 percent over the past three years.
To reduce excesscapacity,Kemerovo'shealth committee
and mandatoryhealth insuranceinsurersintroduced uti-
lizationmanagementprocedures,withan emphasisonreduc-
ing the use of hospitalresources.Excesshospitalcapacity
wasidentifiedand three hospitalswere closedwithina few
months. In Samaraoblast the financialscheme of "poly-
clinicas fund holder" stimulatesprimarycare providersto
assumethe mainburden of health care (Galkin1995).The
schemehas helpedlowerthe shareof inpatientcare expen-
ditures. Other regions are experimentingwith different
approachesto managedcare in order to overcomestruc-
tural distortions.
The USAID-sponsoredZdrav reform program, with
Abt Associates,Inc. as a major contractor,helped develop
managed care principles and techniques in six oblasts
and several cites during 1994-96. Currently, Kaiser
Permanente International and Boston University,as new
contractors, are implementing a new program focusing
on the samemechanisms.Effortsto disseminateoutcomes
of the Zdrav reformprogramhave started, as has involve-
mentat the federallevelthroughthe Dumaand the Ministry
of Health.
Thus Russiahas followedan approach used in many
countries:startingwithrelativelyopen-endedperformance-
related payment methods and then replacingthem with
more advanced managedcare approachesand cost con-
tainment mechanisms.
New roleof management
information
One of the most beneficialoutcomesof mandatoryhealth
insurance has been a growing demand for management
information.When financialmechanismschanged,invest-
FINANCE IN THE RUSSIAN FEDERATION
ment in clinicalandfinancialinformationsystemsincreased.
Nearlyallterritorialmandatoryhealthinsurancefunds and
insurershaveestablishedinformationsystems.Health care
purchasingis increasinglybased on data on utilizationand
costs acrossmedicalfacilities,specialties,patient groups,
and evendiagnosticgroups.Data on cross-boundaryflows
of patients are alsocollected.
Physicians,health purchasers,and decisionmakersare
increasinglyusingcomputersystemsto communicate,mon-
itor, educate,acquiredata, keeprecords, checkbills,store
information,analyzedata, and supportdecisions.The sys-
temsaddbasicvalueto patientandmanagedcare,provider-
payercontracting,and financialmanagement.
Quality control and consumer protection
Mandatoryhealthinsurancefunds and oblast healthcom-
mitteeshavedevelopedmedical-economicstandardsthat
specifyrequirementsforthe processand outcomeof health
care foreach diagnosis.Insurershaveestablishedunitsthat
reviewthe cases and impose sanctionson providersthat
violatestandards.The units alsoidentifycasesof delayed
admissionto hospitalsand penalizepolycinicsforthe delay.
Insurers alsoscrutinizeinpatientcaseswherethe length of
stayis substantiallyshorter than the norm specifiedby the
standard.The scope of this work is growing.For example,
in Kemerovooblast 8 percent of inpatient cases are sub-
ject to quality control by insurers. About 70 percent of
collectedpenaltiesare returned to medicalfacilitiesto sup-
port quality improvements (KemerovoDepartment of
Health 1995).
Insurersareincreasinglyactingaschampionsof patients'
interests. They have set up specialunits that are respon-
sible for settling patients' claims(includingcourt cases),
monitoringpatient satisfaction,and recommendingqual-
ity improvements.Independent expertisebringsdiscipline
into the system,makingphysiciansimprovetheir perfor-
mance.
Thepreoccupationwithuniformrequirementsandnorms
is one drawbackof these innovations.In some casesthese
requirementsmakephysicianstoo defensiveof their clini-
cal practiceand impede innovations.Moreover,it is hard
for insurersto ensurecompliancewith standardswhenser-
vicesare severelyunderfunded.Theseeffortsare alsobeing
INNOVATIONS IN HEALTII CARE FINANCING
72
criticizedfor their emphasison imposingpenaltiesrather
than ensuringquality.There is growinginterest in devel-
opingqualityassurancesystemsbasedon continuousqual-
ityimprovementmodels.There is alsointerestin integrating
such systemswith paymentmechanisms.
Administrative costs
Most oblastsuse a pluralisticmodel of purchasingwith a
numberof insurers.Mandatoryhealthinsurancefundsspend
2.6 percent of collected premiums on administration.
Independentinsurersspend 3.9 percent, for a total of 6.5
percent. Such levelsare comparableto those in Westem
countries (Poullier1992). In addition,mandatoryhealth
insurancefundsuse (temporarily)free resourcesforbank-
ingoperations,whichhavehighfinancialreturns givengal-
loping inflation and high interest rates (the investments
made by mandatoryhealth insurancefunds are protected
by regulation).Federal mandatoryhealth insurancefund
managers claim that the revenue from these operations
exceedsadministrativecosts (Kravchenko1996). This is
only partlytrue, becausepremiumsmighthavebeen used
for financialoperationsnot onlyby purchasersbut alsoby
providers (presumablywith a smaller return due to an
absenceof banking skills).
It is hard to draw anygeneralizationsabout the size of
administrativecosts. These costs should clearlybe taken
into account when analyzingthe effect mandatoryhealth
insurancehas had on the performanceof the health care
system.It is also clear that Russiacannot afford 545 pri-
vate insurerswith luxuriousofficesand well-paidemploy-
ees. In general,multiplefunds havehigher administrative
costs (Poullier1992).
Intensedebatesareunderwayonthisissue.Amendments
to the Health Insurance Act have been submitted to the
Duma that would replaceprivateinsurerswith mandatory
health insurancefunds as major purchasers.An alternative
approach,taken in Moscow,is to decreasethe number of
insurersby mergingsmallentities. Moscow'snetwork of
twenty-fivecompaniesis beingmergedto form eightlarger
ones that willshare the mandatoryhealth insurancemar-
ket on a cartel basis. This approach will likelydecrease
administrativecosts, but at the cost of less competition
amonginsurers.
Implementation Issues governmentspayfor inpatientcare),items of expenditure
(utilitiesand equipment are paid for by the government,
Obstacles
to successful
reform whileotheritemsare paidforbyinsurers),populationgroups,
andspecialties.Thesedistinctionsimpedeplanningandman-
Healthreformin the RussianFederationfacesmanyobsta- agement of care focused on more cost-effectivearrange-
cles.Someare objective,othersare the resultofpoor design ments.In addition,providersact under differenteconomic
in the finance and managementof the system. regimes:contractual(withinsurers)andnoncontractual(with
localgovernments).This approachlimitsthe applicationof
Economicand politicalinstability.Economicinstability performance-related methodsand soreducesthe efficiency
aggravateshealth systemunderfunding,makingit difficult of the contractualpart of providerperformance.
to introduce performance-related payment methods.
Providers are respondingto the new payment methods, Littleornomarketpressure. The transitionto healthinsur-
but incentivesin the informalmarket are much stronger- ance has introduced elementsof competition,but mostly
simpleunder-the table paymentsare easier than sophisti- on the purchasingrather than the provision side of the
cated formalpaymentarrangements. market.Independentinsurerscompetevigorouslyto expand
The unstable political situation also impedes radical the mandatoryhealthinsurancemarket.The maintools of
changesin the system.Local policymakersmay recognize competitionare contractswith the best-equippedmedical
the need to close some facilitiesand fire some personnel, facilities,sound consumerprotectionprocedures,and effi-
but they are reluctantto do sobecauseelectionstake place cientqualitycontrol.But littlehas been done to encourage
twicea year. competitionamongproviders.In primarycare there is no
altemativeto largestate-ownedpolyclinics;In the hospital
Excessive publiccommitments.Giventhe currenteconomic sectorthere areelementsof competitivepurchasingbyinsur-
situation,the governmentshouldno longertryto deliverfree ers, but they are neutralizedby the policyof local govern-
care to all groups.Excessivestate guaranteesdistort eco- mentsand healthcomrittees to supporthospitalsregardless
nomic relationships in health care because unbalanced of their performance.
mandatoryhealthinsuranceprogramshinderfull-scalecon- The privatesectoris the onlyarea with substantialcom-
tracting between purchasers and providers.In addition, petition.Hospitals,polyclinics,and privatehealthinsurance
free care is becomingmore elusiveand publicdiscontentis companies are competing for patients and subscribers.
growing.Havingdeclaredhealthcareentirelyfreeof charge, Voluntaryhealth insurance plans are offering additional
the governmentis losingits capacity to provide it to the services.
groupsthat need it the most. The healthbureaucracytends
to refer to the "constitutionalrights" of citizens for free Lackof management capacity.Despite recentprogressin
careratherthanlookforbetterwaysto attractprivatemoney. buildingcapacity,purchasersstilllackeffectivemanagement
skills.Contractingrequiresskillsthat werenot neededunder
Poorlyspecifiedrolesand responsibilities
of thegovernment directpublicprovision,includingidentifyingcost-effective
andmandatoryhealthinsurance funds. The establishmentof medicalinterventions,planningvolumesof care,and nego-
mandatoryhealthinsurancefundsandinsurancecarriershas tiating,evaluating,and monitoringproviders'performance.
dividedthe rolesand responsibilities
for healthfinanceand These skillsare especiallyneeded at the middle and bot-
provision,creatingmisunderstandings andconflictsbetween tom levelsof the system,wherecapacityis extremelyweak.
old and new actorsin the healthcare system.The systemis
fragrnentedboth verticallyand horizontally.Localgovem- The proposedstrategy
mentsandmandatoryhealthinsurancefundsin mostoblasts
separatetheirresponsibilities
forfundinghealthacrosshealth The followingpoints are drawnfrom "HealthFinanceand
systemsubsectors(insurersmaypayforoutpatientcare,local ManagementReformStrategyin the RussianFederation,"

FROM BEVERIDGE TO BISMARCK: HEALTH FINANCE IN THE RUSSIAN FEDERATION

73
writtenin 1996bya groupof independentexpertsunderthe puttingnonacutecases on a waitinglist.Peoplewho want
USAIDZdravreformproject,withBostonUniversityas the to reduce their waitingtimes should pay for part or all of
maincontractor(Starodulov,
Sheiman,andZelkovitch1996). medicalservicecosts (mostlythrough the voluntaryhealth
insurancesystem).
Reformingstatecommitments.State obligationsto pro-
vide free care shouldbe balancedwith availableresources. Shzftingexcess
capacitytotheprivatesector,
especiallyforinpa-
Declarationsguaranteeingcomprehensivehealthcare must tientcarein bigcities.A preconditionfor implementationof
givewayto an actuarialapproachthat developsa package thisapproachis an efficientpurchasingpolicybased on the
of medicalbenefitsunder the basicmandatoryhealthinsur- analysesandplanningcapacityof both short-and long-term
ance program.Major changes are needed to implement providers.This wouldrequirerestructuringthe hospitalsec-
the new approach. tor mergingsomefacilitiesandchangingtheir casemix.The
strategyalsoproposesrestructuringbig polyclinics
into free-
Improvingmanagement of inpatientcare,withan emphasis standingphysicianpracticesto increasecompetitionbetween
on avoidinginappropriateadmissionsand tighteningrequire- physiciansand phaseout the leastcompetentones.
mentson lengthof stay.This willrequire better planningof Some of these proposalshavebecome part of the cur-
volumes of inpatient care as the main componentof the rent healthpolicyagendaat the federal level.The Dumais
purchasingpoliciesof insurers and local governments.A reviewingproposed amendmentsto the Health Insurance
packageof medicalbenefits should be based on realistic Actprovidingfor anewprocedureto determinethe volume
estimatesof utilizationand identificationof excesscapac- of state guarantees.The basic mandatoryhealth insurance
ity.Accordingto preliminaryestimates,a 12-15 percent program for 1997has been drafted by a group of experts
drop in the number of bed-days(whichcan be achievedin based on the conceptsof resourcemanagementand actu-
1997)would easethe most dramaticshortagesof funding, arialapproaches.As mentionedabove,Kemerovooblastis
and thereby partly balance the basic program of manda- experimenting withplanninghealthcarevolumesanddevel-
toryhealth insurance. opingmore cost-effectivealternatives.

Adoptinga newprocedureforestablishingthe basicmanda- Integratingthesystemverticallyand horizontally.

toryhealthinsuranceprogram.
The minimumsocialstandard,
set by the USAID program, must be approved by the
Governmentandthe Dumatogetherwiththe premiumsize
and budgetallocationsrequired to support this standard.
An analogousprocedureof determiningthe state's duties
shouldbe acceptedat the oblastlevel.The volumeof state
budgetallocations
to implementthe oblasts'mandatoryhealth
insuranceprogram should be determined by subtracting
employerpremiumsfromthe approvedprogramcost.
Whenestimatingcostsfor the territorialmandatoryhealth
insuranceprograms, determiningthelevelof mismatching with
availableresourcesand usingthe datato developcostsharing.
To minimizenegativesocialimplications,it is appropriate
to set limitson primarycare and to provide free medical
servicesto people who cannot afford evenminimumcon-
tributions.Specificforms of cost sharinghave been pro-
posed. There is also the option of rationingutilizationby
The goal
ofthiseffortisto centralizemanagementandfinancerespon-
sibilitiessufficientlyto integratethe oblasthealth care sys-
tem and equalize resources across local areas. Two
mechanismsare proposed.First,establishingan earmarked
federalfund for resourceequalizationbased on mandatory
healthinsurancecontributionsaswellas federalbudgetrev-
enues,with a clear-cutequalizationformulaand closecoor-
dinationbetween the federal mandatoryhealth insurance
fund and Ministriesof Health and Finance.Second,at the
oblast level,centralizingthe mandatoryhealth insurance
contributions of local governmentsand allocatingmost
healthrevenueon a capitationbasis.
In addition, the multisource model of funding care
providersthrough mandatoryhealth insuranceprograms,
healthcommittees,and localgovernments(by typesof ser-
vice,budgetitems,andpopulationgroups)shouldbe scrapped
in favorof a one-channelmodel.Allcashinflowsto the pub-
lic health system shouldbe pooled at a levelhigher than

INNOVATIONS IN HEALTH CARE FINANCING

74
individualcareproviders.Specifically,
managementof 70 to
80percentofthe fundsshouldbe carriedoutby asinglefund-
ingparty-that is, competinginsurers.Mergingcash flows
willallowcompetinginsurersto build up a reasonablecare
providercontractingmodel,work out efficientincentives,
improvethe care deliverystructure,and controlutilization.
Shiftingtopayrolltax asthe mainsourceof healthrevenue.
With mandatoryhealthinsurancefunds controllinga neg-
ligibleportionof funding,it is hardlypossibleto claimthat
a transitionfrom the Beveridgeto Bismarckmodel is tak-
ing place.Rather,a transitionalmodelis emergingthat has
no analogin the world.Eventuallyit shouldbe transformed
into either a tax-financedor a socialinsurancemodel.
The existingfund-raisingschemeshouldbe modifiedto
emphasize earmarked taxes. Their share of total health
revenuesshouldincreasesubstantiallyfromthe current26
percent to at least 50 to 60 percent.To achievethis goal,
the payrolltax should be increased, althoughthe overall
tax burden should be lowered (for example,by lowering
incometaxes).This impliesa transitionto the fund-raising
schemeused under the Bismarckmodel,with employees
obtaininginsurancefor themselvesand their dependents.
Severalargumentsfavorthis shift:
* The health sector remains one of the last budget prior-
ities. An earmarked payrolltax can be a reliableand
predictable source of health revenue, even in a declin-
ing economy.
Less dependence on general budget revenues will allow
local governments, health committees, and mandatory
health insurance funds to clarify their roles and respon-
sibilities. Collecting most revenues, mandatory health
insurance funds will serve as the main funding party,
responsible for the integrated purchasing policy
* A shift to a payroll tax as the main source of funding will
result in vertical integration of the health system, since
under current legislation the centralization of financial
resources can be achieved onlythrough mandatory health
insurance channels.
Conclusion
The transition to mandatory health insurance has had a con-
troversial effect on health sector performance. Clearly, it
FROM BEVERIDGE TO BISMARCK:HEALTII FINANCE IN THE RUSSIAN FEDERATION
75
has induced positivechangesthat wouldbeen impossible
underthe former"frozen"system,withitsemphasison com-
mand and controlmethods.Yetthe experimentationof the
late 1980s,with health committeesas purchasersof care,
has not been supportedby the healthbureaucracybecause
of a feature inherent to the ownersof medicalfacilities-
atendencyto controlprovidersdirectlyandallocateresources
on a noncontractualbasis. Independent insurers as pur-
chasersof carehavestarted the transitionfromahighlyinte-
gratedto a contractualmode ofrelationshipswithproviders.
The reformhas helped stabilizefundingduringa period
of economiccrisis,encourageda transitionfrom input- to
output-basedpaymentmethods,introducedtougherqual-
ity controland elementsof consumerprotection,and cre-
ated new incentivesto collectinformationand use it for
decisionmaking.However,resourceallocationis becoming
lessequitable,the systemis disintegratingverticallyandhor-
izontally,and structural distortions in the health system
havenotbeen addressed.Thusreformhas not fullymetorig-
inalexpectations,but it has had somepromisingoutcomes.
The failuresof reformcan be traced to an unfavorable
economicand politicalenvironmentand to mistakesin the
designof the reform.The biggestmistake is the isolation
of mandatoryhealthinsurancefrom the overallhealth sys-
tem. Because this approach makes the system less coher-
ent, it impedesimplementationof the strategyfor phasing
out excess capacity and overcoming structural distortions
in health care provision.
The way health funds are raised cannot automatically
increase resources and change the performance of the health
system. Rather, the design of relationships between pur-
chasers and providers determine actual changes in the sys-
tem. At the core of the reform is a shift from an integrated
to a contractual model.
The lessons of the Russian reform for other transition
economies can be summarized as follows:
* Moving toward a payrolltax-based model may raiseaddi-
tional funding, provided the economy is not declining
too steeply and too long. But the change in the funding
method cannot be seen as the main area of reform. It is
much more important to ensure the operational auton-
omy of providers, to increase their support for health
policy objectives through performance-related payment
methods and managed care mechanisms.
* If the mandatoryhealthinsurancemodelis used as the
basis of health reform, mechanisms should ensure that
this model prevails and can be easily implemented. To
that end, payroll taxes should replace general taxes as
the main source of health funds, supplemented by state
subsidies (with little or no additional tax burden for
employers); mandatory health insurance carriers should
control most health revenue, with the governmentdirectly
responsible for a small portion of health care funding
and provision;
the roles and responsiblities of manda
and provision; the roles and responsibilities of manda-
' ~~~~~~~~~~~~Care.
tory health insurance funds and the government should
be clearlyspecified;and coordination mechanisms should
be developed. One predominant model of raising and
allocatingfunds may be the best way to avoid vertical
and horizontal disintegration of the health system.
*Contracting should be the prevailing model of health
care finance, planning, and management. The share of
revenue allocated directly by the government should be
limited to clearly specified expenditures or services.
Contracts should be designed in the context of health
policy objectives, the most important of which is to
eliminate the structural disproportion and excess capac-
ity of inherited systems.
* Regardlessof the fund-raising model, purchasers should
encourage competition among providers by allowing
more consumer choice, using competitive procedures
for contracting, and dismantling local monopolies wher-
ever possible. In addition, purchasers should ensure an
ever posibe.nddtio,urcasrssholdensre
open selection of providers in order to fund them for
the real value of their services,and carefully monitor and
evaluate providers' performance.
* Under the mandatory health insurance model it is crit-
ical to base state commitments on an actuarial approach
rather than on declarations about free care for all. 'This
requires planning of volumes of care, improving man-
agement of health care, and making bold political deci-
sions about cost sharing or rationing of health care.
References
Ensor, T. 1993. "Health System Reform in Former Socialist
Countries of Europe. " International
Journalof Health Planning
and Management8: 169-87.
FederalFund of MandatoryHealth Insurance. 1995. "MHI in
the RussianFederation:Figuresand Facts."Moscow.
INNOVATIONS
IN HEALTIICAREFINANCING
Files, A., and M. Murray.1995. "GermanRisk Structure
Compensation:EnhancingEquityand Effectiveness."Inquiry
32 (fall):300-9.
Galkin,R. 1995."Mine zhaluemsana zhizn."Meditsinskaya Gazeta
15(3): 2.
Goldstein,Ellen,AlexanderPreker,OlusojiAdeyi,and Gnanaraj
Chellaraj.1996. Trendsin HealthStatus,Services,and Finance:
The Transitionin Centraland EasternEurope.World Bank
TechnicalPaper 341.Washington,D.C.
Izvestia.4 December1996,p. 4.
KemerovoDepartmentofHealth.1995.QualityArsuranceofHealth
CaeKmro.
Kemerovo.
Klugman,jeni, and George Schieber.1996.A Surveyof Health
Reform in Central Asia. World Bank TechnicalPaper 344.
Washington,D.C.
Kongstvedt,P 1993.TheManagedHealthCareHandbook.2nd ed.
Washington,D.C.
Korchagin,V 1996. "Sostoyaniefinansirovaniyazdravoohrane-
ni"EcoiaZdaohnea6:0
nia.>}
EconomicaZdravoohranenia 6: 10.
Kravchenko, N. 1996.'Analizsostoyaniafinansirovania
zdravoohra-
nenia v 1995."Economica Zdravoohranenia4: 14-20.
Langenbrunner,J.,A.Wouter,T.Makarova,and K. Quinn. 1996.
"HospitalPaymentPoliciesand Reforms:Issues and Options
in Russia."TheJournalof HealthAdministrationEducation2
(spring):164.
OECD(Organization forEconomicCooperation andDevelopment).
1993.HealthSystems:FactsandTrends,1960-1991.Vol.1.Paris.
Poullier,J-P1992."Administrative Costsin SelectedIndustrialized
Countries."Health CareFinancingReview (summer):14-20.
Savas,S., and Igor Sheimnan. Forthcoming."ContractingModels
and ProviderCompetitioninEurope."In EuropeanHealth Care
Reforms. Geneva:WorldHealth Organization.
Reom.Gnv:WrdHatOgnito.
Sheiman,Igor. 1994."Formingthe Systemof Health Insurance
inThe RussianFederation."SocialScienceandMedicine39(10):
1425-32.
Sheiman,Igor,V Shevski,and R. Zelkovitch.1996."Methodsof
Payment for Outpatient Care." The Journal of Health
Administration Education 2(spring):182.
Shishkin,S. 1996. "Ne samie bednie sredibednih." Meditsinski
vsnc2:3
vestnic22: 3.
Starodulov,Vladirnir,Igor Sheiman,and RomanZelkovitch.1996.
"ConceptiaReformupravleniaI finansirovaniazdravoohrane-
nia Rossii."MeditsinskiVestnic18.
Tzaregorodtsev, A. 1996."Ob itogahdeatelnostizdravoohranenia
RossiskoiFederatsii,1993-1995."Econiomica
Zdravoohranenia
8: 5-6.
VanVliet,R., andWVandeVen.1992."Towarda BudgetFormula
for Competing Health Insurers." Social Science and Medicine
34: 25-40.
WHO (WorldHealth Organization).1996.RegionalHealthCare
Reforms:Analysisof CurrentStrategies.
Copenhagen.
76
Private Insurance: Principles
and Practice
DeborahJ. Cholletand MaureenLewis
P rivate health insuranceis a growingphenomenon
in muchof the world.Fueledby risingincomesand
growingdissatisfactionwithpubliclyfinanced(and
often publiclydelivered)healthcare services,privateinsur-
ance coveragetypicallybeginsamonglargecompanyman-
agers in white-collar industries, and in multinational
companieswith large numbers of employees.However,
informalinsurancearrangementsexistevenin the poorest
countries,reflectinga universaldesire for financialprotec-
tion and, in somecountries,accessto better-qualityhealth
care than is offeredin the public system.
A privateinsurancemarketcan offera number of advan-
tages over a purelypublic systemof healthcare financing.
It can allowgovernmentsto developand maintainsmaller
and targeted systemsof healthcare financingto servepeo-
ple who do not have accessto private insurance.Private
insurancecan helphealth careproviders(privateand pub-
lic)rebuildinfrastructureand amortizeneeded investment
whengovernmentpaymentsforhealthcareare inadequate.
In the best casesprivate insuranceencourageshealth care
systems-both the financingand the deliveryof care-to
innovateand to become more efficient,and offers a point
of reference for improvingthe quality and efficiencyof
care in the publicsystem.Typicallyencumberedby politics
and bureaucracy,governmentsmayfindit difficultto inno-
vate without a privatemarket to lead the way.
Despite these advantages,privateinsurancepresentsa
number of problems.First,health insurancecan be unaf-
fordablefor low-incomepeople.In principle,governments
can developsubsidiesto helplow-incomepeoplebuy insur-
directorat theAlphaCenterin Washington,
DeborahJ. Cholletis associate
andSocialDevelopment Groupof the LatinAmericaandtheCaribbeanRegionalOfficeat theWorldBank.Theauthorsaregrateful
to JeffreyHammer,GerardLaForgia,JackLangenbruner, andLenNicholsforhelpfulcomments.
77
ance, but such systemscan be difficultto design and to
administer.Second,insurersmaywant to denycoverageto
peoplewhoare sick,andto limitcoverageforhigh-costcon-
ditionsor services.By refusingpeople who are sick or by
deterringthem from seeking coverage,private insurance
can contributeto higher averagecosts in publicfinancing
programs that enroll larger populationsand serve as the
insurer of last resort. The selectionof low-costpatients
into private insurancecan cause seriousproblemsfor the
publicfinancingsystem(orthe healthcare deliverysystem
that it finances)if it cannotreadilyadjustto higheraverage
costs-even when total costs decline. Governments can
require privateinsurersto accept sickpeople, restricthow
insurerspricecoverage,andrequireinsuranceplansto cover
varioustypes of high-costhealth care. However,insurers
maybe unwillingto enter or remain in marketswith such
requirements.
Other problemswith private insurancederivefrom the
complexityof insurancecontracts.Consumerstypicallydo
notunderstandmanyaspectsofinsurancecontracts.Because
consumersusuallyare unable to detect whichinsurersare
unscrupulousor financiallyunsound,the marketcanattract
these insurers.Consequently,insuranceregulatorsplayan
essentialrole in stabilizinga competitiveprivate insurance
system.However, insuranceregulation is more art than
science.It shouldencourageinnovationsthat improveeffi-
ciencyand service,but it must require financialintegrity
anddiscouragepracticesthat threatenthe stabilityandeffec-
tivenessof the health insurancesystem-and, in turn, the
healthcare systemthat it supports.
in the Human
D.C.MaureenLewisis principaleconomist
 This paper has three sections.The first sectionreviews
privatehealthinsuranceprinciples-what privateinsurance
is,howit works,andwhyinsurancepracticestend to evolve
in particularways.It then discussesthe role of government
regulation-specifically,how regulationcan stabilizeand
guidethe performanceofprivateinsurancemarkets-draw-
ing on examples from OECD countries (especiallythe
United States,whichreliesthe most heavilyon voluntary
competitive private health insurance) and developing
counuties.
Principlesof PrivateHealth InsuranceMarketsand Regulation
Like allforms of insurance,healthinsuranceis a systemof
protection againstfinancialloss.In a health insurancesys-
tem a group of individualsagreeto pay certainsums for a
guaranteethat they willbe compensatedfor costs related
to the use of specifickinds of health care. Formalhealth
insurancecontractstypicallystipulatethat coveredhealth
care servicesmust be medicallynecessaryand providedby
appropriatelytrained health care professionals.Because
health providersgenerallyprefer to see patients who are
insuredrather thanrisknonpaymentforcare,in manycoun-
tries havinghealth insurance(privateor public)is equated
with havingaccessto health care.
The premiseof health insurance is simple:individual
healthcare needscan be unpredictableand costly,but rel-
ativelyfew people need health care at anyparticulartime.
Thus,by poolingthe risk of largehealth care expenditures
over many people, health insurancecan make necessary
health care affordableto all.
Private health insurance systems differ from public
financingsystemsin severalways.Most important, a pri-
vate insuranceplantypicallycompetesfor customers,either
with a public systemor with other privatehealth insurance
plans.Thusconsumersmaychooseamongplans that have
differentfeatures (more or less financialprotection, dif-
ferent accessto physiciansand hospitals,a greaterempha-
sison customerserviceandsatisfaction,and soon).Typically,
these plans alsohave differentprices,and consumershave
to decide if a more desirableinsurance plan is worth its
higherprice.In principle,competitionamonghealthplans
INNOVATIONS IN HEALTH CARE FINANCING
78
Thesecondsectionaddressesthe extentofprivateinsur-
ancecoveragein developingcountries,summarizesselected
countries'experiencewith privateinsurance,anddescribes
emergingregulationin these countries.
The final section offers conclusionsand a number of
lessonsfor developingeffectivehealthinsuranceregulation.
By meldingtheory,practice, and experience,we hope to
providea context for evaluatingthe role of private health
insuranceand fordesigningeffectivesystemsof healthinsur-
ance regulationin allcountries.
willtend to drive up the quality and drive down the price
of availableinsurance.But differentprices are often asso-
ciated with a range of plans that provide different bene-
fits and may be difficult to compare. People with low
incomesare unlikelyto be able to afford anyinsurance at
all, much less insurancethat offers them access to com-
prehensivehealth care.
Bycomparison,in a public financingsystemconsumers
typicallydo not chooseamonghealthplans.Moreover,pub-
licfinancingmayor maynot allowpatientsto chooseamong
hospitals,physicians,and other serviceproviders.A public
systemtypicallyis financedin largepart throughtaxes that
are unrelatedto the use ofhealthcarebut that mayberelated
to abilityto pay.Thusdifferencesin the qualityof coverage
maynotbe a problem.Instead,problemswithpublicfinanc-
ing usuallyderive from the absence of competition and,
therefore,the absence of incentivesfor public systemsto
respondto consumers.Public systemstypicallyhavelittle
incentiveto continuouslyimprovequalityandcustomerser-
vice and alsocontain costs.
Duringthe past two decadesmanagedcare plans have
emergedin a numberofcountries.Managedcareplanscom-
bine the financingand deliveryof health care in the same
contract,offeringenrolleesboth insuranceprotection and
a prescribednetworkof healthcareproviders.In industrial
countriesmanagedcareplanshavebeenofferedasa lower-
costalternativeto fnancial insuranceplansthat donot con-
strainparticipants'choiceof provider.But as enrollmentin
managedcare plans has grown,some plans are also striv-
ing to be known as a better-qualityalternativeto the frag- healthyand high-incomepeople buy private insurance
mented,fee-for-servicesystemofhealthcaredelivery.Unless and othersrelyon the public system.
otherwiseindicated,in thispaper healthinsuranceincludes Supplementalcoverage for servicesnot coveredby a uni-
both financialinsuranceplans (whichpay for coveredser- versalpublic insuranceprogram.Forexample,in the United
vicesfrom anyqualifiedprovider)and managedcare plans Kingdom,wherethe publicinsuranceprogramis popu-
(whichpay only for coveredservicesthat are deliveredby lar and providescomprehensivecoverage,people may
providerswho are under contractto the plan). buy privateinsuranceto financecarefrom specialistsin
privatepractice, "jumpingthe queue" for specialtycare
Rolesof Private Insurance in the public program. Similarly,in Brazilno one may
withdrawfrom the publicsystem,but some peoplebuy
Most countrieshave a private healthinsurance sector.In private insuranceto get more timely or higher-quality
general,private insurancetends to emergewhen the pub- carein the public system.In Australiaprivate insurance
licfinancingsystemis perceivedas financinglower-quality pays only for hospital care, either in private facilities
care (usuallyalso restrictingpatients' choice of provider) (whichoffer patients a choice amongphysicians)or in
or coversonly some types of health care. Reflectingthe public facilities. In the United States enrollees in
diversereasonsthat a privateinsurancesectorwouldemerge, Medicare,the socialinsuranceprogramfor the elderly
the role of privateinsurancevarieswidelyamongcountries and the disabled,can buy private supplementalcover-
that allowor encourageit. In general,these rolesare of three ageto payforthe publicplan'sextensivedeductiblesand
types: coinsuranceamounts,and to pay for majoritems (such
* Coverageforpeople whoareineligibleforpublic insurance. as prescriptiondrugs) that are not coveredby the pub-
For example,in the United States private insuranceis lic plan. About one-third of retirees have a private
considered the main source of coverage,while public Medicaresupplementinsuranceplan.
insurance is intended to cover groups whom the pri- Annex tables 1 and 2 provide additionaldetail about the
vate insurancemarket is likelyto fail-the elderly (in alternativeroles of privateinsurancein selectedcountries.
Medicare) and people who are unable to work (chil- In manycountriesa largeportionof healthcare spend-
dren, the elderly, and the disabled) and poor (in ingis financedprivately,eitherthroughinsuranceor out of
Medicaid).Peoplewho are ineligiblefor public insur- pocket.But rarelydoesprivateinsurancefinancemosthealth
ance do not alwaysbuy private insurance.Somerelyon care use. In the United States,where privatehealthinsur-
publichospitals(fundedby localgovernment)for care. ance is unusuallywell developed (but purchaseis volun-
In communitiesthat do not havea publichospital,peo- tary),private insurancefinancedonly about 37 percent of
ple without insurancemaybe unable to obtain routine all personalhealth care spendingin 1994;59 percent was
care. publiclyfinanced.The Republicof Koreais unique in that
* Coverage for peoplewho withdrawfrom a universal public privateinsuranceis mandatory,and it financesmost health
insurance program.For example,in Germanyindividuals care.
maywithdrawfromthe nationalpayrolltax-financedsys- Even in industrial countries where public health care
tem of sicknessfunds, whichoffer coverageto allresi- financingis universal,privateinsurancemay stillfinancea
dent workers,their families,and retirees. Peoplewho significantshareof healthcare.For example,in the United
withdraware not requiredto buy privateinsurance,but Kingdomprivate insurancefinanced14 percent of health
theyusuallydoso.Fewpeoplewithdrawfromthe national care in 1990. Similarly,in Canada (where public health
publicinsurancesystem,however,sincethey can never care financingis universaland privateinsuranceis prohib-
reenter it. In Chileinsuranceis compulsorybut individ- ited from coveringpubliclyinsured healthcare services)a
ualscan choosebetweenbuyingregulatedprivateinsur- sizableminorityof people buyprivate insuranceto finance
anceor relyingon publiclyfinancedand deliveredhealth services that are not covered by provincial programs.
care.Thisapproachhascreateda two-tiersystemin which Informationaboutthe sharesof healthexpendituresfinanced

PRIVATEINSURANCE: PRINCIPLES AND PRACTICE

79
TABLE
I
Private and pubic expenditures for personal health care services, selected countries
oftotalexpenditures)
(percentage

Privatehealthcareexpenditures Publichealthcare expenditures

Issuedor Out-of- Socialinsurance
Country prepaid pocket Total programs Other Total
Argentina - 23 - 36 22 58
Brazil,
1995 _ - - - - 75
Canada 20 - - 75 - -
Ecuador - 63 - 17 14 31
Egypt - - - 9 30 39
France 21 - - 75 - -
Gen-nany. 1985a 7 7 14 69 12 81
India,1990 _91b 3 75 78 - - 21
Jamaica 9 - - 35 - -
Jordan,1994 - - 53 39 8 47
Kenya,1994 - - - - 43
Niger 14 - 67 - -
Nigeria 44 - - 45 - -
Peru,1995 - 28 - 36 30 69
Africa,
South 1993-9
4c 37 14 55 - - 45
Tanzania 14 - - 68 - -
Thailand,1992 - 74 74 2 24 26
Tunisia 25 - - 67 - -
Uganda I5 - - 47 - -
United Kingdomd 13 - 14 85 2 87
UnitedStates,1994e 32 3 37 30 29 59
Uruguay 14 - - 76 - -
Note:
Unless
otherwise
indicated,
source
didnotprovidea reference
year.Inallcases
dataarethemostrecent
available.
4.3percent
a.Excludes ofexpenditures
financedfrom
othersources.
0.8percent
b.Excludes ofexpenditures
financedbyexternal
donors.
4 percent
c.Excludes ofprivate
spending
categorized
asindustrial
healthexpenditures.
I percent
d.Excludes ofprivate
spending
notallocatable
tocategories.
c.Excludes
3.4percent
ofprivate Social
fromothersources.
spending insurance figures
includeMedicare AandPartB.
Part
Source.
Lewis 1995;
andMedici Musgrove1996;Reinhardt1995;
WoddBank 1995 and1996; andothers1996;
Collins 1995;
Abel-Smith andothers1996;
Levit 1997;
Femandez
TA11997;
Nittayaramphong 1994;
andTangcharoensathien Rafehinthisvolume.

by private insurance, out-of-pocket expenditures, and pub-
lic insurance programs in selected countries is summarized
in table 1.
Concept of InsurableRisk
Risk is defined in terms of both the probability and the mag-
nitude of potential health care expenditures. A high-risk sit-
uation may entail a high probabilityof expenditure (regardless
of how great the expenditure may be), a high magnitude of
expenditure (regardless of the probability), or both.
In general, health care for any illness or condition that
occurs randomly among a population is insurable. But in a
number of high-risk circumstances health care maybe unin-
surable. In these circumstances insurers will be unwilLing
to offer coverage, or will design insurance contracts in par-
ticular ways-usually trying to package coverage for unin-
surable risks together with coverage for insurable risks.
The kinds of situations in which some or all health care
may be uninsurable are described below.
healthcarerisk
Nonrandom
Possibly the main reason that health care would be unin-
surable is if it were nonrandom. For example, during a war
or civil conflict health care risks are systemic: the likeli-
hood that any person will need health care is highly corre-
lated with the likelihood that many others will need care as
well. Similarly,in communities where serious, communica-
ble health problems have reached epidemic proportions (for
example, in communities with a high incidence of AIDS),
much health care may be uninsurable. In these communi-

IN HEALTHCAREFINANCING
INNOVATIONS
80
ties insurersmaybe unwillingto insuremuch of the popu-
lation,or theymayrefusehealthcareforthe specificinjuries
or illnessesthat are most likely(for example,those due to
war or civilconflict).
High-probability
health coreservices
Evenwhenthe incidenceof illnessor injuryisrandom,some
health care servicesmay be uninsurableif the probability
that peoplewilluse those servicesis very high.The reason
suchservicesmaybe uninsurablerelatesto howinsurance
prices are determined. Specifically,the price of an insur-
anceplanthat wouldcoverhigh-probability lossesmayequal
or exceed the cost to consumersof remaininguninsured,
evenif theycould affordto buy coverage.I When thisis the
case,private insurancefor those servicesmaynot emerge.
Instead, insurersmayoffer insuranceproductsthat specif-
icallyexcludecoveragefor high-useservicesor for services
that, when covered,would attract enrollmentby high-use
patients. In the United States mental health care is one
exampleof such a service;most private insurance plans
strictlylimitcoveragefor mentalhealth careor carerelated
to substanceabuse.
Verylow-costhealthcare services
Similarly,verysmallhealthcare expendituresmaybe unin-
surable,whethertheyare likelyor not. Forverysmalllosses,
the administrativecostsof insurancemayexceedconsumers'
demandto be protectedfromthe associatedrisk.This does
not meanthat privateinsurancewouldnot coversuchexpen-
ditures,but it probablywouldnot coveronly suchexpendi-
tures. Instead, insurerswould package coveragefor very
smallexpenditureswithcoveragefor morecostly,lesslikely,
and thereforeinsurableservices(suchas hospitalizations).
Uninsurable
individuals
orgroups
Finally,health care that is insurablefor some people may
be uninsurableforothers. Specifically,insurersare likelyto
viewpeopleas uninsurableif they are likelyto need exten-
siveand costlyhealthcare.Thisis the main reasonthat pri-
vate insurance(whenit isvoluntary)does not financemost
health care spending,evenin countrieswith a well-devel-
PRIVATE INSURANCE: PRINCIPLES AND PRACTICE
81
oped insurancesector.If regulationpermits,insurerswill
shunpeoplewith chronichealthproblems,peoplewho are
terminallyill, or peoplelivingor workingin circumstances
that suggesta high risk of illnessor injury.Even if private
healthinsuranceis availableto suchpeople,it maybe unaf-
fordable.2 Althoughaffordableprivatehealthinsurancemay
emerge for relativelyhigh-riskpopulations (such as the
elderly),it is likelyto be availableonlyto supplementexten-
sivecoveragefrom a publicinsuranceprogram.
Dynamics of Private Insurance Systems
In private,voluntaryhealth insurancesystems,people can
choose whetherto buy health insurance.In a competitive
systemtheycan alsochoosewhichhealthinsuranceplan to
buy.In many countriessome workers "buy" health insur-
ance through their employers,taking insurancein lieu of
higherwages.In this casethe employeris the directbuyer
of the healthinsuranceplan for a group.Alternatively, con-
sumersmay buy health insurancedirectly,either as indi-
vidualsor asafamily-much astheywouldbuyanyproduct.
Aswith most products,buyerswilltend to choosean insur-
anceplan that has a lowerprice if its essentialfeaturesare
acceptable.
Insurers can lowerthe price of a health insurancecon-
tract in four ways:
* Bytryingto insure onlylow-riskpeople,denyingcover-
age to people who are sick, or excludingcoveragefor
some conditions.
* Byofferinglesscoverage,limitingthe scopeor extentof
coveredservices.
* By discouragingexcessiveuse of covered health care
services.
* By reducingthe administrativecostsof the plan.
Eachofthese methodscancreateimmenseproblemsfor
some consumers.Consumersmay be unable to buy ade-
quate insurance(or any insurance), especiallyif they are
sick, and they may find that customerserviceunder their
plan (for example,timelyand accuratepaymentof claims)
is poor.But each methodoffers an economicadvantageto
consumerswho are healthy.Because healthy consumers
are unlikelyto need much health care,they are unlikelyto
use theirhealthplanextensivelyifat all.Thusprivateinsur-
ance canofferthem relativelylow-costfinancialprotection.
Insuring
low-riskpeople:insurance andpricing
underwriting in differentclasses(ortiers) of risk.The practiceof pricing
insurancebased on enrolleehealth status (or variousindi-
Consumers seeking insurance are alwaysmore knowl- cators of medicalrisk) is calledtieredrating.Tieredrating
edgeablethan insurersabout their healthstatus and about is a natural,stableresult of competitiveinsurancemarkets:
3
the likelihoodthat they willneed health care. Moreover, tieredratessimplyreflectdifferencesbetweenthe riskscon-
consumerswhohaveor anticipatehealthproblemsare more tained in different insurancepools.8 However,critics of
likelyto seek insurancethan are healthyconsumers.4These tiered rating viewit as splittingup risk unnecessarily. They
factsdictatea greatdealof howinsurancecontractsare sold. argue that tiered rating makes health insuranceunafford-
Enrollmentby people with greater health care needs than able to people who have health problemsor evento peo-
the insureranticipatedwhensettingthe priceof insurance ple who are in a demographicgroup that might suggest
is calledadverseselection.Adverseselectioncan destabilize higher medical expenses. They argue that a single rate
an insurancepool5 and evencause it to fail.6 Thus insurers class, severalrate classesreflectingbroad geographicdif-
havedevelopedtechniquesto avoidor reduceadverseselec- ferencesin the costof care(purecommunityrating),or broad
tion. Possiblythe most importantof these is underwriting. rate classesbased on demographicfactorsbut not reflect-
Insuranceunderwritingis the practiceofevaluatingindi- ing individualdifferences.in health status (modifiedcom-
vidual health status and either rejectingpotential buyers munityrating)wouldmakehealthinsurancemore affordable
who are deemed to pose excessivelyhigh risk or placing to high-riskpeople by forcingother members of the pool
them in plans with other people who represent approxi- to subsidizethem. But since low-riskconsumerstend to
matelythe same risk.Insurersare inclinedto underwritein prefer low-costinsuranceproducts over subsidizingpeo-
order to avoid adverse selection,but competitiveinsur- ple with predictablyhigher health care costs,community
ancemarketsalsotendtorewardinsuranceplansthat exclude rating does not naturally occur in insurance markets.
or isolatepeople with extensivehealth care needs.That is, (Underwritingand communityratingare discussedfurther
in a competitivemarket consumerssearch for the lowest- in a later sectionon insuranceregulation.)
pricedplanthat providesthemwiththe coveragetheywant. Althoughsome insurersre-underwriteenrolleesat the
Insuranceplans that are able to excludehigh-riskpartici- time of renewal,they are more likelyto rely on a pricing
pants are likelyto be less costlyand more comprehensive strategywhich assumesthat customerswho are renewing
than plans that insureeveryoneand try to controlcost in coverageare likelyto havemore (and more costly)health
other ways.Thusconsumerswhoare searchingforthe low- careneedsthan new customersin the sameplan.The prac-
est-pricedinsuranceplan are likelyto prefer an insurance tice of chargingmore for renewalthan for first-issuecov-
plan that excludes people who are more costlythan they eragein the sameplan is calleddurationalrating.
are.7 Durationalrating assumesthat the claimsexperienceof
Insurersthat underwritecoveragetypicallyrequireappli- anyrisk pool willworsenover time-a phenomenonthat
cants to disclosetheirmedicalhistory(allowingthe insurer in fact is usual in insurancepools. Someparticipantswho
to reviewtheir medicalrecords)and mayrequirethat appli- were healthy at the start of the contract become sick or
cantsundergoa physicalexaminationbyan approvedphysi- injured.Womenbecome pregnant and require maternity
cian.Theymayrequireapplicantsto presentsuch"evidence and obstetriccare. Participantswho were unfamiliarwith
of insurability"at the time the contract is firstissued, and the provisionsof their health insuranceplan at the start of
againeachtime the contractis renewed.Thus,whileinsur- the contract (and thereforehesitant to seek care) begin to
anceunderwritingenablesinsurersto priceinsuranceplans leam what services are covered and what they must pay
more accurately,many consumersfind it to be personally out of pocket.In plansthat restrict choiceof provider,par-
intrusiveand offensive. ticipantsmust selecta doctorwithinthe constraintsof the
Insurers that are able to identify high-riskconsumers plan; theywillhesitate to use care for minor health prob-
mayneverthelessbe willingto sellcoverageto manyof them lemsiftheyhavenot yet taken the time to make theirselec-
if they are able to priceinsurancedifferentlyto consumers tion. For all these reasons,insuranceclaimsrise gradually

INNOVATIONS
IN HEALTHCAREFINANCING
82
over the courseof an insurancecontract.Thus insurersare
inclinedto offer a lowpriceto attract new participants,but
they willraise the price at renewalto reflectthe growthin
averagemedicallossesas the insurancepool ages.
Aswithtieredrating,durationalratinghas its critics,who
arguethat it is evidenceof a noncompetitiveinsurancemar-
ket. Theylikenit to sirnplepricediscrimination,notingthat
insurersraisepriceswhen peoplebecomesickand (inmar-
kets whereinsurersunderwrite)when no other insurerwill
selltheman insuranceplan.Moreover,theyarguethat insur-
ers use durationalrating to "chum" their business:by rais-
ing the price of insurance at renewal, durational rating
encouragespeopleto shop fornew coverageand to change
insurersfrequently.Whentheychangeinsurers,theyareunder-
writtenagainasnewbusiness.Peoplewhohavehealthprob-
lemsmaybedeniedcoverage altogether,ortheymaybedenied
coverageforthe caretheyare mostlikelyto need. Peoplewho
arehealthyare able to find newinsuranceat a lowerprice.
In eithercasedurationalratingwillencourageconsumers
to sort themselvesinto differentinsuranceplans,separat-
ing high-riskconsumersfrom low-riskconsumers.Byiso-
latingpeople into relativelyhomogeneousrisk pools, each
of these practices-underwriting, tiered rating, and dura-
tionalrating-tends to reduce cross-subsidiesamongpeo-
ple who are insured. In many countries this outcome is
sociallyunacceptable;and the more perfectlyinsurersare
able to achieve this result, the more unacceptable it is.
Nevertheless,it is an economicallyefficientresult: it min-
imizesinvoluntarytransfersamongindividuals.Regulating
insurersto "correct"this result (that is, to force more het-
erogeneousrisk pools) increasesexcessburden by artifi-
ciallyraisingprices to low-riskconsumers.
In addition to underwritingand pricing,insurershave
developedtwoother techniquesto reduceorto avoidadverse
selectionin theirhealthinsuranceplans:marketingto groups
rather than to individualsand excludingcoveragefor pre-
existingconditions.
Groupcoverage.To limit the amount of adverse selec-
tion that can occur in a healthinsurancepool, manyinsur-
ers prefer to insure people who havegrouped themselves
for reasons other than the purchase of insurance. Such
groupstypicallyincludeemployeesof aparticularfirm,mem-
bers of a professionalor trade association,or evenmem-
PRIVATEINSURANCE: PRINCIPLES AND PRACTICE
83
bers of a social organization.Many insurers require that
the sponsoringorganizationpaya significantshare (at least
half)of the cost of coveragefor group membersto ensure
that eventhe lowest-riskgroupmemberwouldfindenrolling
advantageous.Also,insurersmayrequire that a minimum
percentageof the group (for example,80 percent) enroll
in the plan regardlessof the sponsoringorganization'scon-
tribution, further reducingthe chance that adverse selec-
tion willoccurwithinthe group.
Group underwritingreducesthe insurer'sneed to bear
the cost of careful,individualunderwriting.Instead, the
insurer can look at the broad demographicsand circum-
stancesof the group and decidewhether it represents an
insurable risk. However, in highlycompetitive markets
(where competitionhas driven insurersto set prices very
low)insurersmayattemptto underwritewithinthe group-
denyingcoverageto some group membersbased on their
health status. When within-groupunderwritingoccurs, it
typicallyis in insurancemarketsfor small-groupcoverage.
For reasonsrelated to howlarge-groupcoverageis negoti-
ated and priced,it is rare in the large-groupmarket.
Preexistingconditionexclusions.To deter people from
seekinginsuranceafter they become sick, insurancecon-
tractstypicallyexcludecoveragefor conditionsthat existed
(or that could have been knownto exist)when the insur-
ance contract started. Preexistingcondition clausestypi-
cally stipulate a "look back" period to deem medical
conditionsas preexisting(for example,medicalconditions
that were manifestor could havebeen knownto existsix
monthsbeforethe start of the insurancecontract).For such
conditionsthe contract will stipulatean exclusionperiod
(forexample,sixto twelvemonthsinto the contractperiod)
duringwhichanycare relatedto a preexistingconditionis
uninsured,but care related to other conditionsis insured.
Preexistingcondition exdusions are particularlyprob-
lematicwhen participantswith ongoinghealth problems
try to changeinsuranceplans, or whenthey loseand try to
regaincoverage.Whenhealthinsuranceis providedthrough
an employer,workersmay changeinsuranceplans or lose
coveragealtogetherwhentheychangejobs.Butthe propen-
sityof consumersto buy insuranceonlywhentheyare sure
to needhealthcareissogreat-and the potentialfor adverse
selectionto destroy a health insurancepool is so signifi-
cant-that in markets like the United States insurersare
almostneverwillingto writean insurancecontractthat does
not excludeor delaycoveragefor preexistingconditions.
Limitingthe scope or extent ofcoveredservices
Insurersmaybe willingto insurerelativelyhigh-riskpeople
if they can deny coveragefor specifichigh-costillnessesor
procedures.Coverageexclusionsforcare relatedto specific
high-costillnesses(for example,hemophiliaor diabetes)
or for specificprocedures(for example,organtransplants)
reducethe amountof insuranceavailableto someor allbuy-
ers in the market.These exclusionsdifferfrom preexisting
conditionexclusionsin that they applyequallyto all buy-
ers of that particularplan,regardlessof health care status
at the beginningof the insurancecontract.Moreover,such
exclusionsare permanent:theyexistforthe life of the con-
tract.9 U.S. insurershaveused coverageexclusionsfor ser-
vices such as obstetric care to develop low-pricedplans
that predictablyattract low-riskbuyers;in this case a dis-
proportionatelylargenumber of youngmen.
excessiveuse of coveredservices
Discouraging
Consumerstend to use morehealthcare serviceswhenthey
are insuredthan whentheyare not, evenwhentheir health
status does not change.This tendencyis calledmoralhaz-
ard,a term that at one time implieda judgmentabout con-
sumerdishonestybut no longercarriesthat connotation.
Moralhazard partly reflectsefficientconsumerbehav-
ior: people use more health care when insurancereduces
the price that they must pay for each service.To a fully
insured consumer,the cost of using an additionalhealth
care serviceis zero;thus they are verylikelyto use it.10 But
sinceconsumerstypicallyrelyon the medicaljudgementof
healthcareproviders,muchmoralhazardbehaviorreflects
the propensityof providersto delivermore serviceswhen
insurancemakesthem affordablefortheir patients."I Most
insuranceplans incorporatecost-sharingfeaturesthat are
designedto deter excessiveuse of medicalcare by making
consumerspaysome of the cost of insuredhealth care.
In someindustrialcountries(andespeciallyin the United
States)managedcare has emergedin part to controlmoral
hazard-as it relatesboth to consumers'propensityto use
INNOVATIONS IN HEALTIH CARE FINANCING
84
more care and to providers'inclinationto offer more care
whenit is insured.12 Managedcare planstypicallyestablish
networksof primarycare physiciansfrom whichenrollees
choose,as well as a networkof specialistsunder contract.
The plans coverexpendituresfor specialtycare onlyif the
primarycare physicianrefers the enrolleeto a specialistin
the network.Managedcare planstypicallyattemptto estab-
lisha lower-cost"culture"of practiceamongtheir primary
carephysicians,emphasizinggreateruse of preventivecare
and less aggressivetreatment for some conditions.'3 Thus,
whilethe featuresdescribedbelowarewidelyusedby finan-
cial insuranceplans, managedcare plans use them much
less often and much less extensively
Deductibles.An insuranceplan deductiblerequiresthe
insuredconsumerto payallchargesforcoveredservicesout
of pocketuntilthe total costreachesthe deductibleamount.
After that, the insurance plan begins to pay Insurance
plansalsomayimposeseparatedeductiblesfor specifictypes
of services(forexample,hospitalcare orprescriptiondrugs),
eitherto deterunnecessaryuse or to avoidthe costofadmin-
isteringvery smallclaims.
Toadministera deductible,the insuranceplanmustkeep
an accountingof each enrollee'sexpendituresfor covered
services(and adjudicatewhichexpendituresqualifyforreim-
bursementfrom the plan), eventhough the plan does not
issuepayment.Althoughthisprocesshas been greatlyfacil-
itated by the growinguse and sophisticationof computer
technology,deductiblesneverthelessrequirecarefulrecord-
keepingand, therefore, administrativecost for the plan.
Thiscostis worthwhileifthe deductibleeffectivelydeters
unnecessaryuse. However,a number of studieshavefound
that patientsmaynot distinguishappropriately betweennec-
essary and unnecessary health care. Instead, they may
respondto deductiblesand other meansof cost sharingby
delayingnecessarycare,potentiallyraisingthe ultimatecost
of careby seekingcaretoo late. Still,in the shortterm (that
is, withinthe durationof an insurancecontract)deductibles
are very effective in reducing the use of covered health
careand, therefore,in reducingplancost (Lohr and others
1986).
Deductibles are rare in managed care plans, in part
becausetheyare administrativelyinfeasibleformost of the
care that the planinsures.That is, becausethe health care
providerswho contract with managedcare plans charac-
teristicallydo not charge a separate fee for each service,
mostplansdo not havean obviouswayto comparethe value
of deliveredcare againsta deductible.Also,the conceptof
a significantdeductibleis incompatiblewith managedcare
plans'emphasison preventivecareand low-costcarewithin
a networkof primarycare providers.
Coinsurance. Coinsuranceis the share of costs for cov-
ered servicesthat the consumeris required to pay out of
pocket.Typicalcoinsurancelevelsforcoveredservicesrange
from 10 to 20 percent. Like a deductible, coinsuranceis
intended to reduce moral hazard by imposingsome of the
cost of care on the insured consumer. Coinsurancemay
apply uniformlyto all coveredservices,or onlyto some-
forexample,onlyto inpatienthospitalcare.Separate,higher
coinsuranceamountsmayapplyto types of care for which
underwritingis particularlydifficultand high use can pose
a seriouscost problem for the plan-for example,outpa-
tient mental health care.
Insurancecontractsthat use coinsurancetypicallyalso
set an out-of-pocket limit above which the coinsurance
amount drops to zero. For example, consumers may be
required to pay a $10 deductible, plus 20 percent of any
expenditure for covered services, until the sum of these
out-of-pocket payments reaches a specified amount. At
that point the insurance plan will pay 100 percent of cov-
ered charges up to the plan's external limit, if any. The
enrollee's cost of using health care in such a plan is depicted
in figure 1.
As with deductibles,coinsuranceprovisionscreate admin-
. .. . ,,
istrative expense: they require insurers to keep a careful
accounting of accumulating health care expenditures for
covered services,and to adjudicate each claim.For the same
reasons that they are unlikely to impose deductibles, man-
aged care organizations are unlikely to use coinsurance to
deter moral hazard.
Copayments.A copayment is a small fee that an insur-
ance plan requires patients to pay each time they obtain a
covered health care service. Like coinsurance, copayments
are intended to encourage patients to exercise judgment
about their need for care. Copayments typically apply tO
specific services that are discretionary in nature (such as
PRIVATEINSURANCE:PRINCIPLESANDPRACTICE
85
physicaltherapyor a secondroutine examinationduringa
calendar year). Unlike coinsurance,copaymentsimpose
no significantadministrativecost; they require only basic
record-keepingby healthcare providers(whotypicallycol-
lect the copayment).Because copaymentsare administra-
tivelysimple,managed care plans are more likelyto use
them than anyother form of cost sharingto managecon-
sumer demand.
Internalandexternallimitsoncoverage.Most groupinsur-
ance plans and allindividualinsuranceplans place a limit
on the cumulativeamount that the plan will pay for cov-
ered services.An extemallimit on coverageis a limit on
the total valueof coverageunder the plan,usuallyover the
fullperiodof the contract(forexample,total planpayments
per year).An internallimiton coverageis alimitonthe total
value of coveragefor a particular servicecoveredby the
plan.Commonly,insurerswillplace internalplan limitson
coveragefor mentalhealth care, but may alsolimitcover-
age for inpatient care. Internal limits may be denominated
as units of currency,or as a maximumnumberof inpatient
days or outpatient visits. Many plans place one or more
internal limits on coverage as well as an external limit.
Plan limits serve a number of purposes for insurers.
Internal limits may deter people who anticipate needing
substantial amounts of a particular type of care from seek-
FIGURE
I
Out-of-pocket health care expendituresfor an
uninsured patientandfor an insuredpatient
in a conventional
healthinsurance plan
~~~~~~~~~Patient
costof care
/
for uninsured
patient
/ Out-of-~pocket
costfor:
/ msured~i
patfient,
netof,
i
applies
to Deductible Out-of-pocket
~~~~~ ~ Extera
~~limitplan
rna,drnum
Useof care
ing healthinsurance,since the planwouldnot coverallthe
care that theywould need. Thus insurersare able to limit
the costof failingto underwriteaccurately.Plan limitsalso
limitthe amountof moralhazardthat the planwillsustain.
With respect to mentalhealth care services,moral hazard
is the most significantreasonthat insuranceplans typically
placean internallimiton coverage.Althoughsuch services
maybecurative,whethermentalhealthis restoredis largely
a subjectivejudgement by the patient or the provider-
neither of whomis directlyresponsiblefor payingmost or
anyof the costof care.Finally,planlimitsreducethe amount
and cost of reinsurancethat an insurer needs, either as a
matter of prudent business practice or to complywith
regulation.
Plan limits have very different effects on consumers
and healthcare providers.Limitson coveragemayleavea
significant"tail"of uninsured expenditures,exposingcon-
sumersto financialrisk and the risk of needingto termi-
nate care because they are unable to pay. For providers,
professionalethics and concern for the patient may con-
flict with the plan terminatingpaymentfor care. Even in
cases where plan limits are very high (for example, in
employergroup insuranceplansin the United Statesmost
externalplan limitsare $1,000,000or more), patientswho
reach these limitsmaybe in-hospital,and the hospitalmay
be ethicallyunable to terminatecare.Whilethese casesare
rare,when theyoccurthe hospitalgenerallycontinuescare
without payment.
Reducing
administrative
costs
Whilethe practicesandfeaturesofinsuranceplansdescribed
above effectivein avoidingor reducingadverse selection
and moral hazard, some can raise a plan's administrative
costs.For example,while careful underwritingcan be an
effectivewayto avoidadverseselection,it canalsoadd sig-
nificantadministrativecost.As a resultinsurershavedevel-
oped a number of practicesthat are effectivealternatives
to carefulunderwriting.
Possiblythe most importantof these practicesis group
underwriting-that is, acceptingor rejectingentiregroups
insteadof screeningindividualhealthstatus.Moreover,writ-
ing coveragefor largegroupsmay offer other administra-
tive-costadvantagesfor insurers.For example,the cost of
INNOVATIONS IN HEALTII
86
marketingto a few large groupsof enrolleescan be much
lowerthan the cost of marketingto many smallgroupsor
individuals.Finally,client-initiatedturnoverin the insurer's
business(and thus the administrativecost of the contract)
can be less.Employeeturnoverin largefirmsis lowerthan
in smallfirms (so fewerworkers enter and leavethe plan
duringthe contract period), and large firmsare less likely
to go out of businessthan are smallfirms.14
The economicadvantagesto insurersof writinggroup
coveragegenerallydiminishwith the size of the group.15
For very smallgroups the administrativecost of a group
insuranceplanapproachesthat forindividualcoverage.U.S.
insurersreport that the marketingand administrativecosts
of individualand small-groupcoveragecan makecoverage
40 percent more costlyper enrolleethan the same cover-
age for a largegroup.Moreover,becausemanyinsurersdo
not want to undertake the considerablecost of aggressive
underwriting,some of the largestinsurers (thosethat can
achieve significanteconomiesof scale for administrative
costs such as claimsprocessing)willnot coverindividuals
or smallgroups.The absenceof largeinsurersmayfurther
explainthe high administrativecostsin these markets.
The greater administrativecost of insurance for indi-
vidualsandsmallgroupsin the UnitedStatescan makecov-
erageextremelycostlyand, formany,unaffordable.Lacking
anylegalrequirementthat everyonebe insured,manyindi-
vidualsand smallgroupswithout accessto group coverage
are uninsured.This experienceis not universal,however.
In the Czech Republicquasi-privateinsuranceplans that
compete with the large, central government plan enroll
groups and individualsalike, and find feasiblethe statu-
tory7 percentlimiton the plans'marginsovermedicallosses.
In Australia,where private insurancesupplementspublic
coveragebut coversonly inpatient care, private insurers
are not particularlyconcernedabout adverseselectionor
the high administrativecostof marketingto individuals,and
are not inclinedto foster a group market.
To reduce the cost of underwritingcoveragefor groups
or individuals,insurersmay practicea form of underwrit-
ingcommonlycalledredlining:denyingcoverageto broad
classesof groupsor individualswithout actuallyconsider-
ingtheirinsurability.
Forexample,ifinsurersdeemresidents
of aparticulargeographicarea-such asa low-incomeurban
area-as high risk, they might routinelydeny coverageto
CARE FINANCING
anyonewho livesor worksin that area. Insurersmightalso
redline workers in particularindustriesor occupationsif
they are likelyto be exposedto hazardousmaterialsor suf-
fer a job-relatedinjury.'6
Finally,variousplan designfeaturesthat curb the high
use of health care under the plan may alsoyieldadminis-
trative efficiencies.For example,a plan deductibleallows
the insurerto avoidfullyprocessingandpayingsmallclaims.
The net effecton administrativecostsis likelyto be small,
however,since financialinsurance plans must adjudicate
evensmallclaimsin orderto accumulateclaimsagainstthe
deductible.
Managedcare plansthat pay their primarycare doctors
a salaryor a fixed fee per patient (instead of fee-for-ser-
vice)can vastlyreduce the amount of paperworkinvolved
in sendingandpayingbils andthus can substantially reduce
administrativecosts.But becausewell-runmanagedcare
plans require active management of a large network of
healthcare providers,the net costadvantageof these plans
typicallyrelatesto lowerratesof hospitaladmissionamong
managed care participants, not to lower administrative
costs.17
GovernmentRegulationof Insurers
Governmentregulationofinsurancetypicallyhas threegoals:
maintaininga stableinsurancemarket,protectingconsumers,
andmaximizingconsumerparticipationin the privatemar-
ket. Variouspractices related to each of these goals are
describedbelow.
Stabilizing
the insurance
market:standards
forinsurer
entryandexit
Regulationto stabilizeinsurancemarketsincludessetting
financial standards for market entry and ongoingopera-
tions, ethical standards for market entry, and conditions
for insurer exit from the market. Entry or exit standards
that are set too high willallowfewerinsurersinto the mar-
ket and reduce competitionamonginsurersthat do enter.
In turn, insurancepricesare likelyto be higherin a lesscom-
petitive market, and buyers will have less choiceamong
insuranceproducts. But standardsthat are set too lowcan
allowfinanciallyunsound or unscrupulousinsurers into
PRIVATE INSURANCE: PRINCIPLES AND PRACTICE
87
the market. Such insurers prey on unsophisticatedcon-
sumerswho find manyaspectsof insurancecontractsdif-
ficultto understand.
Financialstandardsforentryandoperation.The mosteffec-
tive and efficientwayto minimizethe chancethat people
willbuy insurancefrom financiallyinsecure insurersis to
bar such insurersfrom enteringthe market. Thus govern-
ments mayrequire that insurersbe licensed.1 8 As a condi-
tion of receiving a license, insurers maybe required to
meet minimumstandards for financialsoundnessand to
demonstratepast or intendedethicalbusinesspractices.
As evidenceof financialsoundness,governmentsmay
requirethat an insurermeet minimumcapitaland surplus
requirements.Acapitalrequirementestablishesa minimum
levelof financialassets for insurersseekingto enter (and
remainin)the market.Governmentsmayrequirethat insur-
ers hold these assetsin highlysecureinvestments.19
In addition,governmentsthat regulateprivateinsurance
typicalyestablisha surplusrequirement:a minimumlevelof
financialassetsthat an insurermust hold relativeto its esti-
matedliabilities(principaly,
itsestimatedmedicallosses).Both
capitaland surplusrequirementsmaybe set bylaw,andthey
mayvary by class of business.For example,insurersthat
writeboth personalhealthinsuranceand personallifeinsur-
ancemaybe requiredto meetseparateminimumcapitaland
surplusrequirementfor eachclassof business.20,21Thegov-
ernmentalsomayset the surplusstandardhigherforinsur-
ers that write a larger amount of coverage.By scalingthe
surplusstandardto the insurer'svolumeof business,the gov-
ernmentcanencourageinsurersto enterthe marketandreduce
the likelihoodof a major marketdisruptionfrom the insol-
vencyof a large insurer.Finally,governmentsmayrequire
that insurersenteringthe market bring a higheramountof
initialsurplus,whichthey maythen deplete (so long as it
exceedstheminimumsurplusthattheymustholdon anongo-
ingbasis)in financingtheirinitialoperations.
Successfulsolvencyregulation entailsintensive peri-
odicreviewofeachinsurer'sfiscalcondition.Typically, insur-
anceregulatorsare givenbroad statutoryauthorityto audit
and investigateinsurancecompanies.U.S. insuranceregu-
lators even have statutory authorityto assume control of
domesticinsurersin financialdifficulty,with the intent of
"rehabilitating"the insurer-modifying the insurer'sman-
agementand financialpracticesto improveits prospectsfor
remainingsolvent. 22
In developingcountriesestablishingsolvencystandards
forinsurersentailscarefulreviewand standardizationinsur-
ers' accountingand actuarialpractices.If regulatorystan-
dardsfor capitalandreservesare to be meaningful,insurers
mustuse a conmmon accountingconventionto measurethe
valueof assets and liabilities.The governmentmust estab-
lishthe basisfor evaluatingassetsand healthinsurancelia-
bilities,and require all licensedinsurersto report on that
basis.Therisk of financialdistortions(and the opportunity
for unsound financial practice) by insurers rises when
accountingprinciplesareundear andindustrypractice varies.
However,evenwithcommonaccountingpractices,antic-
ipatinghealthinsuranceliabilitiesin developingcountries
canbe extremelydifficult.The introductionof privateinsur-
ance may cause health care prices and real spendingfor
health care to acceleratesharply.Thus the integrityof the
plan's actuarialestimatesrely fundamentallyon how well
the plan is able to control the cost of covered services.
Such controlsare generallymore reliablein managedcare
plans,especiallyif theypayprovidersa fixedfee per patient
(calledcapitation)for most or allcare coveredby the plan.
Because plans without meaningful expenditure con-
trols may be unable to anticipateliabilitywith sufficient
accuracy,regulatorsmust carefullyconsiderwhether to let
them enter the market. If the insurer'spremiumlevelsare
likelyto be insufficient,it is at great risk of insolvency-
evenwhenit holdsthe requiredreserves.Especiallyin newly
developingmarkets,regulatorsshouldreviewthe adequacy
of insurancepremiumsas vigilantlyas theyreviewcompli-
ance with financialstandards.
Governmentscan also prohibitspecificorganizational
forms of insurance that are deemed financiallyunstable
(eventhough they maypurport to broaden enrollmentin
private insurance plans). These may include multiple-
employertrusts, fraternal organizations,or membership
associations.Since not all regulatorsagreeon whichorga-
nizational forms are inherently unstable, some jurisdic-
tions may alloworganizationalforms that othersprohibit.
In the UnitedStatesand manyother industrialcountries
insurersare prohibitedfrom directlyengagingin anybusi-
ness that is not reasonablyrelated to insurance.For exam-
ple, insurersmaybe prohibitedfrom affiliatingwithbanks,
INNOVATIONS IN HEALTH CARE FINANCING
88
based on concernsthat the insolvencyof one could con-
tributeto the insolvencyof the other.Butinsurersmayform
a holdingcompanyto affiliatewithan unrelatedenterprise. 23
Still,anyaffiliatedenterprisecan destabilizean insurer,and
insolvencyproblemshaveresultedfrom holding company
affiliations(U.S.House ofRepresentatives1990and 1994).
Especiallyin countrieswith a newlyemergingprivateinsur-
ancesector,governmentsshouldconsiderregulatorysafe-
guardsin order to prevent such situations.24
Finally,manycountrieshavesome arrangementto guar-
antee insurancebenefits if an insurer becomes insolvent.
In some countriesthe governmentoperates the guaranty
system;in others a government-authorizedprivateagency
orassociationperformsthat role,but allinsurersare required
25 Government-
to participateas a condition of licensing.
authorizedprivateguarantyfunds maybe financedby an
assessmenton allhealthinsurers(usuallyat the time of the
insolvency).These assessmentsmaybe limitedby law (for
example,to 2 percent of grosspremiumvolumeper year).
Government-runguaranteefundsalsomaybe financedfrom
salestaxes on insurancepremiums.If the guarantyliability
exceedswhat is collectiblewithin the annual assessment
limit, participatinginsurers maybe reassessedeach year.
Thusguarantee-fundpaymentsto policyholders maystretch
over a number of yearsin the case of a largeinsurerinsol-
vency.Insolvencyguaranteefunds may pay policyholders
onlya portionof each claim(for example,60 percent),and
may leave policyholdersresponsiblefor much of the cost
of care that would have been paid by the insurer had it
remainedsolvent.
Nonfinancialstandardsfor entryandoperation.Regulation
governing the ethical practice or intent of insurers may
take anyof severalforms. Somegovernmentsrequire that
officers,board members,andincorporatorsmeet residency
or citizenshiprequirements.Some countriesrequire that
certain minimumownershiprights be held by nationals.
Governmentsalsomayinvestigatethe experienceor char-
26 They may require
acter of individualsseekinglicensing.
that insurersdomiciledin another jurisdictionor country
demonstratethat they are lawfullyorganizedand licensed
in their home jurisdiction.They mayrequire that insurers
seekinga licensesubmita proposedbusinessoperationplan,
assessingthe economicsoundnessof that plan as part of
the processof admittingthe insurerintothe market.Finally,
states may prohibit some organizationalforms of health
plans if their practicesare deemed too difficultto monitor
andregulate.Forexample,at leastoneU.S.state(Minnesota)
will not licensemanaged care plans that are operated as
for-profitenterprises.
Conditionsfor insurerexit.Constraintson marketexitcan
be as importantas restrictionson market entry.Allowing
insurersto haphazardlyenterandleavethe marketincreases
market instabilityand erodesconsumerconfidencein pri-
vate insurancesystems.Ideally,insuranceregulatorsshould
examinean insurer'scommitmentto stayingin the market
as a conditionof admittingthem to the market.But such
an appraisalcan be difficult,and regulatorsmayinsteadrely
on establishingconditionsfor exit.
Exitrulesforinsurersmaytake severalforms,allofwhich
are designedto minimizemarket disruption as a result of
insurer exit.At a minimum,departinginsurers should be
required to givereasonablenoticeto policyholdersand to
submita plan demonstratinghowclaimsand other obliga-
tions willbe satisfied.They alsomaybe required to pay a
processingfee to the regulatoryagencyto coverthe costof
overseeingan orderlyexit (Skipper1992).
Reporting requirements,examinations,andlprofessional
over-
sight.Reportingrequirementsconstitutethe coreof insurer
surveillance. Governmentsshouldrequirealllicensedinsur-
ersto submitfullfinancialreportsannually,andmayrequire
abbreviatedfinancialstatementsquarterly.At a minimum,
these reports shouldallowthe regulatoryagencyto under-
stand whetherthe insureris operatingwithinthe financial
limitsdictatedby law or by prudent businesspractice.
In mostdevelopedinsurancemarketsgovernmentscon-
duct periodic onsite examinations of insurers' financial
records.For example,allU.S. states requireonsite exami-
nation of domestic insurers, typicallyevery three to five
years.Regulatorsalso mightconduct a targeted examina-
tion if they suspectthat an insurer is in financialdifficulty
(Skipper1992).
Governmentsalsomay enlistthe accountingand actu-
arialprofessionsto discourageinappropriateinsurerbehav-
ior and to reveal it if it occurs. For example,insurance
regulatorsmayrequire insurersto submit an independent
PRIVATEINSURANCE: PRINCIPLES AND PRACTICE
89
actuarialopinion certifyingthe adequacyof their reserves
relativeto the nature of the risk they have assumed. (All
U.S. states now require that licensedinsurers submit an
actuarialopinion each year.)Regulatorsalso may require
that insurers submit to a periodic financial audit by an
independentcertifiedpublicaccountant,andthat the results
2 7 The costsof these professional
be disclosed. certifications
shouldbe paid by the insurer.
Consumerprotection
Consumerprotection regulationtypicallyis of two types:
regulationthat governsthe languageand marketingof insur-
ancecontractsand regulationthat governsthe relationship
between insuranceplans and health care providers.Given
the complexityof evenrelativelysimnple insurancecontracts,
governmentregulationof the languageof insurance con-
tractscanbe extremelyimportant.Governmentscan require
that insurersexplaincoverageusingcommontermsandthat
specificfeaturesof healthinsuranceplansbe explainedin
all insurance contracts. Such features might include the
plan's deductibles, coinsurance provisions, copayment
amounts,out-of-pocketlimits,and internal and external
limitson coverage.Plansthat contractwithprovidergroups
(physiciangroup practicesor specifichospitals,for exam-
ple) and restrict or modifycoverageaccordingto policy-
holders' use of those providersshouldalsobe required to
fullydisclosethe nature and detailsof those limitations.To
enforce the use of clear languageand honest marketing
practices,insurance regulatorsshould require insurersto
submitallmarketingandenrolleematerials,andtheyshould
be preparedto reviewthese materialsin a timelymanner.
In practice,much of the consumerprotectionprovided
by regulatoryagenciesoccurs only after a problem arises
and in response to consumer concerns and complaints.
Prompt andvigorousresponseto consumergrievancescan
be an effectiveand efficientwayto augmentreviewofinsur-
ers' contractlanguageandmarketingmaterials.Thisis espe-
ciallytrue in largeor fast-growinginsurancemarkets,where
the paperworkburden of meticulousdocumentreviewcan
overwhelrna smallregulatoryagency
Regulationthat govemsthe relationshipbetweenhealth
plansandprovidersis sometimesalsoconsideredconsumer
(or patient) protection.Typically, such regulationseeksto
preserve the professional autonomy of health care
providers-especiallyif they participate in managedcare
contracts,which may discourageproviders from recom-
mendinghigh-costtreatmentsor fromreferringpatientsto
specialistsfor care.
Twomain forms of regulationhave emergedaffecting
the relationshipsbetween health care providersand man-
aged care plans. "Anywillingprovider" regulationspro-
hibitinsurersfrom "lockingout" physiciansor other health
careproviderswhoare willingto acceptthe plan's payment
levels,practiceguidelines,and reportingrequirementsfor
participatingproviders.Conversely,antitrustregulationmay
prohibithealthinsuranceplansfrom"lockingin"providers-
that is, requiring that providers who contract with them
not acceptpayment(or patients)from anyother insurance
plan.
Health care providersgenerallyviewregulationof either
type as essential to maintainingtheir professionalinde-
pendence and, therefore, protecting the best interests of
theirpatients.Health insuranceplans,however,viewthese
regulationsas limitingtheir abilityto establisha professional
cultureofconservative(low-cost),high-quality medicalprac-
tice amongproviderswho contractwith them.
There is no simpleresolutionof these competingper-
spectives.But manyanalystsbelievethat the development
of generallyacceptedmeasuresof healthcarequalitywould
reduce the need for such patient protection regulations.
They contendthat qualitymeasureswould allowboth reg-
ulatorsand enrolleesto monitorthe qualityof carein com-
petingmanagedcareplansandin financialhealthinsurance
plans. They believethat health planswouldbegin to com-
pete on quality(aswellas on price)-reducing the need to
regulaterelationshipsbetween healthinsuranceplans and
providersas long as the qualityof healthcare financedby
the plan remainsacceptable. 28
Improving
the fairness
ofprivateinsurance
In the United States a growingbody of state regulationis
emergingto reconcileinsurer practiceswith socialpercep-
tions of fairness.These regulationsgenerallyare of three
types: those prohibitingor restrictingmedicalunderwrit-
ing, those prohibitingor restrictinginsurersfrom setting
pricesbasedon healthstatus,andthose requiringinsurance
INNOVATIONS IN HEALTH CARE FINANCING
90
plans to cover specific types of services or health care
providers.Each typeof regulationis discussedbelow.
Guaranteed issueand renewal.Guaranteedissueregula-
tion requiresinsurersto accept allapplicantsfor coverage,
regardlessof their health status. By contrast,jurisdictionrs
that requireguaranteedrenewal(andbyinference,notguar-
anteed issue) allowinsurersto underwritewhen they first
issuecoverage,but prohibitthem from underwritingwhen
the insurancecontract is renewed.Thirty-sixU.S. states
requireguaranteedissueof someor allinsuranceproducts
in the small-groupmarket;thirteenrequireguaranteedissue
in the individualhealthinsurancemarket.2 9
Guaranteed issue regulation may require insurers to
accept applicantsinto anyhealth insuranceplan that they
offer,or into onlyone or two selectedplans.Byitselfguar-
anteedissueregulationdoesnot addresshowinsurersmay
set prices for these plans.
Communityrating.Sixteen U.S. states require insurers
to use someformof communityrating-to chargeone price
for allplan participantswithinbroad geographicor demo-
graphicgroups,withoutregardto healthstatus.Community
ratingregulationprohibitsinsurersfrom usingthe tiered or
durationalratingtechniquesdescribedearlier.Alternatively,
somestates requireinsurersto pricewithin specified"rate
bands"-that is,theyallowinsurersto pricecoverageaccord-
ingto health status,but limitthe variationin rates around
the medianrate that insurerschargefor the same plan.
Insurancerate regulationis intended to subsidizeplan
participantswho are more likelyto need health care.Thus,
to achievea public policygoal (more affordablecoverage
for peoplewith ongoinghealth problems),rate regulation
createssome economicinefficiencyby forcinginvoluntary
cross-subsidiesamongplan participants.
Experiencesuggeststhat guaranteedissueregulationnei-
ther destabilizeshealthinsurancemarketsnor causesa sig-
nificant increasein the averageprice of health insurance.
However,actuarialresearchindicatesthat it does raisethe
priceof coveragefor some low-riskconsumers(American
Academyof Actuaries1993). In states whereinsurers are
requiredto guaranteeissue and to communityrate health
insuranceproducts, somesmallinsurershave left the mar-
ket,and someinsurershavecomplainedthat plancostshave
risenbecauseyoungerand healthierenrolleeshaveleft their
plans (Chollet and Paul 1994). However, it is unknown
whetherparticipantswhoexitedthese plansbecameunin-
sured or simplychangedplans.
In newlyemergingprivateinsurancemarketsthe transi-
tioneffectsof guaranteedissueandrate regulationmightbe
greaterthaninwell-established mar-
markets.Well-established
kets mayhaveone or two relativelylargeinsurancecompa-
nies that anchorthe market and that absorb much of the
market's high risk. In most U.S. marketsBlue Cross and
BlueShieldplansplaythisrole.Many"Blues"planshaveperi-
odic open enrollment(onceor twicea year,for threeto six
weeks)duringwhichtheyguaranteeissue.(In some states
periodicopen enrollmentis a conditionof the Blues'non-
profit, tax-exemptstatus.) Moreover,the Blueshistorically
havecommunityratedtheir products(again,sometimesas a
conditionof theirnonprofitstatus),althoughmanyno longer
do so.In marketsthat haveoneor twolargeinsurersanchor-
ing the marketand acceptingallapplicants(at leastperiodi-
cally),the largeinsurermayalreadyholdmuchofthe market's
high risk,and smallerinsurersmaydevelopnew marketing
practicesthat in effect preservetheir underwritingadvan-
tage.30 Thusstrictguaranteedissueandcommunityratingreg-
ulationmayhavefewearlyeffectsin well-developed markets.
But in developingmarkets that have only a few small
insurerswritingcoverage,suchregulationmayhavea chill-
ingeffecton marketdevelopment.Health insurersin devel-
opingmarketsare likelyto experienceimmnediate entry of
high-costparticipants,to havea smallerbase overwhichto
spread risk, and to have no ready reinsurancemarket to
financethehighriskthat regulationrequiresthemto accept.31
Raterevieworapproval. Governmentsthatdonotrequire
insurersto communityrate theirhealthinsuranceproducts
maystillundertakerate review.Theymayevenrequirethat
insurersreceivegovernmentapprovalfor their rate levels
andincreases.Whererate reviewand approvalare required,
insuranceregulatorstypicallyexaminethe reasonableness
ofinsurers'lossratios:that is, the ratioof paymentsfor med-
ical care to premium income. In effect, approvalof the
insurer'sratesis approvalof their loss ratio.
Anumberof U.S.stateshaveestablishedtargetlossratios
for healthinsurers;eightstates requireinsurersto meet or
exceeda minimumlossratioin orderto haveanyrateincrease
PRIVATE INSURANCE: PRINCIPLES AND PRACTICE
91
approved.The federalgovernmentalsohas set a targetloss
ratiofor insuranceproductsthat supplementMedicarecov-
erageforthe elderlyandthe disabled,althoughit leavesrate
reviewto the states.
Mandatedbenefits.Healthcareproviderscanbe the great-
est opponentsofinsuranceplansthat limitthe scopeof cov-
ered benefits as a way to offer low prices. Health care
providerswhofavorspecificbenefitmandatesusuallymake
two arguments:that without a mandatethe insurancemar-
ket is evictingpatientswhoneed a specifictypeof care and
who cannot afford to pay for it out of pocket; and that
theirservicesrepresentlower-costalternativeto othertypes
of care. The second argument is usuallyvery difficultto
proveconclusively.
In everyU.S.state varioustypesof healthcare providers
(psychiatristsand psychologists,chiropractors,physician
assistants and nurse practitioners, marriage counselors,
homeopathicmedicalpractitioners,and even faith heal-
ers) have sought and won legislationrequiringthat insur-
ance planscovertheir services.Theirsuccesshas led states
to regulatethe contentofhealthinsuranceplansextensively,
encouragedemployersto self-insure(by federal law,the
states cannot regulate self-insuredplans; Jensen, Cotter,
and Morrisey1995),and increasedthe costof healthinsur-
ance (Jensenand Morrisey1990;Jensen 1993).Most state
legislatorshave come to understand the economicineffi-
ciencythat these regulationscan create, and the rate of
32
new enactmenthas markedlydiminished.
Some mandated benefitsare not so clearlyinefficient,
however.Instead,theymaycorrectmarketfailuresthat arise
from consumers'mnisinformation about the coveragethat
their plan provides.For example,everyU.S. state requires
health insurersto cover newbornsimmediatelyunder the
parents'familyhealthinsuranceplan.Thisrequirementpro-
hibitsinsuranceplans from deeminganycongenitalhealth
problemof the newbomasapreexistingconditionand,there-
fore, denyingcoveragefor care that is often life-savingand
usuallyverycostly.If few parents anticipatethat their pol-
icywould not cover such care when they purchasefamily
insurance,suchregulationmaycorrectforconsumers'imper-
fect information.However,it also mayraise the price of
coverage,andprobablyencouragesinsurersto develophealth
plansthat excludecoverageof obstetriccare altogether.
 Private Insurancein DevelopingCountries
The firstsectionof the paper has providedthe context for
reviewingthe coverageand circumstancesof privateinsur-
ance in low-and middle-incomecountries.Althoughthe
theoryandpracticeof insurancehavebeenhonedin OECD
countries(particularly the UnitedStates),non-OECDcoun-
trieshavedrawnonthat experience,experimentedwithdif-
ferentapproaches,andhad distinctexperienceswithprivate
insurance.This sectionemphasizesthe practiceand expe-
riencethat has evolvedfrom these efforts.
The extentof private insurancein anycountryis related
to a number of factors.Amongthe most importantare per-
sonalincome,maturityof financialmarkets, extent of pri-
vate health care services,culturalfactors,and government
policy.These characteristicsalsodeterminehowinsurance
evolvesand its pace of growth.Annextable 2 summarizes
the characteristics
andcoverageofprivateinsurancein thirty-
fournon-OECDcountries,as wellasin Mexicoand Turkey.
The tableincludescountrieswithrelativelystronginsurance
markets,aswellas somewith emergingmarkets.Countries
with negligiblemarkets, like Bangladeshand Ghana, are
not includeddespiteavailabledata.
In Africaprivate insurancecoverageis highlyvariable,
rangingfrom SouthAfrica(witha well-establishedmarket
and 16percent coverage)to countrieslikeAngola(withno
measurable market). Outside of C6te d'Ivoire, Kenya,
Nigeria, and Zimbabweprivate insurance is negligible,
althoughthe potentialforgrowthexiststhroughoutthe con-
tinent.The Commonwealthcountriesin the regionare the
most likelyto have an establishedprivate insurancemar-
ket. Althoughthese markets have a foothold in Tanzania
and Uganda,theyare imited.Higher-incomecountrieswith
large employershave pioneered and established private
insurance,but small,poor communitiesalso havechosen
to sharerisks.
Privateinsuranceis extensivein Latin Americaand the
Caribbean,where a long traditionof privateprovidersand
payers offers a solid base for expansion. Moreover,the
proliferationof socialinsurancethroughsocialsecurityinsti-
tutes and employer-basedsicknessfunds throughout the
regionhas maderelianceon insuranceto financehealthcare
INNOVATIONS IN HEALTII CARE FINANCING
92
commonpractice.The existenceof mature insurancemar-
kets has helped strengthenthe health insuranceindustry
and support its growth.
Traditionally,much of Asiahas relied on out-of-pocket
expendituresto financehealth care, and even public pro-
gramstend to requirecopayments.Privatehealthinsurance
is negligiblein these countries(see annextable 1).Indeed,
China has no private insurance. The one outlier is the
Republicof Korea,whichhas the most privatelyfinanced
health care system in the world, with 94 percent of the
populationcoveredby compulsoryprivatehealthinsurance.
India, with its 3.3 percent coverageand 1 billionpeople,
has the most individualscovered(33 million).In allthese
systems,copaymentsare attached to the use of most ser-
vices.At the other extreme,Bangladeshand Pakistanare
seeingthe beginningsof an insurance industry,with cur-
rent enrollmentsin the thousands.
Althoughprivate health care is common in the Middle
East,privateinsuranceis justbeing established.Lebanon's
emergencefrom years of civil war make it an exception,
but its private sector-oriented health system will likely
lead to rapid growthin insurance-typefinancing.With an
establishedprivatesector and risingincomes,demand for
privateinsuranceis likelyto increasein other MiddleEastern
countries,whereit alreadyhas a significantfoothold.Egypt
is a prime example(seeRafeh in this volume).
Eastern Europe and Turkeyhavelimited privateinsur-
ance coverage.In Eastern Europe low coveragecan be
ascribedto comprehensivesocialinsurance,and in Turkey
to high inflation,traditionaldistrust of life insurance,and
cultural factors (Fuenzalida-Puelma1996). The region's
potentialforprivateinsuranceis stronggivenrisingincomes
and growthin private providers.Moreover,some Eastern
Europeancountriesperceiveprivateinsuranceas a source
of funds to recapitalizetheir health care deliverysystem.
In somecountriescommunitygroups,ruralcooperatives,
and missionhospitalshave created networksof informal
insurancefunds to financecatastrophiccare or health ser-
vicesformembers.Nonprofitorganizations likethe Grameen
Bank in Bangladesh have expanded into health insur-
ance-type arrangements, and cooperatives and other com-
munity endeavors have led to private arrangements for
financing health care. Government has taken a role in Zaire
and Guinea-Bissau, providing physical infrastructure and
often seed capital to establish community insurance funds
that are usually designed, managed, and operated by com-
munities to pay for health care services for a defined pop-
ulation (TAI 1997; Shaw and Griffin 1995; La Forgia and
Griffin 1993)..
Private health insurance coverage data give an idea of
the extent of health insurance in developing countries. But
the nature of those systems-how they operate, how they
are regulated, and how well they meet the objectives dis-
cussed in the first section of this paper-vary significantly.
Indeed, the orderliness conveyed in the first part of the
paper is not always apparent in these countries. The fol-
lowing sections discuss the characteristics of these coun-
tries' health insurance systems, emphasizing the differences
relative to U.S. approaches; impediments to development
of a private insurance industry in different countries; and
the role of health insurance regulation, and its scope and
limitations, in developing countries.
Health Insurance in Developing Countries
As in OECD countries, private health insurance is rarely
the main source of health care financingindevelopingcoun-
tries. Instead, most private insurance is supplementary (an
outgrowth of social insurance that covers services not
included by the socialinsurance or "primaryinsurance" ben-
efit plan) or additional to services financed under an exist-
ing public system. Characteristics of insurance are very
different from those in industrial countries. Indeed, even
the rationale for insurance in non-OECD countries is dif-
ferent, with basic costs covered but rare, high-cost proce-
dures excluded.
Benefitplons
Private health insurance plans in developing countries typ-
ically exclude costly or chronic diseases and have low ben-
efit ceilings (in terms of number of inpatient days or total
expenditures). In many casesthese limitstranslate into insur-
ance plans that do not cover catastrophic care but cover
PRIVATE INSURANCE: PRINCIPLES AND PRACTICE
93
everything else. In Turkey cancers, heart conditions, and
even high blood pressure are outside the package of most
benefit plans, and in Brazil plans typically exclude infec-
tious diseases, chronic conditions, mental illness, kidney
dialysis, and AIDS (Fuenzalida-Puelma 1996; Lewis and
Medici 1995).
Argentina's sickness funds have highly differentiated
benefit plans because benefits are a function of earnings
(sincethe revenue base is drawn from a percentage of wages).
Thus high-wage industries provide insurance that covers a
full range of surgeries, psychiatric services, cancer treat-
ments, and dental care, while low-wage industry insurance
funds cover onlylimited care, strictlyration access to costly
care, and force members to rely on public facilities (World
Bank and IDB 1997).
Uruguay's sickness funds, the Institucionesde Asistencia
M&dicaColectiva,have a legallymandated benefit package
and exclude services that are covered by the reinsurance
fund (see below on reinsurance). Private insurance is often
purchased for specific types of benefits: emergency, med-
ical, surgical, diagnosis, or hospitalization. There are no
requirements for private insurance benefits.
In Korea expensive high-technology medical services
(CAT scanning,magneticresonanceimaging,PET scan-
ning, some chemotherapy) are excluded from private insur-
ance plans, and hospitalization coverage-is limited to 180
days a year (210 days for the elderly; Yang 1996). In addi-
tion, the qualityand scope of benefits variesby income level.
This pattern is also apparent in Chile's private insurance
companies-higher-income industries receive a more gen-
erous package of services because their contributions are
higher (World Bank forthcoming).
Benefit plans are often regulated by governmentto ensure
a basic level of coverage under private insurance plans (as
in Uruguay for social insurance and much of Europe for
both social and private insurance). In countries where gov-
ernment does not play a role in defining primary insurance
coverage, benefits are less likely to include low-risk, high-
cost events, and high-riskindividuals are less likelyto obtain
private coverage.
In the absence of government requirements on the scope
of benefits, private insurers typically cover only low-risk
groups, leaving government to finance catastrophic care
through supplementary insurance and to provide services
to the chronicallyill as the insurer of last resort.Although
tem but require that citizenshave healthinsurancecover-
market segmentationmaybolster the insuranceindustry age-which is whyboth countries have a robust private
andmayalsomaximizethe numberof peoplewhoare insured insuranceindustry.The costs of this private coverageare
voluntarily,it has socialconsequencesthat many countries borne by consumers, reducing overall government
find unacceptable. Hence the role of health insurance expenditures.
regulation. In Argentinaand Uruguayusers are not satisfiedwith
the cost and qualityof servicesprovided under the sick-
Government policyandhealthinsurance ness funds, which is putting pressure on governmentsto
make changesand spurringthe purchaseof insuranceand
Most governmentshavenot taken a strongpositionregard- servicesfrom the private sector. It is not clear, however,
ing private health insurance.And since insuranceregula- whetherthese developmentsare reducingthe government's
tion is almostnonexistentin most countries,policymakers health care financingburden. Indeed, in Uruguaycostly
arenot alwaysawareofthe extentornature ofprivatefinanc- high-technologyservicesremain the financialresponsibil-
ing.Some policies,however,encourageprivateinsurance. ity of the government.
Amongthe most importantis not havinga policyagainst
privateinsurance,or at least allowingcompaniesto exper- Reinsurance
iment with different ways of financing health care for
employees. Reinsuranceallowsinsurancecompaniesto spreadthe risks
Policiesthat permit people to opt out of socialinsur- that they assume from policyholders.Because spreading
ance (althoughusuallywith some nominalpaymentto the risks allowshealth insurers to include rare events in the
government)are also important, as in Egypt, Singapore, packageof coveredservices,reinsuranceis in effectinsur-
and Uruguay.But optingout mayhavenegativeeffectson ance for insurancecompanies.Reinsurancecan take the
the riskpoolthat remainsundersocialinsurance.Theyoung form of sharing risks among affiliated insurers, sharing
andthe healthyare the most likelyto opt out,leavinglower- high risksamong a large number of participatingcompa-
incomeand lesshealthyindividualsin the socialinsurance nies, andtransferringrisksto anotherinsurer,amongother
systemand raisingthe averagecost of servicesin the pub- options.Somegovernmentsrequiresharingamonglicensed
liclysponsoredsystem.Even wherethere are coststo opt- insurers,but this is rare outsideOECD countries(Mclsaac
ing out-as in Egypt, where 1 percent of earnings must and Babbel 1995).
continueto be paid to the government(Rafehin this vol- The abilityto spreadrisk across enrolleesis often rein-
ume)-these tend to be nominal,and do not discourage forcedbyreinsurancethat ensuresinsurancecompanysol-
people from buyingalternativecoverage. vencyin the event of high, unforeseen risk. Indeed, the
Developing country governmentsoften allow opting amount of risk that an insurer can absorb increaseswith
out in order to sharethe costof healthservicedeliverywith the availabilityofreinsurance;otherwisecompaniestend to
the privatesector,providecitizenswith choiceof payersand be smallandvulnerableto unforeseenhighrisks.Reinsurance
providers,and reduce the overallburden on public provi- coverageis rarein countrieswherecapitalmarketsare weak,
sion. Morerecently,governmentshaveseen privateinsur- and unavailablewhere actuarialdata are poor. In Turkey
anceasthe possiblefinancialbaseforrecapitalizinga decayed reinsurancecompaniesconsider health insurancea high-
healthdeliveryinfrastructure;thisismostrelevantin Eastern risk businessfor many of these reasons.Limited benefit
Europe. Some countrieshave responded to demands for packagesare one symptomof a lack of reinsurance,and
choiceby allowingcompetition among approvedprivate mostdevelopingcountriesimplicitlyrelyon governmentas
providers under a regulated or mandated (social)insur- the insurer of last resort. In Braziland Egypt the govern-
ance system (Brazil,Chile, Korea, Philippines,Uruguay) mentfinanceshigh-technology tertiaryservicesthroughsup-
or bycontractingout specificservicesto privateplans.Chile plementaryinsurance,and coverspeoplewho lose private
and Korearelyextensivelyon optingout of the public sys- coverage.

INNOVATIONS IN HEALTEI CARE FINANCING

94
 Sicknessfunds in Argentinaand Uruguayhave public for limitingor not investingin privateinsurance(TAI1997).
reinsurancesystems-that is, the governmenttaxes and. But in someinstancesthesedifficultieshavehad littleeffect
managesreinsurancefunds. Argentinadoes this through on the growthor profitabilityof private health insurance
the Fund for HighlyComplex[Procedures](Fondode Alta markets.Argentinaand Brazil,forexample,weatheredyears
Complejidad),whichtheoreticallyfinancesrare and costly ofhighinflationaccompanied bygrowthin privateinsurance-
treatmentsbut in realityfinancesoverflowfrom the sick- suggestingother mitigatingfactorsfor investors.
nessfunds (WorldBankandIDB 1997).Uruguay'sNational Governmentpolicytowardthe sectoralsoinhibitsinvest-
Fund for Resources(FondoNacionalde Recursos)is a form ments.Indeed,lackof transparencyregardinglawsand reg-
of socialinsurancethat financessophisticatedmedicalpro- ulation and extensivebureaucratic hurdles have limited
cedures at specializedmedicalcenterslocatedin the sick- investmentin healthinsurancein countriessuchasthe Czech
nessfund facilities.Mandatorycontributionsfrom workers Republic,Egypt,and Indonesia.Inconsistencyin financial
finance, amongother procedures,cardiovascularsurgery, andhealthpolicesdiscouragesprivateactivity,and inTurkey
renal dialysis,and transplants.The fund functions as de has contributedto slowgrowthand limitedproductsin the
facto reinsurancefor the sicknessfunds, whoserisk pools insurancemarket (Fuenzalida-Puelma1996;Rafehin this
are heterogeneous. volume;TAI 1997;WorldBank 1992).
Policiesthat subsidizeor makecapitalaccessibleforinfra-
Impedimentsto Health Insurance structure expansion,often for designatedareas or under-
Development servedgroups(as in Brazil,Colombia,andJordan), have
resultedin a largenumber of private health care facilities
Impediments to the establishment and development of (WorldBank 1994 and 1996). Such facilitiesencourage
privatehealthinsurancestemfrom rigiditiesin the market, developmentof a privatehealthinsuranceindustry.
inmmature capitalmarkets, and direct and indirectgovern- Finally,limitedadministrativeand managementcapac-
ment policies regarding insurance, capital markets, and ityof companiesor intermediarieshaveshrunkprivateinsur-
health sectoractivities. ance in Kenya and Tanzania. In both countries high
Health insuranceis often a by-productof other forms incidencesof fraud amongconsumersand providershave
of insurance-as in Egypt,Jamaica,andTurkey-and often resultedin a shift from third-partypayersto self-financed
serves as a loss leader for other insurance products. plansamonglargecompanies.Self-regulationclearlyis not
Inadequate actuarial data raises risks for investorsand viable in these circumstances(TAI 1997). Similarlapses
restrictsreinsuranceto catastrophiclosses.Accessto and haveoccurredin other countries,particularlywherethe pri-
the cost of capital and reinsurancehave posed problems vateinsurancemarketis small(asitis in KenyaandTanzania).
forprivateinsurancegrowthin C6te d'Ivoire,Hungary,and Similarproblemsarisein countrieslike the Dominican
Jamaica(Fuenzalida-Puelma 1996;TAI1997;Lewis1988). Republic,whereprivateinsuranceis evolvingbasedon small
In CostaRica,whereprivateinsuranceisillegal,healthinsur- employersor individualswhopurchasehealthinsuranceand
ance existsbut is arrangedby employersin creativeways obtain care from smallclinicsand individualpractitioners
through contractsand other paymentmechanisms.There under fee-for-servicearrangements.Becausepublic over-
are no insurancecompanies. sightis minimal,transactionscostsare high and.abusesare
Agovernment's roleinfosteringorimpedingprivate health- difficultand costlyto monitor,the insurancemarketisinher-
relatedinvestmentis a key elementin the growthof private entlyunstable.
healthcarefinancingand servicedelivery.Moreover,private Some governmentshave made concerted efforts to
financingrequiresprivateservicedeliveryto be viable,and improve the climate for private investmentsin insurance
privateservicedeliveryrequiresprivatefinancing.Importcon- and insurance-provider arrangements.SouthAfricahas done
trols Jordan),hightariffs(Nigeria),foreignownershiprestric- sothroughderegulation, Ugandaandthe Philippines through
inflation(Peru,Turkey),andpolitical tax reform, and Sri Lanka through a policyencouraging
tions(Brazil,Philippines),
uncertainty(C6te d'Ivoire) are commonlycited as reasons privateinsuranceand reducingtariffs on medicalimports.

PRIVATEINSURANCE:PRINCIPLESANDPRACTICE
95
The abilityto repatriateprofitsin Eastern Europehas pro-
duced a favorableinvestmentclimate despitethe current
government-dominated health financing system. The
absenceof suchimpedimentsmakeother difficultiesmore
manageablefor investors(TAI 1997).
Health Insurance Regulation
The problemsdiscussedearlier-adverse selection(com-
paniesavoidinghigh-riskenrollees),moralhazard(increased
consumptionbyconsumerswhoare coveredbyinsurance),
imperfectinformation(consumerswho do not understand
the marketin whichthey are purchasinghealthinsurance),
andhigh administrativecosts-are highlyrelevantto devel-
oping countries.This section summarizesavailableinfor-
mation on the regulation of private health insurance in
developingcountries,and the experiencesto date.
The incentivesof privateinsurersare often incompati-
ble with the social objectivesof affordableuniversalcov-
eragethat manygovernmentshavepledgedto ensure.This
is whyfew countrieshave been able to build their health
care systemsexclusivelyon private insurance financing.
These two sets of objectiveshavebecome so intertwined
that regulationof insurershas evolvednot only to ensure
that the insurancemarket is competitiveand financially
sound,but alsothat it meetssocialobjectivesof access,ade-
quate benefits, and consumerresponsiveness.This dis-
tinction is important. Chile, Korea,South Africa,and the
United States can rely on private insurance because the
socialcontractdoes not guaranteeequalaccessto the same
set of services.Efficiencyand relianceon market forces
are bigger priorities,and the private insurancemarket is
seen as the best way of achievingthose objectives.
Insuranceregulationis essentialin societiesthat endorse
the conceptofequalaccessto payersof healthcare,to ensure
that companiescannotexcludehigh-riskindividualsor costly
preexistingconditions. Moreover,the cost increasesfre-
quentlyassociatedwith largelyprivatepayers(asin Brazil,
SouthAfrica,and the United States) cannotbe contained
effectivelywithout incentivesfor controllingthe cost and
volumeof care. In short, governmentscan playa keyrole
in promotingcost containmentthrough insuranceregula-
tion. But this function is secondaryto fairness,efficiency,
and financialaccountabilityfunctions.
INNOVATIONS IN HEALTH CARE FINANCING
96
Regulationof private insuranceis most commonlyan
extensionof insurancelaw,and the institutionsthat over-
see health insurance are often part of or affiliatedwith
(generic)insurance regulation.This is the case in Brazil,
SouthAfrica,and Turkey,amongothers. Insuranceregula-
tion overseesfinancialviability,reserves,reinsurance,and
exitfrom the industry.Evenwherethose functionsarecom-
petently executed for the non-health insurance industry,
as in SouthAfrica,their record in regulatingand enforcing
rules for the health insuranceindustryis poor.
Regulationis generallyweak andunevenin much of the
developingworld.Thisshortcomingis partlyattributableto
an inadequateregulatoryframework,but alsoto weakinsti-
tutionalcapacity.Withafewexceptions(Colombia,Hungary)
there are almostno comprehensiveregulationsfor health
insurance.Even Korea,with its heavyrelianceon private
insurance,has almostno regulations.Moreover,evenwhen
there are regulationson the books, enforcementis often
limitedor ineffective.
Regulotory
institutions
in developing
countries
The body regulatingprivatepayersvariesby country.It can
be central,state, or evenlocal.What is crucialis that reg-
ulatorsbe independentofthe industryandhavethe author-
ity to influenceinsurerbehaviorto preventabuseanddamage
to the industryand consumers.
Argentina.The Argentineregulatoryagency,Adminis-
traci6nNacionaldelSegurodela Saluid(ANSSAL),is respon-
sibleforoverseeingthe financialadequacyof sicknessfunds
(ObrasSociales).The agencyis largelypolitical,with close
ties to organizedlabor and with inadequate and inappro-
priatelytrainedstaff.Thesefeatureshaveunderminedeffec-
tive regulation and enforcement of rules, since political
imperativesprevailin the faceoffinancialdifficulties.Current
reformsaim to improve the scope,benefits, and solvency
of the better-run Obras Sociales.Doing so, however,may
exacerbatethe regulatoryagency'sweaknesses,since it will
requireit to playa largerand more complexrole.
Brazil.Privatepayershaveburgeonedin Brazilin the past
decade,andthe varietyof financingmechanismshas grown
commensurately. The Superintendencia
de SegurosPrivados,
the Sao Paulo state insurance regulatorybody, oversees
only indemnityinsurance,whichrepresentsjust 4 percent
of allprivatefinancing.Almostnocontrolor oversightis pro-
videdin the establishment,operation,andmedicalandfinan-
cialperformance of the prepaid group practices,medical
cooperatives,and companyhealthplans (self-insuredcom-
panies)that togetherwithinsurancefinancehealthservices.
Colombia. Colombia'snewhealthcarereformlawincludes
a set of decreesestablishinga healthregulatoryinstitution
(Supeznntendencia de Salud)that is meant to take on many
of the functionsoutlinedin the firstsection of this paper.
The regulatorystructurewas modeledon those of OECD
countriesand is designedto addressa range of issuesthat
traditionallyare not includedin middle-incomecountries'
healthinsuranceregulations.TheSupenintendencia isintended
to ensure adequate structure,performance,and outcomes
by accreditingprivateprepaidplans(tobe financedthrough
vouchers), establishingbasic quality standards and han-
dlingconsumercomplaints,and ensuringthe financialsol-
vencyof insurersbymonitoringreserves,liquidityandcapital.
Hungary.Allprivateinsuranceproductsin Hungarycome
underthe aegisof theStateSupervisoryAuthorityofInsurance.
Originallyset up to overseeallinsurance,the authorityhas
been expandedto includefor-profithealthinsurancecom-
panies.Thelawwasrevisedin 1993to conformto European
Unionstandards.Nonprofitinsuranceisregulatedbythe law
governingvoluntarymutual insurancefunds.Limitedregu-
latoryopportunitieshavelefttheseinstitutions
largely
untested,
but the necessarystructureis in placeto deal with a poten-
tialmarket (Fuenzalida-Puelma 1996).
Korea.Korea'scompulsorynationalhealthinsurancelaw
of 1989expanded private healthinsurancebased on fee-
for-service
payments,withthe governmentsettingfee sched-
ulesin consultationwith other players.Furthergovernment
regulationshave not been developed.
Turkey.Recentinsurancereformin Turkeyhas included
establishment of the Department of Life and Health
Insurancein 1995and the creationof an insurancesuper-
visionboardto regulatethe insuranceindustry.Healthinsur-
ance oversightis almostnonexistent.
PRIVATEINSURANCE: PRINCIPLES AND PRACTICE
97
Uruguay. Althoughits systemisbased on sicknessfunds,
Uruguaytakes a differentapproachto regulation,and has
a relativelycomprehensiveset of regulations.In 1981the
governmentmadea legaldistinctionbetweenfor-profitpay-
ersand the nonprofitsicknessfunds (calledInstitucionesde
AsistenciaMedicaColectiva,or IAMCs), and regulations
for the two groupsdiffer.
For-profitinsuranceplansare accreditedbythe Ministry
of Public Health, which requires them to meet certain
standards,definesnorms fortheir establishmentand oper-
ation, and requiresthem to provide it with basic informa-
tionon aregularbasis.The MinistryofEconomyandFinance
regulatesfinancialaspectsof the IAMCs.
Financial
oversight
Monitoringthe financialpracticesand solvencyof private
health insurersis probablythe best understoodand most
widelyaddressedissuein regulation.However,experiences
and policyvaryconsiderablyacrossdevelopingcountries.
In Chile regulationof financialsolvencyhas been the
soleresponsibilityof the Superintendencia de ISAPREs,the
regulatorybodyofprivateinsurers,or ISAPREs(Instituciones
de SaludPrevisional). The superintendencymonitorsfinan-
cialreservesand ensures that the ISAPREsremain finan-
ciallyviable.
Asthe firstsucharrangement inLatinAmerica, the ISAPREs
wereestablishedforindividualratherthan grouppolicies.As
such, they have engagedin aggressive"cream skimming"
(attractinglow-risk,low-costpopulationsinto privatehealth
plans).Becausetheyare notrequiredto acceptanyenrollees,
the ISAPREshaverefusedto insurehigh-riskworkersor the
elderly.Peoplewith preexistingconditionsand chronicdis-
eases,as wellas the elderly,havebeen forcedto relyon the
publichealthsystem.The lackof regulationhas led to classic
problemsof adverseselectionin privateinsurance.
Financialoversightis the most developedand system-
aticelementof Colombia'snewregulations.Insurancereg-
ulationsrequirethat insurersbe visitedeverythreemonths,
and an informationsystemhas beendesignedto allowtrack-
ing of financialflowsand productivity.Insurers can invest
in other enterprises,but the health investmentsmust be a
legallyautonomousentity (that is, a subsidiarycompany)
with separate investmentinterests (Londofio1997). The
scope of the new regulations and the country's limited expe-
rience with this form of regulation have made implemen-
tation difficult.Indeed, the regulations place a large burden
on regulators, who must deal with institutions that are not
accustomed to systematic government oversight.
Among sickness funds in Argentina, South Africa, and
Turkey,allbut SouthAfrica'sMedicalSchemes(sicknessfunds)
run chronic deficits that are financed with government sub-
sidies (TAI 1997;World Bank and IDB 1997).Although the
Medical Schemes are experiencing financial difficulties,the
lack of regulation inhibits the government from stepping in,
and the weaker schemes are in danger of collapsing.
Argentina's Obras Socialeshave recentlyrun highdeficits,
and some are insolvent. While the Argentine regulatory
agency,Administraci6n Nacional del Seguro de la Saluid,is
responsible for overseeing the financial practices of the
Obras Sociales, it has been largely ineffective. As a result
the government has periodically been forced to step in to
prop up ailingObras. The systemis currently being restruc-
tured to address these and other deficienies.
Prohibitionson underwriting
Underwriting by insurance companies is economicallyeffi-
cient for the reasons discussed in the first section of this
paper: it is efficient to group homogeneous risks and to price
insurance coverage accordingly.The social consequences of
such practices are considered inequitable and unacceptable
in many countries, however. As a result health insurance
regulations often include measures to prevent or discour-
age such practices.
In Colombia insurance policies are either group or indi-
vidual, but individuals cannot legallybe denied a basic pol-
icy-although certain conditions cannot be claimed within
the first six months or year of the insurance contract.
Uruguay's nonprofit IAMCs are tightly regulated in all
financial and access issues; for-profits have relative freedom
in the scope and pricing of health insurance products. The
IAMCs have defined enrollment procedures and premiums;
copayments and physician fees are set by the Ministry of
Economy and Finance.
The IAMCs underwrite potential enrollees and must
grant enrollment rights if there are no preexisting condi-
tions. People with preexisting conditions have immediate
INNOVATIONS IN HEALTHI CARE FINANCING
98
access to emergency care, but must wait one year for surgery
or two years for medical care. Under pressure from the
IAMCs, the government passed a decree that imposes entry
restrictions on individuals over the age of 50 or 60. These
restrictionscan be total (that is, rejection of applicant because
of age) or partial (constrained access to services such as
medication or outpatient services).
The strict regulation of nonprofits and the absence of
regulation of for-profits have divided the market in much
the same way it has in Chile. The for-profits (ISAPREs in
Chile) have designed contracts that appeal to the young and
the healthy, and the IAMCs (Ministry of Health in Chile)
are increasingly serving the elderly and the chronically ill.
Thus the costs of the IAMCs keep rising, as they do in the
public system in Chile. The IAMCs, however, are overreg-
ulated; it is difficult to become efficient when prices, costs,
and beneficiaries are beyond the insurer's control.
Nongovemmental regulation
In countries where government has not provided a regula-
tory framework for privatehealth insurance, private groups
have started to step in to try to maintain the industry's
credibility and integrity. The industry's goals are to protect
its market and survival-but unlike government, it does not
seek to protect consumers. The standards set by the health
insurance industry are necessarily voluntary, and enforce-
ment is problematic. The limited experience in middle-
income countries is consistent with this principle.
Financial integrityand fairnessin access are neither mon-
itored nor required under any specific Brazilian legislation
or legal decree. The concerned trade group for prepaid
group practice, ABRAMGE (Associacdo Brasileira de
Medicina de Grupo), officiallyrequested government regu-
lation to protect the integrity of the industry, but the gov-
ernment refused. The group has created its own regulatory
body, CONAMGE (National Council for Self-Regulation
of Company-Based Prepaid Group Practice), which repre-
sents about 90 percent of all enrollees in such plans. The
association is attempting to provide guidelines that will
reduce fiscal irregularities and false advertising. A spin-off
association, the National Council for Self-Regulation of
Advertising, is considering setting up a regulatory arrange-
ment for advertising among its members. However, both
are voluntary programs, and punitive actions against trans-
gressors have not yet been spelled out, limiting the clout of
such efforts (World Bank 1994).
Korea, lacking regulation or careful monitoring, has seen
a rise in inefficient risk pooling, excessivetechnology acqui-
sition, and fraudulent billings.Some self-regulationhas been
introduced by the Federation of Korean Medical Insurance
Societies, which represents a large portion of insurers. The
federation has established a National Insurance Appeals
Committee to handle enrolleeproblems,developed a process
for identifyingdesignated treatment facilities,and overseen
a program that provides training, routine inspections, and
audits to ensure proper operation of member insurers
(FKMIS 1986). Although these mechanisms do not neces-
sarily work well (Yang 1996), they provide an infrastruc-
ture on which to build a potentiallyviable regulatory system.
Turkey's private health insurance industry views regula-
tion as essential to maintaining its integrity. Given the lack
of government action, health insurance companies are set-
ting standards for their members' organization and payment
of providers. They are not, however, setting financial per-
formance standards, and enforcing their policies is diffi-
cult (Fuenzalida-Puelma 1996). However, these initiatives
suggestthe importance of regulation in protecting both con-
sumers and the health insurance industry in Turkey, as in
all countries.
Orderlyentry and exit
Controlling entry into the insurance market promotes finan-
cial viabilityand ethical business conduct. The careful mon-
itoring that occurs inthe United States is difficult to replicate
in developing countries. Still, simple controls such as lim-
iting the types of investors and banning previous abusers
can help establish a process of orderly entry. Exit regula-
tions ensure that firms leaving the business do not leave
their clients without coverage or with large debts (because
they have not been reimbursed for services consumed).
Strict entry regulations are rare in developing countries; exit
policies are more common.
In Colombia entrants into the market must be licensed
by the government and have minimum capital reserves.
Troubled insurers are counseled, and firms that exit are
taken over by the Superintendenciaand their assets liqui-
PRIVATE INSURANCE: PRINCIPLES AND PRACTICE
dated. Insurance contracts are then transferred to other
insurers. Chile's regulatory body cannot intervene when
insurers falter, but upon exit it takes over the firm, pays off
creditors (includingpatients awaiting reimbursements), and
liquidates assets (Oyarzo 1997). In Uruguay IAMCs were
originallyrequired to have a minimum of 20,000 enrollees;
those below that level were merged or exited from the
market, with the insured transferred to other IAMCs.
Consumerprotection
Consumer protection regulation is just beginning in non-
OECD countries. India's 1986 Consumer Protection Act
was passed with great fanfare and was an important start.
However, professional groups claimed that the law under-
mined the physician-patientrelationship,and weak enforce-
ment has resulted in few changes. Public interest groups are
now pressing for judicial intervention and government
enforcement of the law.Despite progressin filingcomplaints,
there are serious backlogs in bringing these cases to adju-
dication (Bhat 1996).
A similarsituation exists in Brazil, where consumer pro-
tection laws exist but adjudication under these statues is
difficultto assure. Options for individualswho have encoun-
tered problems tend to be protracted, and often result in
unsatisfactory decisionsfrom a less than transparent process
(World Bank 1994).
In Chile the Superintendenciadoes not regulate bene-
fits, is not responsible for quality lapses in provision, and is
prohibited from giving information to consumers to influ-
ence their choice among competing private health funds
(Oyarzo 1997).
Uruguay has attempted to introduce competition into its
socialinsurancesystem.The nonprofit IAMCs areclosely reg-
ulated, but consumerscan choose among them (although, as
discussed,there are restrictionsamong the elderly).Enrollees
can transfer among the IAMCs subject to certain enrollment
requirements and a maximum hiatus of two months between
departing from and enrollingin another IAMC.
Enforcementexperience
Many countries lack an adequate regulatory structure, and
those that have regulations in place often cannot enforce
99
them.Indeed, enforcementat alllevelsof governmenthas
encountereddifficultiesin the countriesforwhichthereare
data.
Despitethe logicand strengthof Colombia'snew regu-
lations,enforcementis havingmixedresults.Partof the dif-
ficulty,and an issue that has emerged elsewhere,is the
politicizationof the regulatoryfunction, whichultimately
underminesthe effectivenessand credibilityof regulation.
Still,the gains from establishingan appropriatestructure
are a significantachievement(Londofio1997).
Sicknessfund systemsin Argentina,SouthAfrica,and
Turkeysharesome of these difficultiesin regulation.First,
becausethese systemsare employment-based,there is no
competitionamongfunds.Second,governmentregulations
oftencontrolthe decisionmaking of the sicknessfundadmin-
istrators,leadingto problems.Third, few of these arrange-
ments ensurethat abuses associatedwith healthinsurance
(moral hazard) and enforcement of the social contract
(adverseselection)are mitigatedthroughlegaland regula-
torymeasures.Finally,none of the funds is accountableto
enrollees,firms,government,or reguiatorybodies.
Argentina'sregulatoryagency,ANSSAL,has had trou-
ble with enforcementpartly because of its complexrela-
tionshipwiththe sicknessfundownerswholeadtradeunions.
Indeed, recentreformssuggestthat ANSSALwillneed to
strengthenits role to include overseeingenrollment,pro-
vidingtechnicaland financialassistancefor mergers,and
closingObrasSociales.Despitethisincreasedrole,ANSSAL
continuesto be challengedby enforcement(WorldBank
and IDB 1997).
In Brazilthereare no institutionsto protectagainstfinan-
cial insolvencyor sloppy exits from the market, a threat
that couldunderminethe country'sentireprivateinsurance
market. Someforms of regulationhave been tried in vari-
ous Brazilianstates,but weakand politicizedenforcement
has undermined the effectivenessof these experiments
(WorldBank 1994).
Inappropriate
regulations
Regulationis difficultfor most countries,especiallyin the
healthsector.Whilethere is an importantrole for govern-
ment in many areas, there is also a risk of inappropriate
regulationsthat controlthe sectorin an unreasonableman-
INNOVATIONS IN HEALTII CARE FINANCING
100
ner rather than regulateit. Rigidprice controlson private
insurance,genericspace requirements,and infrastructure
and technologyspecificationscan generallybe classifiedas
nuisanceregulations,and ultimatelyprovidelittlein the way
of true regulation.3 3 The objectiveof healthinsurancereg-
ulationis to moveawayfrom controlsto a set of incentives
and disincentivesthat ensure the proper conduct of busi-
ness. Most countries,however,have focused on bureau-
cratic requirements and have ignored the more difficult
but essentialfunctionsfor regulatingprivatepayers.
Theconceptofpricecontrolsis an importantone.During
periods of high inflationin the 1980sBrazilfroze private
healthinsuranceprices,includingthose for managedcare
plans.This led to overcompensationin price adjustments
whencontrolswere lifted,in orderto hedgeagainsta return
to price controls. A similar pattern occurred with price
controlsin Uruguay.Both initiatives'soleobjectivewas to
stop price increases-the reasonablenessof health insur-
ancepriceswasnever considered.Suchpoliciesdistortthe
marketforhealthinsurance,andin the end havelittleimpact
on overallprices.
In the UnitedStatessomestatesapproveinsurancerates.
Thesetypicallyinvolvegeneraltargetsforpremiumincreases
based on lossratios (that is,costs/revenues).Thisapproach
allowsprofit levelsto be containedwhile allowinghealth
insurancecompaniesmaximumflexibilityin management
and sufficientprofit to ensure competitionin the market.
Someregulationsare too specificand haveofferedper-
verse incentives.In 1989Uruguay'sgovernmentdecreed
the necessaryinfrastructure,equipment,andinpatientcapac-
ity that the privatesicknessfunds (IAMCs)wereto possess
based on an assessmentof resource availabilityand the
perceivedeffectivenessof those resources.Financialfeasi-
bilitywasdefinedaccordingto nationalandregionalhealth
capacityaswellasfinancialconstraints.Thesewere adopted
in another decree that set IAMC investment levels and
increased monthlypremiums accordingly.The result has
been overcapacity, risingmaintenancecosts,and higheruti-
lization,part of which government finances and part of
whichcomesfrom privateinsurancepremiums.
Most health sectorregulationsin developingcountries
are weak, and regulatorystructures(wherethey exist) are
often unenforced or inappropriate.34 Weak regulations
discourageprivate insurancebecause there is a dearth of
competentprovidersfrom whom to purchase health care
services.Moreover,when payersand providersmerge (as
theydo under managedcare),regulationson the organiza-
tion anddeliveryofhealthserviceshavean increasingimpact
on healthinsuranceregulation,its structure,and enforce-
ment. Hence, although a separate issue, regulation of
providers (hospitals,clinics,physicians,and other health
Conclusions
and Lessons
Learned
Privateinsuranceexistsin almosteverycountry,anditspoten-
tialis significant.In countriesthat alreadyhavealargeprivate
insurancesector,it is likelyto remain,sincereversinggov-
enmnentpolicytoward amajorconsumerindustrycanbepolit-
icallydifficult. Whether countries choose to encourage,
discourage,or simplyacceptprivateinsurance,it isimportant
that governmentsunderstandthe healthinsuranceindustry.
Private health insurancehas both benefits and pitfalls
relativeto purely public financingsystems.Private insur-
ance can improvethe availabilityand quality of health. It
canhelphealthcaresystemsrebuildinfrastructureand amor-
tize needed investment when government financingfor
health care is inadequate.It can offer consumersa choice
of providers,and therebygenerateincentivesthat reward
high-qualityprovidersand penalizepoor quality(at leastas
perceivedby consumers). Similarly,private insurers are
becomingincreasingly adept at distinguishing
efficientfrom
inefficientproviders,and can developmethods that sys-
tematicallyreward efficiency.(By contrast, public financ-
ing systemsoften reward efficiencyand inefficiencyalike.)
Finally,althoughprivateinsurancecansiphonofflower-cost
peopleandleavethe publicsectorwith higheraveragecosts,
it can reduce the total fnancial burden on government.
The pitfallsof private insuranceare, in many cases,the
flipside of their advantages.Privateinsurancesystemsare
naturallydynamic.When privateinsuranceplanscompete
with one another for enrollees(or when one or more pri-
vate insurancecompaniescompetewith a publicinsurance
program),the compositionof anyone insurer's risk pool
can change as consumersmove from plan to plan. Plans
that experienceadverseselection,evenwhenthey arewell-
run and efficient,can fail.
PRIVATE INSURANCE: PRINCIPLES AND PRACTICE
101
professionals)has a directbearingon healthinsurancereg-
ulation and the regulatoryenvironmentin general.
These lapses-either in terms of impedinggrowth of
the private sector or allowingit free rein with no controls
or policysignalsfrom government-lead to an uncertain
businessenvironmentand an inhospitableclimatefor pri-
vate investmentin healthinsurance.
Thus insurershavedevelopeda number of underwrit-
ingandpricingpracticesto avoidacceptingsickpeoplewhen
theyapplyfor coverage.Froma purelyeconomicperspec-
tive, these techniques can improve economicefficiency,
althoughtheymaybe sociallyunacceptable.Thesepractices
expeditethe formationofhomogeneousriskpoolsin which
participants' expected costs are similar.When these risk
poolsform and are priced efficiently,theyreduce involun-
tarycross-subsidies amongparticipants-a resultthat is eco-
nomicallyefficient.But they can also reduce the abilityof
sickpeopleto findinsurance,increaseadministrativecosts,
and reduce the percentage of health care expenditures
that actuallyfinances health care. By attracting healthier
risks, insurers can quicklyerode the broad risk pooling
that socialinsuranceattemptsto create,strainingthe social
contract's abilityto ensure the universal,equal coverage
that manycountrieshave embraced.
Moreover,the abilityof private insurance systemsto
improveaccessand quality,and to help recapitalizehealth
care capacity,can cause health care costs to accelerateas
private insurancegrows.Greater accessproducesgreater
use of health careservices;the acquisitionand use of costly
technologiesfurtherincreasecosts.The challengeto regu-
lators-and to privateinsurersthat hope to expandprivate
coverage-is to keep private insurance affordable while
encouragingongoingimprovementsin the qualityof health
care.This is not a simpletask.
Other potentialproblemsof privateinsuranceare unre-
lated to its advantages:financialinstability,
unethicalbehav-
ior,andunreliableconsumerorientation.Governmentscan
resolvethese problems,but onlyiftheyestablishclear,prob-
lem-focusedregulationsand enforcethem systematically.
 Regulationof private insurancecan take many forms. * Regulation canbetooperfect.Governmentsthat havehad
Dependingon its nature, scope,and enforcement,regula- verylittle regulationmay find it impossibleto leap to a
tion canhavedifferenteffectson consumersand providers, comprehensive systemofinsuranceregulation.Forexam-
and thereforeon costs.In Germanyallaspectsof the pri- ple,Colombia'snewcomprehensiveset ofinsurancereg-
vate,noncompetingsicknessfunds are so closelyregulated ulationisan impressive movebythe government to ensure
as to be quasi-publicinstitutions;companybehaviorand oversightof insurersand health care providers.But the
performanceare tightlycontrolled.At the other extreme, regulationimposesan abrupt shift in culture.Asa result
most non-OECDcountrieshavealmostno oversightorreg- the breadth of the changesand the adoption of sophis-
ulation.The absenceof regulationcan have effectson the ticated methods havebeen difficultto establishand to
industrythat are as perverseas those caused by excessive enforce.A better approachmaybe to start with a core
and irrationalregulation. set of regulationsthat rectifyegregiousproblems;once
Regulationis essentialfor anycountrythat has a private these regulationsbecome effective,regulatorscan turn
insurancesector.Some of the most importantlessonsfor systematically to other areas of concern.Tradeoffsneed
regulatorsabout healthinsuranceinclude: to be madebetween regulatoryneeds and the abilityto
* Knowyour industryand the majorplayers.Information effectivelyenforceregulations.
about the practicesof companiesand their managers, * Regulationsmustbe effectiveand offerproperincentives.
and about the performance and financial integrityof In manycountriesnonfinancialregulationsare nuisance
health insurers,is essentialto preservethe industryand regulations.Theycreatesignificantbureaucraticred tape
to protectconsumers.Regulatorswhoareunfamiliarwith and have littleimpacton qualityor other desirableout-
the specificsof their country's insuranceindustrywill comes.Forthe insuranceindustrysuchregulationsmerely
be unableto anticipateand preventabuses. raisecosts.Tobe effective,regulationsmust havea clear
* Allhealthinsurance systems
mustbalanceefficiencywithcul- objective,and the lawsand decreesput in place must
turalnotionsoffairness.Countriesthat have adopted a servethat end as directlyas possible.The experienceof
sweepingsocial contract, and seek to ensure absolute a number of countriessuggeststhe need to reviseand
equalityto all,willhaveimportantproblemswiththe eco- craftlegislationto strengthenappropriateincentivesand
nomic efficiencyof the insurancesystem they create. to discourageperniciouscompanybehavior.
Conversely,countriesthat allowinsurersto isolatesick Regulationwithoutenforcement islikea tigerwithoutteeth.
peopleintohigh-costinsurancepools,andoffernomeans An effective regulatory system relies on an effective
to subsidizepeoplewith highhealthcare costs,willhave enforcementstrategy.Regulationsthat succumbto polit-
a systemthat maybe economicallyefficient (in that it icalimperativesor that allowexceptionsunderminethe
doesnot forceinvoluntarycross-subsidies) but that most process and purpose of regulation. Sound regulatory
peoplewillconsiderunfairorinadequatelyprotective.All requirementswillbe ineffectiveiftheycannotbe enforced
insurancesystems,public and private,must strike a bal- or are not enforced systematically
ancebetweeneconomicefficiencyandequity.Wherethey Finally,evencountriesthat want to retain a broad pub-
lieon the spectrumwillbe determinedbyeachcountry's lic insurance program to finance health care can benefit
socialandpoliticalculture,and byits economicresources. fromunderstandingandjudiciouslyusingprinciplesof pri-
* Healthinsurance companies, whetherfor-profitornonprofit, vate insurance. If it is encouragedand directed toward
mustmakea profit.Governmentcan regulateinsurance explicit systemgoals, competition is a powerful tool for
companies' profit levels (usuallyby regulating price improvingefficiencyand quality.For example,the Czech
increases),but preventingprofitwillunderminethe com- Republichas initiated competitionwithin its public insur-
pany and the industry.Withoutprofit,insurershaveno ance program, allowingregionalgovernment-sanctioned
means of financinginvestmentin efficientmanagement insuranceplansto formand competewith the nationalplan.
systerns and effective systems of health care cost As mightbe expected,some of the new regionalplansare
management. doingwell:theyareattractinggrowingenrollment.Theyare

INNOVATIONS IN HEALTH CARE FINANCING

102
also causing the national plan some discomfort,since it
believesthat it is being left withsickerenrollees.This expe-
riencebears observation,but in the short run it willmoti-
vate an ongoingsearchfor waysto improveadministrative
efficiencyand customersatisfactionin all plans.
Private health insurance offers benefits and costs to
society.Its importance and applicabilityvary by country
and culture, but in most countries it is well-established.
The challengeto governmentsis to regulate privateinsur-
ance effectively:to assure the financialstabilityof insur-
ers and the integrityof insurance contracts,and to strike
a balancebetween the socialacceptabilityof private insur-
ance and insurers' competitive drive toward economic
efficiency.

PRIVATE INSURANCE: PRINCIPLES AND PRACTICE

103
Annextable I Roles,characteristics,
and extentof privatehealthinsurance
andmainalternatives,OECDcountries
Country Roleof privateinsurance Characteristics
Australia Supplementalto universal
publiccoverage
Belgium Supplementalcoveragefor
the self-employed
Canada Supplementalto universal
publiccoverage
France Supplementalto universal
publiccoverage
Germany Altemativeto universal
publiccoverage
Ireland Supplemental to universal
publiccoverage,with two
levelsof benefitcoverage
coverage
Shareof
populationwith
of privateinsurance privateinsuranceMain alternative
Paysonly for hospitalcarein private 45 percent
facilities,includinginpatientphysician
chargesthat exceedthe publicfee
ratefor inpatientcare.
Paysfor ambulatory,
nonsurgical
care. n.a.
Only coversservicesnot covered n.a.
by the provincialpublicplans.
Paysfor privatehospitalandphysician n.a.
chargesthatexceedcontractual fees
paidbythe publicprogram.About25
percentof physicians (mostlyspecialists)
balancebill pabentsfor care.Covers
requireddeductible.
Peoplewho opt out of the statutory 9 percent
systemin favorof privateinsurance (1986)
cannotever rejointhe statutorysystem.
Voluntaryprivateinsurance,soldby a 30 percent
monopolistcstatutoryhealthinsurerto (1991)
peoplewho are partlyeligiblefor public
programcoverage,coversphysician fees
andprivatehospitalaccommodabons
(including
privatebedsin publichospitals).
INNOVATIONS IN HEALTH CARE FINANCING
104
sourceof financing
Publicuniversalprogram(Medicare)
financedfromthe
generalincometax.
Compulsorysocialinsurancecoversrisksfor the entire
populabon
andminorrisksfor allbutthe self-employed.
Universalpublicinsurance planfinancedfrom nabonal
andprovincial
generalrevenues. Physirians
arein private
practiceandpredominantly paidon a fee-for-service
basis.Balancebillingto patientsisprohibitedby law.
Comprehensive publicinsuranceprogram.Private
providerfeesare negobatedby the publicinsurance
program.Supplementary privateinsuranceto cover
costsharingiscommon.
Universalentitlementfor the statutoryhealthinsurance
systemcomprising1,200self-financing sicknessfunds
organized by geographic area,finm,trade,or craft.
Membership iscompulsory for peoplewithina specified
incomelimitandfor rebredpeoplewho participated
whileworking.Aboutthree-quarters of participants
in
the statutorysystemare mandatory;one-quarter
participatevoluntarily.
Publicprogramisfundedfromgeneraltaxationand
providescomprehensive benefits. Aboutone-thirdof
the populationreliesexdusivelyonthe publicinsurance
programwith noout-of-pocketpayments for care;the
restof the populationispartlyeligible.Of these,more
thanhalfpayfor somecareout of pocket,but haveno
supplemental privateinsurance.
Annextable I Roles,characteristics,
andextentof privatehealthinsurance
coverage
and mainalternatives,OECDcountries(continued)
Shareof
populationvith
Country Roleof private insurance Characteristicsof privateinsurance privateinsurance Mainalternativesourceof financing
NetherandsAlternativeto universal Mandatoryregulatedprivateinsurance 30 percent Compulsoryuniversal publiccoveragefor chronic
publiccoverage coveragefor acutecarefor higher- (1991) care;publiccoveragefor acutecareiscompulsoryfor
incomepatients;insuranceplans about70 percentof the population.
reimbursepatients.
Spain Supplementalto universal Typically
coverssupplementary n.a. Compulsorynational systemof publichospitals
and
publiccoverage voluntarypaymentsfor hospitaland publiclysalariedphysicians,
fundedfrom general
physiciancare. revenuesandsocialinsurancecontributions.Some
provinceshavetheir own socialinsurancesystem.
United Supplementalto universal Coversacutecareservices,supplements 10percent Comprehensive, compulsorynationalsystemof public
Kingdom publiccoverage or substitutes
for paymentby the (1990) hospitalsand publiclysalariedphysicians,
financedfrom
NationalHealthSystem. generalrevenues.Patientsareassigned to primaryand
institutional
careproviderslargelyby placeof residence.
Efficiencies
haveemergedfrom recentreformsthat
introducedsystems of provideraccountability,
noton
competitionamongplansor providers.
United Primary,voluntary Primarilya tax-exempt,voluntary 71 percentof Publicassistancecoverageisavailable to poor individuals
States coveragefor people benefitfrom self-insuredor insured population in specificdrcumstances(underage 13,over age65,
ineligiblefor the national employerplans.About7 percentof under65 or permanently disabled);eligibilityandcoveragevaries
socialinsuranceprogramfor peopleunder65 buy privateindividual (1995) by state.Neariyallpersonsover 65qualifyfor cover-
the elderiy(Medicare)or (nongroup)coverage.Aboutone-third agefrom the socialinsurance program(Medicare).
federal-statepublicassistanceof elderlypublicinsuranceenrolleesbuy
healthprogram(Medicaid). supplemental privatecoverage.In any
Publicinsuranceenrollees year,i 6-19 percentof the population
maybuy supplemental under65 isuninsuredallyear.
privateinsurance.
Source:
Aftman
andJac6son
1991;
Chollet1996;DayandKlein1991;
Enthoven
1991;Hurst1991;Reinhardt
1995.

PRIVATE INSURANCE: PRINCIPLES AND PRACTICE

105
Annex table 2 Roles,characteristics,and extent of private health insurancecoverage
and main alternatives,non-OECD countries

Shareof
populationwith
Country Roleof privateinsurance Characteristics
of privateinsurance privateinsurance
Argentina Additionalto socialwelfare White-collarworkerstypicallyredirect
fundsrun by tradeunions 75 percentof mandatoryhealthcon-
tributionsto prvate, prepaidmedical
plansthat varyin coverage.Theplans
normallycontractwith prvate hospitals.
Brazil Additionalto publicly Mostenrollmentisin prepaidgroup
financedcarefrom largely practices,
medicalcooperatives,
or
privatefacilities employerplansthatown or contract
with healthcarefacilities.
Chile Altemativeto the public Benefitsare tiedto amountof contri-
healthsystem butionandage,sex,and numberof
dependents.Largecompaniesprovide
complementarycoverageor create
their own privateISAPRE.
Colombia Supplemental
to public New entitiescalledEPSs maybe state
coverage run (6), private(I 2),or nonprofit(7).
EPSs competewith the publicsystem.
CostaPica Additionalto public Thereare no private(health)insurance Low
coverage companies.Thewealthyuseprivate
facilitieswhoseservicesare financed
by new payersandproviders.
Cote Additionalor supplemental Mostinsuredare coveredby their 7 percent
d'lvoire to publicoptions employers.Insurance paymentsare
madeto providersbasedon a "payment
guarantee"issuedto patientsprior to
hospitaladmission.
Czech Complementary
to public
Republic insurance
Dominican Addiionalto socialsecurity Largelyemployer-provided
carein
Republic andsupplemental to other prepaid,PPO,andindemnityplans.
publicservices
Ecuador Altemaiveto public Employer-operated medicalplans
coverage;additional
to financeuseof privateclinicsand
socialsecurity doctors.
Egypt Altematve/supplemental Employer-run and -financedclinics. 3 percent
to publiccoverage Typicalprivateplansare noncontributory,
coverall employees,andreimburse
costsup to a fixedlevel.
GuatemalaAddiionalto public Employer-provided, prepaidgroup
coverage healthinsurancefor non-union
employees.Comprehensive major
medicalplans.HMOs andPPOshave
emergedas otheroptions.
Mainalternative
sourceof financing
19percent Compulsorysocialwelfarefundsoperatedby
govemment-affiliated trade unions,aswellas
federal,provincial,andmunicipalhealthservices
(22 percent),socialinsurance(36 percent),and
out-of-pocketpayments(23 percent).
25 percent Compulsorypublicinsuranceprogramandfederal,
(includes state,andmunicipalservices.Out-of-pocket
managed care): expendituresare significant.
indemnityplans
are 4 percent
of total
27 percent FONASAcollects7 percentcompulsoryhealth
insurancepremiumand allocates themto ISAPREs
or the Ministryof Health.Theministry'ssystemis
subsidized
by the govemment.
I I percent Halfthe populationisaffiliatedwith the public
insurancesystem,which issupportedby a
compulsorywagetax that alsofinancesthe EPSs.
Good-qualitypublicsocialsecuritysystem.
Ministryof Healthfacilities.Socialsecuritysystem
for formalworkersprovidesgenerousbenefitsand
lessgenerousmutuellefor govemmentemployees
anddependents.Morethan halfof expenditures
are private.
None Compulsoryuniversal socialinsurancewith
contributionsto I of 20 licensedinsurance
companieswith generous,publiclymandated
medicaland dentalbenefits.
14percent Poor-quality
socialsecurtyandMinistryof Public
Healthservices.Expenditures are 10percent
socialsecurity,20 percentministry,and 52 percent
private(NGOsandfor-profit).
12percent Poor-qualitysocialsecuritysystemandMinistryof
Healthservices.Expenditures are 17percent
socialsecurity,14percentministry,63 percent
private,and6 percentother.
Poor-quality
compulsoryHealthInsurance
Organizationand Ministryof Healthservices.
Expendituresare 30 percentgovemment,54
percenthouseholds,and9 percentsocialinsurance.
5 percent Poor-quality
socialsecurityand Ministryof Health
services.

INNOVATIONS
IN HEALTHCAREFINANCING
106
Annextable2 Roles,characteristics,
andextentof privatehealthinsurance
coverage
and mainalternatives,
non-OECDcountries(continued)
Shareof
populationwith
Country Roleof privateinsurance Characteristics
of privateinsurance privateinsurance r Mainalternative
sourceof financing
Hungary Supplemental
to public Manycompanies grantfree medical Low Socialsecuritybasedon contributions.
coverage consultationsin-houseor in private
clinics.Recentmodificationof law
will increaseprivateprovision.
India Aftemativeto public Expensive medicalinsurance
provided 3.3 percent Systemfundedandoperatedby federalandstate
coverage(operatedat to the richby employers.Govemment govemments.Expendiuresare75 percent
statelevel) employeeshavetheir own system. out-of-pocket,6 percentfederal,and 16percent
stategovemment.
Indonesia Alternativeto public Employer-provided planswdh Minimal PublicsystemOAMSOSTEK). Expendituresare
coverage intemallimitationsor coinsurance 70 percentout of pocket,4 percentof which isfor
to limit costs.HMOs and PPOs privateinsurance.
becomingincreasingly popular.
Jamaica Altemativeto public Typically
coversoutpatientandsome 15percent Ministryof Healthserviceswith someuserfees,
coverage inpatientcare. butfreefor the indigent.
Jordan Altemativeto public Mainlyprovidedasa healthbenefit 12percent CMIInsurance Programfor govemmentemployees
coverage of largefirms. andthe indigent,and RoyalMedicalServices for
the military.Bothare financedby payrolltax and
generalrevenues;cover58 percentof the
population.About53percentof spendingisprivate.
Honduras Addiionalto public Employer-provided
plansthrough 1.5percent Expendituresare 9 percentfor socialsecurity
coverage prepaidandindemnityplans. (formalsectorworkers),29 percentfor Ministryof
Health,and 61 percentout of pocket.
Kenya Altemativeto public Employer-operated, noncontributory 11.4percent State-runcompulsoryinsurancesystemcovers
coverage plansof urban-based companies that peoplein formalemploymentandtheir
offer coveragefor dependents. Five dependents-about25 percentof the population.
companiesareactive.Directemployer
contractingwith providers(self-insuring)
ismore common.
Korea, Compulsoryhealth 417 autonomoushealthinsurance 94 percent 6 percentof populationcoveredby public
Rep.of insurance funds;restrictedaccessto high Medicaidinsuranceprogram.Govemment
technologyand ceilingson inpatient subsidizes
privatefunds.
care.All fundshavedeductibles and
copayments for doctorvisits,outpatient
services(30-55 percent),and inpatient
care (20 percent).
Lebanon AJtemativeto NGO and 90 privateinsurance
companieshave 8 percent Three publicinsuranceschemescoverprivate
publicinsurance enteredthe marketrecentlyanticipating employees(28 percent),civilservants(9 percent),
economicrecovery.Copaymentsare andthe military(I I percent).Expenditures
are 31
standardfor mostservices. percentgovemment,61 percentprivatesources,
and 8 percentdonors.
Malaysia Aftemativeto public Largeor foreignemployersprovide 3 percent Well-runpublicsystem.
coverage privateplanswith differentlevelsof
coveragefor categoriesof employees.
Mexico Additionalto public Largeemployersprovidecoverageto 1-2 percent Socialsecuritysystemsfor formalsectorworkers
coverageandcompulsory non-unionemployees.Typicalplansare andgovemmentemployees,and Ministryof
socialinsurancefor formal noncontributory,
but there isa trend Healthfor the poor.About47 percentof
sectorworkers towardprovidinghigherbenefitsand expendituresare private.
employeecontributions.

(Table
continues
nextpoge)

PRIVATEINSURANCE: PRINCIPLES AND PRACTICE

107
Annex table 2 Roles,characteristics,and extent of private health insurancecoverage
andmainalternatives,non-OECDcountries(continued)
Shareof
populationwith
Country Roleof private insurance Characteristicsof privateinsurance private insurance Mainalternativesourceof financing
Nigeria Supplemental
to public Recentintroductionof medical 0.4 percent Freetreatmentat govemmentdinics,though
coverage insurance on a limitedbasisintandem patientsoftenpayfor medication,medicalaids,
with a self-fundedarrangement. andso on.
Typicallynoncontributory.
Pakistan Altemativeto public Multinabonaland publicsector Low State-runsystemthroughMinistryof Healthand
coverage employeesreceivealmostblanket limitedEmployeeSocialSecurityInsurance.
benefitsfromtheir employers.Limited
otherwise.
Panama Supplemental
to public Companies commonlyprovidegroup Low PublicsocialsecuritysystemandMinistryof Health
coverage medicalinsurance
for white-collarand services.
sometimesplantworkers.Typicallynon-
contributory.
Peru Altemativeto public Group medicalinsurancefor white- 6 percent Expenditures
are 36 percentpublicsocialsecurty
coverage;additionalto collarworkers,characterized
by system,30 percentMinistryof Health,and28
socialsecurity deductibles,coinsurance,
cost sharing, percentout-of-pocket.
andmaximumbenefitlevels.
Philippines Supplemental
to public Two categoriesof groupmedical 1.6percent Medicareprogramthroughcompulsorysocial
coverage insurance:basichospitalizaton
(for securitysystem(covers8.5 percent),andgrowing
typicalmedicalexpenses)andmajor managedcareoptions.About 54 percentof
medical(paysfor a percentageof covered spendingisout of pocket.
medicalexpensesin excessof a specified
deductibleamount).Usuallypaidby
employers.
Poland Limited Highsocialsecuritytaxes(46 percent Minimal Freegovernment-provided healthcare.Patients
of grosssalary)restricthealthasan cover halfof allcoststhroughinformalpayments,
employeebenefit,but investmentis and29 percentindicatethatthey pay'gratuities."
pending.
Singapore Altemativeto public Insuredor self-financed
private Low High-qualitypublicsystemof CentralProvident
coverage medicalschemesare providedby Fund,Medisave,and Medishield allowsconsumers
companieson a scheduled to chooseextentof coverage.
reimbursement basisfor outpatient
andhospitalizationbenefits.Major
medicalnot common.
South Altemativeto employment- Traditionalmedicalschemes (sickness 16percentof Limitedgovemmentsystem.Expenditures are 45
Africa basedsicknessfunds;public funds)andinsuredplans,mostof which population(69 percentpublic,34 percentmedicalschemes,3
coverageisnot widespread requireemployeeand employer percentof percentinsurance,and 14percentout-of-pocket.
contributions
(evenlysplit).AJImedical whites;7 percent
plansunderseverecost pressures. of blacks)
SriLanka Altemativeto public Employer-provided medicalinsurance 2 percent Low-qualitypublichospitals.About47 percentof
coverage for executives;
somecontributory the populationusesprivatehealthcareservices,
insurancefor others.Practicevaries and 60 percentof totalexpendituresare out of
amongthe five insurancecompanies, pocket.
activemostlyin Colombo.
Thailand Addibonalto public Commonfor companies to provide 0.7 percent In 1992expenditureswere 24 percentpublic
coverage medicalinsurancefor allemployeeson services,2 percentother publicinsurance,
and 74
a noncontributorybasis,inthe formof percentout of pocket.
scheduledindemnityprograms.

INNOVATIONS IN HEALTH CARE FINANCING

108
Annex table 2 Roles,characteristics,and extent of private health insurancecoverage
and main alternatives, non-OECD countries(continued)

Shareof
populationwith
Country Roleof privateinsurance Characteristics
of privateinsurance privateinsurance Mainalternative
sourceof financing
Turkey Additionalto compulsory Mostinsuranceplansare individual 500,000 Threesocialsecuritysystemsof variablequality.
socialinsurancefor 65 ratherthangroup policiesoffered policyholders Ministryof Healthcoversthe rest.Two-thirdsof
percentof eligible by 29 insurancecompanies. paymentare private.
population Multinationalcompaniesofferhealth
carebenefitsto Turkishmanagers and
someofficeworkersanddependents.
Uruguay Supplemental
insurance Sickness fundscover55 percentof the 5.9 percent Privatesicknessfundsfinancedefinedbenefits
populationthroughown andcontracted (48 percentof expenditures).
Expenditures also
facilities.35 supplementaryinsurersoffer coveredbysocialsecurnty,Ministryof PublicHealth
medical,surgical,dental,diagnostic,and (16 percent), andout of pocket( 14percent).
mobile/emergency plans.
Venezuela Additionalto public Employer-contractedmedicalinsurance Low PublicsocialsecurityinstituteandMinistryof Health.
coverage commonin officeswith 50 or more
employees.Typically
contributoryand
includesdependents.
Zimbabwe Altemativeto public Privatemutualmedicalorganizations 5 percent Poor-quality
state-runsystemfor low-income
clinics with reasonable
levelof benefits. people.
Employersandemployeestypically
splitcosts.Membershipcommon
for middleandupperclasses.
Source:
Bhat1996;California
Consortium
1996;
Femandez 1997;Fuenzalida-Puelma
1996;LaForgia
andCouttolenc1993;LaForgia
anrdGriffin1993;
Lewis1988;Lewisand
Medici1995;
WilliamM.MercerLmited1996;Nittayaramphong
andTangcharoensathien
1994;Rafeh inthisvolume;
Shaw
andGriffin1995;
TAI 1997;WorldBank1992,1994,
1995,1996,
andforthcoming;WVorld
BankandIDB1997; Yang1996.

PRIVATE INSURANCE: PRINCIPLES AND PRACTICE

Notes adequatelypoolrisk.Insurersthat experienceanunderwritingspi-
ral in a particular insurance product typicallywill cancel that
1. In competitivemarkets the price of insurancecoverageis product if they are unable to stabilizeexperienceby redesigning
equalto the expectedvalueof loss,plus administrativeand mar- the product to retainlowrisks.
ketingexpense,plus an additionalfee (calleda loading).In gen- 7. Over time this behaviorwill cause insurancepools to dis-
eral, insurance loadingsare larger as consumersare more risk solveand formin waysthat reduce differencesin the riskembod-
averse (that is, as consumersare willingto pay more to avoid ied in any one pool.Calledseparating equilibrium(Rothschildand
risk).As the probabilityof loss approaches1 (that is, certainty), Stiglitz1976),the propensityof consumerstoseparatethemselves
the cost of insurancemay exceedconsumers'cost of sustaining into relativelyhomogeneousrisk pools is a well-recognizedphe-
the loss,even if consumersare risk averse.Consumersare more nomenon,especiallyin competitiveinsurancemarkets.
likelyto insureagainsthigh-probability lossesand high-magnitude 8. Tieredrating helps stabilizeinsurancemarkets as follows.
lossesincompetitivemarkets(wherethe priceof insuranceequals Becauseconsumersgravitatetowardriskpoolsthat includemem-
the marginalcost of producinginsurance)than in monopolistic bers whoseriskis similarto or lessthan their own, insurersthat
markets (wherethe priceof insuranceexceedsmarginalcost). attemptto aggregatedissimilarrisk (andset a singlepriceregard-
2. For high-riskpopulationsthe priceof insurancereflectsthe lessof differencesin individuals'risk)willsee an exodusof low-
highprobabilityof highloss.However,the pricemayalsoinclude risk participantsfromthe pool. By aggregatepricingonlysimilar
a deterrentsurcharge,reflectingthe absenceof a competitivemar- risk and pricinginsuranceaccuratelyat the outset, insurershope
ket for these risks. to avoidan underwritingspiral(seenote 6).
3. In the professionalliteratureconsumers'greaterinforma- 9. Manyinsuranceplansexcludecoveragefor any procedure
tion about riskis calledasymmetric information.Ligonand Thistle thatis deemedexperimental.In the United Statesinsuranceexdu-
(1996)provide a technicalpresentationof possiblecontractual sionsfor experimentalproceduresarecontroversial;manypatients
remediesto asymmetricinformationproblemsin insurancemar- (and their physicians)feel that the transitionfrom experimental
kets, althoughnot specifically with regardto health insurance. procedureto acceptedpracticeis a matter of professionaljudge-
4. In the extremecaseconsumerswouldbuy healthinsurance ment.No singleagency,publicor private,judgeswhichprocedures
not to managerisk, but to gain a subsidyfor perfectlyanticipated are experimentaland whichare not. Whilethe insurancecontract
medical expenditures.The magnitudeof the subsidyto a con- mayidentifyspecificproceduresthat it deems experimentaland
sumer with asymmetricinformationis positivelyrelated to the willnot cover,in mostcaseseachinsurer'smedicaldirectormakes
differencebetweenthe anticipatedexpectedvalueof personalloss a decisionabout whethera specificprocedureis deemed experi-
and the "pure"premium-that is, the averageexpectedvalue of mental at the timethat the insuredpatient requirescare.
lossamongthe insuredpopulation.Consumerdemandfor insur- 10. In the broadestsense,the rapid adoptionof new and costly
anceinmarketswithasymmetricinformationgenerallyis assumed medicaltechnologythat occurswhen most consumersare insured
to be greater as the opportunity for such a subsidy increases. (despiteimprovementsin the qualityof care)alsomaybe viewed
Suchconsumerswouldincludepeoplewho haveeitheran unusu- asmoralhazard.Thusprivateinsuranceinmanycountriesmaynot
allyhighprobabilityof illnessor evena slightlyhigherprobability coveradvancedformsof medicaltechnology. Forexample,inKorea
of a verycostlyillness. privateinsurancetypicallyexcludesexpensivehigh-technology med-
5. An insurancepool (or riskpool)is a collectionof individual ical servicessuch as computerizedaxialtomography(CAT)scan-
risksthat correspondsto a specificprice for an insuranceprod- ning, magnetic resonance imaging (MRI), positron emission
uct.Each insurermayforn manypools,each of whichis defined tomography(PET)scanning,andsomechemotherapy(Yang1996).
by the specificinsurancecontractand the groupor classof insured 11.This sourceof moralhazardis generallyconsideredindis-
risksoverwhichthe price is calculated.A risk poolmay include tinguishablefromthe phenomenonof supplier-induceddemand:
eitherunrelatedindividualsor groups. the propensityof healthcare providersto recommendmore care
6. The logicof this is as follows:If an insurersets the price of when consumersare insensitiveto price. Of course, full insur-
coveragetoo low relativeto the plan's medicallosses,it will be ance is the mainreasonfor consumers'priceinsensitivity
forcedto raisethe price at the firstopportunity.But byraisingthe 12. Criticsof managedcarearguethat managedcareplansalso
pricethe insurerwilllose low-riskenrolleesand retainhigh-risk candiscouragesickpeoplefromenrollingin anynumberofways-
enrollees,because the new priceis highrelativeto low-riskcon- for example,by establishinga relativesmallnetworkof specialty
sumers'expectedvalue of loss.When that happens,the insurer providers,by excludingprimary care providerswho have estab-
mayenter an underwriting spiral(sometimescalleda deathspiral), lishedcaseloadsof sickpeople,and bylimitingcoveragefor high-
a repeatingsequenceof setting higherinsuranceprices and los- cost servicessuch as mentalhealth care.
ing low-riskenrollees.The ultimateconsequenceof an under- Indeed,reviewingthe researchliteratureon selectionbias in
writingspiralis the loss of the insurancearrangement'sabilityto U.S.managedcare plans,Hellinger(1995)concludesthat health

INNOVATIONS IN HEALTH CARE FINANCING

110
plansthatrestrictthe choiceof enrolleesto a specificset ofproviders
(that is, health maintenanceorganizationsof alltypes and exclu-
siveproviderorganizations)often do experiencefavorableselec-
tion.Thisresultholdsforboththe elderlyandnonelderlypopulation
enrollingin managedcare plans. However,overtime the selec-
tion advantageexperiencedbyHMOs tendsto declineasenrollee
utilizationregressestowardthe mean (Welch1985).
13.Extensiveresearchonthe costsof managedcareplanscon-
sistently concludesthat the main reason for the lower cost of
managedcare relativeto financialinsurance plans is the lower
rate of hospitalization,controllingfor patientdiagnosisand other
relevantcharacteristics.
14. In addition,large employersusuallyemployprofessional
benefits managerswho are accessibleto the insurerand able to
promptly resolvequestionsabout enrollment and claims. U.S.
insurersconsiderthe abilityto communicateefficientlywithlarge-
groupcustomersto be an importantaspectof the loweradminis-
trativecost of that business.
15.The reasonsforthis arethe inverseofthosethat makelarge-
groupcustomersattractive.Specifically, marketingto manysmall
groupsis more expensivethan marketingto fewerlargegroups,
and may not differ appreciablyfrom marketingto individuals.
Moreover,smallgroupsare much lesslikelythan largegroupsto
be ableto afforda full-timeprofessionalplanadministrator,mak-
ingthe resolutionof questionsand disputesaboutcoveragemore
difficultandtime-consuming for the insurer.Becauseofthe greater
risk and cost of insuringsmallgroups,insurers typicallyunder-
write smallgroupsverycarefullyto avoidadverseselection.
16. For at least two reasons,insurersmay redlineworkersin
some occupationsor industriesevenif those workershave sepa-
rate insuranceto pay for job-related injuriesor illnesses.First,
suchworkersaremorelikelytoposedifficult(andthereforecostly)
problemsfor claimadjudication:everyclaimwould have to be
examinedasbeingpotentiallyrelated totheiremployrment. Second,
workerswho sufferhealthproblemsfrom past exposureor injury
cannot easilybring a claimagainstpast insurersof job-relatedill-
ness.Instead,theywillbringthe claimagainsttheir currenthealth
insurer.Thusinsurersmayredlineto protectthemselvesfrominad-
equaciesin other insurancesectors(specifically, insurancefor job-
relatedinjuries)and from adverseselectionby workerswho are
awareof past exposuresor injuriesbut mayconcealthem froman
insurer.
17.In the UnitedStatesthe bestmanagedcareplansareincreas-
inglyactivein developingand usingmedicalpracticeprotocolsto
improvethe efficiencyand qualityof care. However,the devel-
opment and implementationof practice protocolscan generate
significantadministrativecosts. In general,financialinsurance
plans are muchless inclinedto investin measuringand improv-
ing the qualityof care, sincetheydo not affiliatewith anypartic-
ular set of health care providers.
18. In the United Statesand many other countriesindividu.
als and businessesarenot prohibitedfrom buyinginsurancefrom
PRIVATEINSURANCE:PRINCIPLESANDPRACTICE
111
an unlicensedinsurer.However,when they do buyfrom an unli-
censedinsurer,theyare not protectedby regulatorystandards.
19. Most U.S. statesallowonlycertainclassesof assetsto be
admittedto the balancesheet,andmanyrestrictthe shareofassets
that insurerscan holdin certaincategories.Forexample,in most
statescommonstockcannot accountfor more than 10percentof
alifeinsurer'sassets.Somestatesalsolimitinvestmentsin medium-
to lower-gradebonds (Skipper1992).
20. In the UnitedStatesthe NationalAssociationof Insurance
Commissioners-avoluntaryassociationthat advisesstatesabout
regulatorystandards and practices-is developingrisk-based
capitalmeasuresand standards.Risk-basedstandardswouldvary
minimumlevelsof insurercapitaland surplusby the financialrisk
of the instrumentsin whichthese funds are investedand by the
insurer'soperationalcharacteristics.Thisinitiativeis expectedto
encouragegreateruniformityof capitaland surplusrequirements
amongstates.
21. SomeU.S.statesdo not requirethat reinsurersbe licensed,
sincedirectinsurersareconsideredsophisticatedbuyers.However,
uncollectiblereinsurancehas beena sourceof somedirectinsurer
insolvencies.Somestates indirectlyregulatereinsuranceopera-
tionsby not allowingdirectinsurersreservecreditfor reinsurance
placedwithunauthorizedreinsurers.
22. Suchactionsgenerallyrequirea courtorder.In somestates
the regulatormay assume control of an insurerwithout a prior
courtorderifthe regulatordecidesthatimmediateactionis needed
to preserveassets.However,such "quicktake" actions are sub-
ject to subsequentjudicialreview(Skipper1992).In somecases
where an insuranceregulatorhas assumedcontrol of an insurer,
the regulatorhas discoveredillegalbehaviorby membersof the
insurer'smanagementor governingboard, and has initiatedcivil
or criminalprosecution.
23. A holdingcompanyis a parent corporationthat typically
ownsall or most of the stock of a number of constituent sub-
sidiaries,but occasionallymayownlessthan a majoritysharebased
on control or investmentmotives (Munn, Garcia, and Woelfel
1991).
24. U.S. state insuranceregulatorscommonlyrequire that an
affiliatedinsurerconstruct a corporate"firewall"-that is, a pro-
visionofincorporationthatformallyprotectstheinsurer'sfinances
frominsolventaffiliates.
25. Each U.S. state operates two guaranteefunds (by con-
vention,for life and healthinsurersand for propertyand casualty
insurers) to compensatepolicyholdersfor financialloss due to
insurerinsolvency. As an artifactof their historicaldevelopment,
healthinsurersmaybe cdassified in eithergroup.Differencesamong
statesin the amountsguaranteed,personseligiblefor protection,
extent of coverage,premiumbasefor assessment,capson assess-
ments, and rightsto recoupguaranteepaymentscan lead to very
uneven economicresults and drawn-out litigation over large
amountsof moneywhen a multistateinsurerbecomesinsolvent
(Skipper 1992).
 26. In the United States the National Association of Insurance
Commissioners maintains two electronic databases to help states
in this effort. The Regulatory Information Retrieval System con-
tains information on people and companies against whom regu-
latory actions have been taken. The Special Activities Database
facilitates the confidential exchange of information among regu-
lators, allowing them to inquire into the activities of companies
and individuals of regulatory concern, including people who may
be involved in fraudulent activities.
27. Every U.S. insurer is required to retain an outside firm of
certified public accountants that is accountable to the insurer's gov-
erning board. Regulators sometimes have the right to approve the
choice,and insurerstypicallyretain a largenational accounting firm.
The accounting firm periodically audits the insurer (reviewingits
fnancial statements) and presents the audited financial statements
to the insurer's board with an opinion on whether the insurer's
financialstatements conform to generallyaccepted accounting prin-
ciples as well as statutory accounting principles. The accounting
firm usually also submits a management letter to the board-in
effect, a "report card" on the insurer's fnancial management.
This system is not perfect, however. Outside accountants are
not alwaysthorough, and insurers can hide financial distortions.
The accounting firm that conducts the audit may be friendly to
the insurer's management, especially if the insurer also retains
them for consulting servicesthat generate more revenue than the
audit. While the board should expect the accounting firm to be
responsive to management in operational matters, but commit-
ted to the board in evaluating fiscal condition and management,
striking that balance can be difficult (Akula 1997).
28. De Sa (1996) summarizes the current state of quality mea-
surement and prospects for quality competition among managed
care plans.
29. No state regulates underwriting or pricing in the large-
group market (typicallydefined as groups of 50 to 100 or more).
In general, this market functions very differently than small-
group and individual markets. Specifically,insurers "experience
rate" large-group clients; experience rating allows them to assume
little or no risk and to negotiate a fee to cover administrative cost
and profit.
30. For example, smaller insurers may offer insurance agents
incentives (such as salary or fee bonuses) to screen informaUyfor
poor health status and to send applicants with health problems to
another insurer. Even if such practices are illegal,they are extremely
difficult for regulators to detect and document.
31. Reinsurance is a set of products sold to insurers to help them
finance unusually high risk. It may be written as aggregate rein-
sua ,(pyn all clim afe th inue exed a thehl of
surance (paying all dlains after the insurer exceeds a threshold Of
total liability)or as specificreinsurance (payingallclaims per enrollee
above a threshold amount). Reinsurancecompanies pool risk across
insurers, but all of the basic principles of insurance apply in rein-
surance markets. Specifically,if reinsurers believe that all or many
insurers in a market are likely to exceed the reinsurance threshold,
INNOVATIONS IN HEALTH
112
risk is nonrandom, and reinsurers will be reluctant to do business
in those markets. Such a situation may arise if the introduction of
health insurance is likelyto cause a sudden accelerationin providers'
charges-eitherbecause more care isdelivered orbecause providers
are able to shift costs (that is, to charge privately insured patients
more than publicly insured or uninsured patients for the same ser-
vice). Because managed care companies typically pay negotiated
rates to hospitals and specialtyproviders and may capitate primary
health care services, they may be more attractive to reinsurers in
developing markets than financial insurance plans that have fewer
options for controlling plan costs.
32. During the past decade manyU.S. states have enacted laws
requiring that health care providers prove their servicesare a lower-
cost alternative to more conventional medical treatment before
the legislature will consider proposals to require that their ser-
vices be covered. Few states that have enacted such a require-
ment have subsequently enacted any laws requiring coverage for
specific services or providers.
33. Variants of these controls are obviously valid in some set-
tings-for example, space requirements for technology to pro-
tect health workers and patients and minimal standards for specific
providers (for example, requiring that providers have the neces-
saryequipment before government finances those services),among
others.
34. Other elements of the health system-controlling hospi-
tal construction, licensing health-related businesses, medical prac-
tice, consumer protection-are much less common, not enforced,
and/or counterproductive. Among these the most common are
licensing of health business (as in Bangladesh, Hungary, Jordan,
Philippines, Thailand, Turkey,and Uganda). South Africa licenses
new hospitals, but responsibility for enforcement is at both the
federal and province levels, allowing enforcement to fall between
the cracks. The same is true in India. Regulations on space allo-
cations are often enforced in Brazil and the Czech Republic, but
have little or no impact on quality, effectiveness, or efficiency of
health service delivery in either the public or private sector (TAI
1996; Bhat 1996).
References
Abel-Smith, Brian. 1995. 'Assessing the Experience of Health
Financing in the United Kngdom." In David W Dunlop and
Jo. M. Martins, eds., An International Assessment of Health
CareFinancing:Lessonsfor Developing Countries.Washington,
D.C.: World Bank Economic Development Institute.
Akula John L. 1997. "Insolvency Risk in Health Carriers:
kl,Jh .197 Islec ls elhCres
Innovation, Competition and Public Protection. " HealthAffairs
16(1): 9-33.
Altman, Stuart, andTerrijackson. 1991. "Health Care in Australa:
Lessonsfrom Down Under." Health Affairs 10(3): 129-46.
American Academy of Actuaries. 1993. 'An Analysis of Mandated
Community Rating." Monograph 3. Washington, D.C.
CARE FINANCING
Behlil,Ali. 1995."PrivateHealth Sector Experiencein Turkey"
Workingpaper.Halk Yasam,Sigorta,Istanbul.
Bennett,Sara. 1992."Promotingthe PrivateSector:A Reviewof
DevelopingCountryTrends."HealthPolicyandPlanning7(2):
97-110.
Bennett, Sara, George Dakpallah, Paul Garner, Lucy Gilson,
SanguanNittayaramphong,BeatrizZurita,and AnthonyZwi.
1994."Carrotand Stick:StateMechanismstoInfluencePrivate
ProviderBehaviour."HealthPolicyandPlanning9(1): 1-13.
Bhat, Ramesh.1996. "Regulationof the PrivateHealth Sectorin
India."InternationalJournalofHealthPlanningandManagement Jensen, Gail A., and MichaelA. Morrisey 1990. "Group Health
11:253-74.
CaliforniaConsortium. 1996. "Options for Health Reform in
Ecuador."Paper preparedfor the Ministryof Health and the
NationalHealth Councilof Ecuador.SanFrancisco.
Chollet, Deborah J. 1995. "RedefiningPrivate Insurancein a
ChangingMarket Structure."In Stuart H. Altmanand Uwe
E. Reinhardt,eds.,StrategicChoicesfor
a ChangingHealthCare
System.Chicago,Ill.: Health AdministrationPress.
1996."TheChangingRateof Employer-Based Coverage."
Workingpaper.AlphaCenter,Washington,D.C.
Chollet,DeborahJ.,and RebeccaR.Paul.1993.Community Rating:
IssuesandExperience. Washington,D.C.: AlphaCenter.
Collins,David H., Jonathan D. Quick, StephenN. Musau,and
DanielL.Kraushaar.1996."HealthFinancingReforminKenya: Levit,KatharineR., Helen C. Lazenby,Lekha Sivarajan,Madie
The Fall and Rise of Cost-Sharing, 1989-94." Stubbs
MonographSeries1. ManagementSciencesfor Health,Boston,
Mass.
Day,Patricia, and RudolfKlein. 1991. "Britain's Health Care
Experiment."HealthAffairs10(3):39-59.
DeSa, Jeanne. 1996. The Marketfor Accountability:Measuring
and ManagingPlanPerformance.Washington,D.C.: Alpha
Center.
Dunlop,DavidW, andJo.M. Martins.1995."LessonsLearned."
In DavidW DunlopandJo. M. Martins,eds.,An International
Assessmentof Health CareFinancing:Lessonsfor Developing
Countries. Washington, D.C.: World Bank Economic
DevelopmentInstitute.
Enthoven,AlainC. 1991."IntermalMarketReformsof the British
NationalHealth Service."HealthAffairs10(3):60-70.
Fernandez,JesusM. .1997."Revisi6ndeEstudiosde Financiocion Londofio,JuanLuis.1996."ManagedCompetitionintheTropics?"
del Sistemade SaludDel Perui.ImplicacionesPara el Proceso
de Reforma del Sector." Consultant report. Overseas
DevelopmentAdministration,London.
FKMIS(FederationofKoreanMedicalInsuranceSocieties).1986.
"Outlineof MedicalInsurancein Korea."Seoul.
Fuenzalida-Puelma,Hernan L. 1996. "Regulationof Health
Insurance:ConceptualFramework,Case Studieson Hungary
and Turkey,Notes on Croatia and Lithuania."World Bank,
Human DevelopmentDepartment,Washington,D.C.
Gross,AlfredW, and CarlosZarabozo.1994."Reportsto USAID Munn, Glenn G., F.L. Garcia,and CharlesJ. Woelfel.1991.The
on January20-28, 1994visitsto the CzechRepublic."U.S.
Departmentof Health and Human Servicesand U.S.Agency
for InternationalDevelopment,Washington,D.C.
Hellinger,FredJ. 1995."SelectionBiasin HMOs and PPOs:A
Reviewof the Evidence."Inquiry32(2): 135-42.
Hurst,Jeremy.1991."ReformingHealth Carein SevenEuropean
Nations."HealthAffairs10(3): 7-21.
Jensen,GailA. 1993."Regulatingthe Content of Health Plans.'
In Robert B. Helms, ed., AmericanHealtb Policy:Critical
IssuesforReform.Washington,D.C.: AmericanEnterprise
Institute.
Insurance:A HedonicPrice Approach."Reviewof Economics
andStatistics72(1): 38-44.
Jensen,GailA., KevinD. Cotter,and MichaelA.Morrisey.1995.
"StateInsuranceRegulationand Employers'Decisionsto Self-
Insure."Journalof Risk andInsurance62(2).
LaForgia,GerardM.,and BernardCouttolenc.1993."Guatemala
Health Sector Reform Lending Program:A Reviewof the
EpiodemiologicContext, GovernmentResourceAUlocation
and Efficiency,and the PrivateSector."Reportpresented to
the Inter-American Development Bank. Urban Institute,
Washington,D.C.
La Forgia,Gerard,and CharlesGriffin.1993.HealthInsurancein
Practice:FifteenCaseStudies.Washington,
D.C.:UrbanInstitute.
W Stewart,and others. 1996. "Data View:National Health
Expenditures."HealthCareFinancing Reviewv17(3):205-42.
Lewis,MaureenA. 1988.Financing HealthCareinJamaica.Report
3714-04.Washington,D.C.: Urban Institute.
Lewis,MaureenA.,andAndreCesarMedici.1995."PrivatePayers
of HealthCareinBrazil:Characteristics,
Costs,and Coverage."
HealthPolicyandPlanning10(4):362-75.
Ligon, James A., and Paul D. Thistle. 1996. "Information
Asymmetriesand InformationalIncentivesin Monopolistic
InsuranceMarkets."TheJournalof Risk andInsurance63(3):
434-60.
Lohr,KatherineN., RobertH. Brooks,and others.1989."Use of
Medical Care in the Rand Health Insurance Experiment:
Diagnosis-and Service-SpecificAnalysesin a Randomized
ControlledTrial."MedicalCare24(9): S1-87.
Paper presented to the International Health Economics
AssociationInauguralConference, May,Vancouver,British
Columbia.
. 1997. Personal communication. Inter-American
DevelopmentBank, Washington,D.C.
McIsaac,DonaldA.,and DavidF Babbel.1995."TheWorldBank
Primeron ReinsurancePolicy."PolicyResearchWorking Paper
1512.WorldBank,FinancialSectorDevelopmentDepartment,
Washington,D.C.
St. JamesEncyclopedia of BankingandFinance.Chicago,Ill.

PRIVATE INSURANCE: PRINCIPLES AND PRACTICE

113
Musgrove, Philip. 1996. Publicand PrivateRolesin Health: Theory
and Financing Patterns.World Bank Discussion Paper 339.
Washington, D.C.
Nittayaramphong, Sanguan, and VirojTangcharoensathien. 1994.
"Thailand: Private Health Care Out of Control?" Health Policy
and Planning9(1): 31-40.
Oyarzo, Cesar. 1997. Personal communication. Director of the
faculty of economic sciences and administration, University
of Chile.
Pauly, Mark V 1986. "Taxation, Health Insurance, and Market
Failure in the Medical Economy."JournalofEconomicLiterature
24: 629-75.
Price, Max. 1994.'The Impact of PoliticalTransformationin South
Africa on Public/Private Mix Policy Debates." Health Policy
and Planning 9(1): 50-62.
Reinhardt, Uwe E. 1995. "Germany's Health Care and Health
Insurance." In David W Dunlop and Jo. M. Martins, eds., An
International Assessmentof Health Care Financing:Lessonsfor
DevelopingCountries.Washington,D.C.: World Bank Economic
Development Institute.
Rothschild, Michael, and Joseph E. Stiglitz. 1976. "Equilibrium
in Competitive Insurance Markets: An Essayon the Economics
of Imperfect Information." QuarterlyJournalof Economics 90:
626-49.
Schneider, Markus, Juerg Sommer, and Aynur Keceki. 1987.
Gesundheitssysteme im Internationalen Vergleich. Augsburg:
BASYSGmbH.
Shaw,R. Paul, and Charles C. Griffin. 1995. FinancingHealth Care
in Sub-SaharanAficathroughUserFeesandInsurance. A Directions
in Development book. Washington,D.C.: World Bank.
Skipper, Harold D. Jr. 1992. "Insurer Solvency Regulation in the
United States." ResearchReport 92-4.Georgia State University,
Center for Risk Management and Insurance Research, Atlanta,
GA.
TAI (TaylorAssociates International). 1997. 'Country Profiles in
Private Hospital Investment Opportunities: A Global Study
Conducted on Behalfof the International Finance Corporation."
Annapolis, Maryland.
INNOVATIONS
IN HEALTHCAREFINANCING
114
U.S. House of Representatives, Committee of Energy and
Commerce, Subcommittee of Oversight and Investigations.
1990. Failed Promises.Washington, D.C.: U.S. Government
Printing Office.
. 1994. Wishful Thinnking: A World View of Insurance
Solvency Regulation. Washington, D.C.: U.S. Government
Printing Office.
Welch, W Pete. 1985. "Regression Toward the Mean in Medical
Care Costs: Implications for Biased Selection in Health
Maintenance Organizations." Medical Care 23: 1234-41.
WllliamM. Mercer Ltd. 1996.InternationalBenefitGuidelines1996.
19th ed. Brussels.
World Bank. 1992. "Indonesia: Health Insurance Issues in the
1990s." Green Cover Report 9999-IND. Washington, D.C.
. 1994. "The Organization, Delivery, and Financing of
Health Care in Brazil: Agenda for the 90s." Report 12655-
BR. Latin America and the Caribbean Region, Human
Resources Division, Washington, D.C.
. 1995. "India: Policy and Finance Strategies for
Strengthening Primary Health Care Services." South Asia
Region, Country Department II, Population and Human
Resources Division, Washington, D.C.
. 1996. "Hashemite Kingdom of Jordan: Health Sector
Study" Report 15418-JO.Middle East and North AfricaRegion,
Country Department II, Human Resources Division,
Washington, D.C.
. Forthcoming. "Honduras National Health Accounts."
Preliminary report. Washington, D.C.
World Bank and IDB (Inter-American Development Bank). 1997.
"Argentina:Facingthe Challenge of Health Insurance Reform."
Gray cover report produced jointly by the World Bank's Latin
America and the Caribbean Region, Human and Social
Development Group, and the Inter-Arnerican Development
Bank's Regional Department I, Social Programs Division,
Washington, D.C.
Yang, Bong-Min. 1996. "The Role of Health Insurance in the
Growth of the Private Health Sector in Korea." International
Journal of Health Planningand Management 11: 231-52.
Private Health Insurance in Egypt
Nadwa Rafeh

T his paper reviewshealth care financingin Egypt,

particularlythe role of privateinsurance.It begins
That same year, all private voluntary associationswere
broughtunderthe regulationof the Ministryof SocialAffairs,
byprovidingan overviewofEgypt'shealthcaresys- and those providinghealthserviceswere requiredto regis-
tem, reviewingthe political and socioeconomicenviron- ter with the Ministryof Health. Under Nasser the govern-
ment, current health indicators, the epidemiological ment owned,operated, and financedhealth care facilities.
transition of diseaseover time, resource distribution,and In addition,twohospitalnetworksnationalizedbythe Nasser
equityin servicedelivery. regime were formed: the Health InsuranceOrganization
The paper then describesthe various sourcesof health and the Curative Care Organization.Until then Health
carefinancing-publicandprivate-that are in place,includ- InsuranceOrganizationhospitalshad been private hospi-
inggovernmentfinancing,publicfinancingand socialinsur- tals run by foreignersand CurativeCare Organizationhos-
anceschemes,and privatefinancing.Attentionthen focuses pitalshadbeenrun bythe CharitableAssociationof Modem
on the currentstatus of the privatehealthinsurancesector, Women(Kemprecos1993).
analyzingits strengthsand weaknessesfrom the perspec- In 1973 Egypt entered a new phase of economicand
tive of society,consumers,and insurers. politicaldevelopmentunder President Sadat, who initi-
Finally,the paperconcludesbyreviewingthe factorsthat atedpoliciesaimedatincreasingeconomicgrowthby encour-
willinfluencethe developmentof the private healthinsur- agingforeigninvestmentand private sector development.
ancemarket in Egyptand describingthe publicsector'srole Thesepolicies,continuedtodayunder PresidentMubarak,
in providinggroup coverage. havehad a majorimpacton the provisionand financingof
health care services.In the past twentyyears the number
Egypt'sHealth Care System of privatemedicalfacilitieshasincreaseddramatically.
Today
a significantportionof health careis deliveredthroughpri-
Egypt'sMinistryof Healthwas foundedin 1936.The min- vate hospitalsand clinics.
istryquicklybecameresponsibleforawidevarietyof health Since 1986, under the guidanceof the Intemational
servicesand began developingnew services.The firstnew MonetaryFund and the WorldBank,Egypthas beenimple-
servicewas a health insuranceprogramfor schoolchildren mentingsignificant economicreform.However,reformsaimed
in Cairo (Kemprecos1993).Duringthis period significant at endingpublicsectordominanceof the economyare pro-
portionsof healthcare servicesin Egyptwere providedby ceedingslowly.There is littleconfidencein govemrent pri-
privatevoluntaryassociations(gami'yat). vatization efforts, and legislationto support and expand
In 1962PresidentNasser's NationalCharter declared theseeffortshas not been welldevelopedor implemented.
medicalcare,education,employment,minimumwages,and The governmentcontinuesto playa major role in pub-
healthinsurancebenefitsto be basic rightsfor all citizens. lic health through a wide range of programs and special-

NadwaRafehis a healthservicesmanagement
andpolicyconsultantat theWorldBank.

115
ized centers.The Ministryof Health's strategyemphasizes circulatory,respiratory,andinfectiousandparasiticdiseases.
prevention,primaryhealth care, drug manufacturing,free By 1990 circulatorydiseasesaccounted for 42 percent of
care for the indigent, and environmental protection. alldeaths amongmen and 44 percent amongwomen,and
Successfulattemptshavealsobeenmadeto improvehealth the proportionof respiratoryand infectiousand parasitic
indicatorsthroughmaternalandchildhealthprograms,pop- deathshad decreased.
ulationandfamilyplanningprograms,vaccinationprograms, Althoughinfectiousdiseasesare no longerthe leading
laboratoriesand blood banks, and control of infectious cause of morbidityand mortality,there is still a great deal
diseases. of workto be done in the preventionof infectiousdiseases
such as hepatitis, trachoma, and schistosomiasis.
Basichealthandepidemiological profile Epidemiologicaldata suggestthat there are areas where
modestinvestmentin health care can significantlyreduce
Historically,infectiousdiseaseswerethe maincauseof sick- infectiousdiseaserates. Populationgrowth willcontinue
nessandprematuredeathin Egypt.Government-sponsored to exertpressure on all aspectsof the economyfor yearsto
effortsto control infectiousdiseasehavebeen successful, come. In 1990more than 39 percent of the populationin
however.The prevalenceof schistosomiasis,a majorhealth Egyptwas under the age of 15 (WorldBank 1992).With
problem,dropped from 36 percent in 1981to 10percent the successfulcontrolof infectiousdisease,the population
in 1991 (MOH-IDC 1993). The incidence of neonatal is livinglongerand chronicdiseasesare becomingthe main
tetanus alsohas dropped,from 20.7percent in 1986to 93 contributorsto morbidityand mortality.
percent in 1990,reflectingsignificantgovernmentefforts
throughthe Child SurvivalProject. Hospitalandproviderprofile
In addition,since1980the governmenthas implemented
a seriesof familyplanning and child and maternalhealth Duringthe 1980sthe governmentbegan allowingthe estab-
careprogramsthat havehelpedimprovehealthindicators. lishmentofprivatehospitals,leadingto a significantincrease
The crude birthrate fell from37 per 1,000people in 1981 in the number of beds in the private sector.Between 1975
to 28 per 1,000in 1993.Infant mortalityfellfrom70 deaths and 1990the number of hospitalbeds in Egyptrose by 60
per 1,0001ivebirthsin 1981to 38 per 1,000in 1990(MOH- percent,to more than 110,000beds.Duringthe sameperiod
IDC 1994).Vaccinationcoverageis more than 80 percent the number of private beds increasedby 180 percent, to
foreveryantigen,and the percentageof childrenfullyvac- about 11,000(Boutros1992).Almosthalf the privatehos-
cinatedhas risen to 75 percent, from 58 percent in 1987. pital beds are in Cairo.
Coveragefor tetanus toxoid jumped from 12 percent in Hospitaloccupancyratesare generallylow,witha national
1987to 63 percent in 1992. occupancyrate of 49 percent.Althoughthere are no accu-
MOH-IDC (1994) indicatesthat 28 percent of hospi- rate data on occupancyrates in privatehospitals,evidence
talizedmen were admitted as a resultof accidents,poison, suggeststhat occupancyratesrangebetween60 and70 per-
or violence.An additional21 percent had diseasesof the cent, and many of these hospitalsare strugglingto main-
digestivetract, 14 percent had respiratorydiseases,7 per- tain profitability(Kemprecos1993).Publichospitalshave
cent had circulatorydiseases,and 6 percenthad genitouri- occupancyrates as lowas 40 percent.
nary tract diseases. Women were mainly admitted for There are many private clinicsthroughout the country,
childbearing, with obstetrical-related hospitalizations particularlyin rural areas. Many of these clinicsprovide
accounting for 35 percent of female admissions.Other limnitedinpatient servicesfor recoveryafter minor proce-
causesof femalehospitaladmissioninduded digestivedis- dures. Many clinicsare attached to mosques, churches,
eases,accidents,andrespiratoryandgenitourinarydiseases. and charitableorganizationsand provide a wide range of
Mortalitydata alsoshowthat an epiderniologicaltransi- outpatientservices.
tion has been taking place over the past decade. In 1982 There are 19.6physicians,2.5dentists,5.6 pharmacists,
the distributionof deathswasfairlyevenlydistributedamong and 19.6 nurses per 10,000people in Egypt. About half

IN HEALTHCAREFINANCING
INNOVATIONS
116
the physiciansare employed by the Ministry of Health. hold (out-of-pocket)expenditureson health care, which
The governmentpolicyguaranteeinga job for eachphysi- accountfor55 percent of spending.Another33 percent is
cian upon graduationhas led to overstaffingof physicians financedby government ministries, 9 percent by public
withinthe ministry. financing (mainly through the Health Insurance
Organization),and 3 percent by private firms,insurance
PublicSourcesof Financing companiesandunions,andprofessionalorganizations.Out-
for Health Services of-pocketexpendituresmainlycoveroutpatientcare.
The averagehouseholdspendsLE 380.5 (about $113)
In 1991Egypt spent about 4.7 percent of GDP on health a yearon outpatient care, comparedwith LE 35.4 (about
care.Althoughthe healthcaresystemis predominantlypub- $9) on inpatient care (table2). Drugsaccount for 53 per-
lic, severaldifferent government,nonprofit, and private cent of outpatient expenditures.Thus, of the averageLE
organizationsprovide and finance health care. There are 410 spent on health care each year,92 percent is spent on
four main financingmechanisms: outpatientcare and 8 percent on inpatient care.Per capita
* Governmentftnancing-direct paymentsmadebythe gov- expendituresin urban areas (LE 106)are almosttwiceas
emnmentfor health care. The Ministryof Health is the high as those in ruralareas (LE 59).
maingovernmentagencyfundinghealthcare.Othermin-
istriesthat own and operate healthfacilitiesincludethe MinistryofHealth
MinistriesofEducation,Defense,Interior,Transportation,
and SocialAffairs.The Ministryof Educationplaysan The government guaranteesall citizens the right to free
importantrole in financingmedicaleducationand uni- health care through a networkof 225 hospitalsand 2,000
versityhospitals,thus funding a significantportion of clinicsoperatedby the Ministryof Health.The freehealth
tertiarycare. care policyservesas a safetynet for a largesegmentof the
* Publicfinancing-including socialinsurance(suchas the population, mainlylow-incomegroups.The government
Health Insurance Organization)that providescare to also provides free medical and nursing education and,
selected groups as well as the Teaching Hospital through its employmentpolicy,guaranteesjobs in Ministry
Organization,CurativeCare Organization,and other of Health facilitiesto allgraduatingphysicians.
public firms.These organizationshave severalsources
of funding, including revenues, premiums paid by TABLE
I
enrollees
and government contributions. The Healt Distribution
of healthcareexpenditures
in Egypt
enroJIees,and governmentcontributions.The Health bysueofiacn,19
InsuranceOrganizationis the largest sourceof public
financing,providingcare to public and private sector of
Millions
employees. Institution Egyptian
pounds Share
(%)
* Privatefnancing-including privatelyowned organiza- Government
Ministryof Finance 182 4
tions,privateinsurancecompanies,unions,cooperatives Ministryof Health 782 19
and professional organizations, and nonprofit non- Ministry
of Education 270 7
governmentalorganizations(NGOs).NGOs areone the Oter ministries 107 3
fastest-growingsectorsin Egypt.Fundingforthese orga- Public 370 9
nizationsisprovidedby nationalandintemationaldonors, Prvate
mosquesand churches,and individuals.NGOs are con- FIm 70 2
sideredmore cost-effectivethan public providersand Private
insurance/unions 30 1
providehigher-qualityservicesfor the charges. Total 100 3
* Householdpaymentsthroughdirect payments. Household
poyments 2,263 55
The distributionof expenditureson healthcareis shown Total 4,115 100
in table1.Mosthealthcarefinancingis throughdirecthouse- Source:
World Bankdata.

PRIVATE HEALTH INSURANCE IN EGYPT

117
 In principle,ministryfacilitiesprovide comprehensive About 60 percent of the ministry'sbudget is allocated
coverage,includingemergencycare. In practice, the care to ministryoperations,26 percent to the Health Insurance
providedin ministry-runhospitalsis limitedin volumeand Organization, and 8 percent to the Curative Care
qualitydue to budget constraints.The ministryalsosubsi- Organization. Ministry support for Curative Care
dizescostlytertiarycareto indigentpatients,includingopen- Organizationoperationsis in the form of loans and inter-
heart surgery,renal dialysis,and treatment of malignancy. est payments.Additionalsupportis providedforoperating
The ministryisthe mainproviderand financierof health andcapitalexpenses.Theministrydoes notsupportCurative
care.Almosttwo-thirds(62percent)of the hospitalsin Egypt Care Organizationsalaries.
are run by the ministry.In urban areas,wheremost private, About 42 percent of the ministry'sbudget is allocated
university,teaching,and CurativeCare Organizationhos- to curativecare,of whichabout half is allocatedto hospi-
pitals are concentrated,ministryfacilitiesaccountfor 25 tals. Curative care is defined as treatment of acutelyill
percentof hospitals.In ruralareassuchas UpperEgyptand patients,includingpregnancyand childbirth,on an outpa-
the Sinai, ministry facilities account for 83 percent of tient or inpatientbasis.Most of the curativecare budget is
hospitals. allocated to urban areas such as Cairo and Alexandria.
Throughthe variousministries,the governmentaccounts Primarycare receivesabout 37 percent of the budget, and
for 33 percent of the country's annualhealth care expen- preventivecare only 8 percent.Almostthree-quarters(71
ditures (see table 1). The largest portion is spent by the percent) of the primarycare budget is allocatedto rural
Ministryof Health (19 percent of the total). In 1993the areas. Most of this moneyis spent on rural health center
ministry'sbudgetwasaboutLE 1 billion(MOH-ILDC 1993). operationsand constructionor renovationof rural health
The ministry'sbudgetincludesgovernmentallocations,rev- care units.
enuesgeneratedbyministryfacilities,and grantsfromdonor The government's"free health care for all" policy has
agencies.The budgethas showna steadynominalincrease significantimplicationsforthe deliveryofhealthcarethrough
for eachofthe pastfiveyears.Theseincreaseshavenot kept government-ownedfacilities.Budgetlimitations,a rapidly
pace with inflation,however,soin constant terrnsthe min- growingpopulation, the inabilityto charge fees for ser-
istry'sbudgethas beendecreasing.Thebudgethas alsobeen vices, and the policy guaranteeingjobs for all graduated
decreasingas a share of governmentspending;duringthis physiciansare among the biggest constraints facing the
period the ministry'sbudget was about 1.8percent of the Ministryof Health.Moreover,the ministrycannotprovide
overallgovernment budget. Nearly two-thirds-65 per- the comprehensivehealthcare coverageavailablethrough
cent-of the ministry'sbudget is used to pay salaries,21 privateinsurersandprivateproviders.Theministryalsocan-
percentis used for operatingcosts,and 14percentfor cap- not compete with the private sectorin terms of qualityof
ital costs. care and patient satisfaction.And becausethe ministryis
poorlystaffed and poorlyfunded, it is the provider of last
TABLE
2 resort forpeople whocannot affordto purchasecare from
Annualhousehold expenditures
for inpatientand other sources.
outpatientcarein Egypt
Inpatient Outpatient HealthInsurance
Organization
EgyptianShareof EgyptianShareof
poundstotal(%) pounds total(%) The Health InsuranceOrganization(HIO) is the largest
Doctorfees 20.0 32 76.2 20 insurance organizationin Egypt providinghealth insur-
Drugs 7.2 28 101.2 27
Other 8.2 40 203.1 53 anceto a definedbeneficiarypopulation.HIO is a manda-
Averagehousehold tory socialinsuranceprogram;participationis required by
expenditure 35.4 100 380.5 100 allcompanyemployees.Article32 of Egyptianlawrequires
Annualnumberofvisits 0.034 4.62 the participation of governmentworkers and Article 79
Source: fromBerman
Adapted 1995. requiresthe participation of privateworkersin MHO's
health

INNOVATIONS
IN HEALTHCAREFINANCING
118
insuranceprogram.1HO coverseligibleemployees,widows
of deceasedbeneficiaries,and pensioners.It does not cover
spouses,children,or other familymembersof employees.
In 1993,however,the programwasextendedto coverabout
10millionschoolchildren.Todaythe programcoversmore
than 15 millionbeneficiaries,almost a fourth of Egypt's
population.
HIO operates twenty-fivehospitalscontaining about
4,500 beds as well as 116 outpatient clinics.Overallbed
occupancyin 11O hospitalsis about 69 percent, and the
averagelength of stayis 5.9 days.FIO staffssmallaid sta-
tions with one doctor and one nurse at work sites with
more than 3,000 employees.
RevenueforHO is providedbyemployeeand employer
contributionsand governmentsubsidies. HIO is predom-
inantly an employment-basedinsurance program, with
employersand employeespayinga portionof salaryas pre-
mniums.Under Article32 governmentemployeespay 0.5
percent of their base salaryand their agencyor ministry
employerpays1.5percent to -H1O. In addition,governmnent
employeesmust make smallcopayments.The copayments
are quitesmalland donot discourageinappropriateor excess
use of RIO services.
Private sector workers coveredunder Article79 must
pay1.0percentof theirbasesalaryto RIOandtheiremployer
must contribute 3.0 percent. Because of the higher pre-
mium,no copaymnents are requiredfor privatesectorwork-
ers. Pensionersandwidowsare requiredto pay 1.0percent
of their basic pension as a health insurance premium.
Pensionersand widowsare the fastest-growingsegmentof
IO beneficiaries,increasingby 15.5percentbetween1991
and 1992.Other beneficiarygroupsincreasedby 4-6 per-
cent during the same period. Pensionersand widowsare
alsothe most frequent users of healthcare services,posing
an ever increasingfinancialburden on the 11O program.
Since 1995the Student MedicalInsuranceProgramrhas
been another source of funding for HIO. Smallannual
subscriptionsare paid by students at everylevelof educa-
tion as part of their tuition.
In 1984the governmentpermittedemployersto request
waiversfrom RIO participationif their employerprovides
similarhealthinsurancecoverageto allemployees.However,
this lawrequiresemployersto continueto payHIO a pre-
miumequalto 1.0percent of each employee'sbasic salary.
PRIvATE HEALTH INSURANCE IN EGYPT
119
By 1993,561 companieshad receivedwaiversto opt out
of HIO. The characteristicsof these companiesand their
reasons.fornot participatingin HIO have not been ana-
lyzed.Anecdotalevidencesuggeststhat companiesrequest-
ingwaiversdoso becauseof widespreaddissatisfactionwith
the quality of health care provided under HIO. HIO is
alsocriticizedformakingit too easyfor employeesto take
sickleave.
Mostcompaniesoptingout ofHIO becomeself-insured,
providinghealth care coveragethroughcontractswith the
CurativeCare Organization,private hospitals,and health
careproviders.Othercompaniespurchasegrouphealthcare
coveragethroughprivateinsurancecompanies.Mostemploy-
ers that opt out of HIO are in the private sector and are
financiallysound.Similarly, HIO memberswhocan afford
higherpaymentsoften chooseto payout of pocketfor ser-
vices or to buy alternative private health insurance.
Individualschoosingnot to use HIO servicesare morelikely
to be eligiblefor health insuranceprovidedthrough pro-
fessionalorganizationsor cooperativeprivate voluntary
organizations.
RIO providesa comprehensivepackageof health care
benefits. Currently, however,there is a substantial gap
betweenthe premiumspaid to HIO andthe costsincurred
by the program.In 19931IO experienceda net operating
lossof LE 14.9 millionafter accountingfor all premiums
paid and all governmentsubsidies.There are severalrea-
sonsEIO has been unable to operate profitably.
First, healthcare coveragethroughHIO is providedto
alleligiblebeneficiariesregardlessof preexistingconditions
or other high-riskcharacteristics.Denyingcoveragebased
on health status and requiringhigherpremiumsfor high-
risk groups is contraryto the government'sgoals for the
HIO social insuranceprogram: HIO's inabilityto apply
standard underwritingpracticesmakes it extremelydiffi-
cult for the programto meet its financialobjectives.
Second,RIO facessignificantproblemsregardinginap-
propriateandexcessuse of services.Forexample,HIO pro-
vides a generousdrug benefit overwhichit exerciseslittle
or no control.As a resultmanybeneficiariesuse their HIO
benefitsto obtain medicationsat minimalor no cost.
The governmentis likelyto expand health care cover-
age throughHIO. The recent expansionof HIO coverage
to studentsis an exampleof the government'spolicyobjec-
tives. Yet the premiums and subsidies collected by HIO
are insufficient to pay for the care provided to the increas-
ing beneficiary population. It is unclear whether HIO has
the management capabilities to deal with an increasingben-
eficiary population and implement the reforms needed to
allow it to play an expanded role in health insurance cov-
erage in Egypt.
CurativeCare Organization
The Curative Care Organization (CCO) is a parastatal over-
seen by-the Ministry of Health. The CCO runs twelve hos-
pitals containing a total of 4,846 beds. The hospitals are
located in six governerates, with the largest in Cairo and
Alexandria.As noted, CCO hospitals were private until they
were nationalized in 1964 under the Nasser regime.
CCO provides services to four groups of users: employ-
ees through contracts with employers, individuals on a fee-
for-service basis, low-income groups, and accident victims
free of charge. In 1993the CCO signed a contract with HIO
to provide health care services to students attending voca-
tional schools. The CCO is financed through contracts with
employers opting out of the HIO insurance program, con-
tracts with FRO, out-of-pocket hospitalizationfees, and firom
Ministry of Health grants that cover the free treatment of
low-incomepatients. Almost half of inpatient admissions to
the CCO are covered by contracts with companies,and more
than a third of patients pay out-of-pocket fees.
Sources of revenue for CCO hospitals include fees from
laboratory and inpatient services, premiums and fees from
companies with contracts, a percentage of drug sales, gov-
ermnent subsidies for free beds, and donations and grants.
Private Sources of Financing for Health
Services
Private financing of health care is limited. As noted, private
funds accounts for about 3 percent of national health care
expenditures. The private sector includes privately owned
organizations, private health insurance companies, and
NGOs. Privately owned organizations serve as financiers
and providers of health care, while private health insurance
companies finance health care through funds collected from
individuals and employers on behalf of their employees.
INNOVATIONS IN HEALTH CARE FINANCING
120
There are three main forms of private insurance cover-
age: policies purchased through private health insurance
companies, group insurance policies purchased through
unions, professional organizations, and cooperatives, and
self-insurance policies where care is provided under con-
tract with hospitals and physicians and funded from inter-
nal resources. About 100,000 people are covered under
policies purchased from private insurance companies and
160,000 people are covered under union or professional
organization policies. These policies may be used to sup-
plement other social health insurance programs or to pro-
vide comprehensive coverage for people who can afford it.
Householdout-of-pocketexpenditures
As noted, household out-of-pocket expenditures are the
largest source of health care spending. In 1991 household
out-of-pocket payments accounted for 55 percent of health
care expenditures, or LE 2.3 billion. In addition, employ-
ees must contribute to HIO premiums or to private health
insurance plans (for companies that opt out of the HIO
program). Employees spend an estimated LE 90 million
on HIO premiums and LE 10 million on other health
insurance.
Health maintenanceorganizations
Several attempts have been made to establish health main-
tenance organizations (HMOs) in Egypt. The Middle East
Medicare Plan, established in 1989, was the first nonprofit
organization operating according to principles of managed
care. This plan now covers 10,000 subscribers through
contracts with thirty-six companies. The plan offers a full
range of services,includingcomprehensiveoutpatientand
inpatient services as well as emergency care. The policydoes
not, however, cover dental care, care associated with preg-
nancy,home visits,regular checkups,or eyeglasses.Marketing
of the Medicare plan has focused on group enrollment,
which tends to minimize problems of adverse selection.
Members or their employers pay the annual Medicare pre-
miums on a per capita basis.
The Medicare plan has adopted a number of practices
to control costs and remain profitable. The company under-
writes coverage and maintains the right to evaluate indi-
vidualsabovethe age of 45 and to excludehigh-riskindi-
vidualsand those with preexistingconditions.Tominimize
overutilization,the companyuses affiliatedphysiciansto
act as gatekeepers and to continuouslyevaluate medical
needand appropriateness.Premiumsare adjustedeachyear
based on historicalutilization,.
Medicarepremiumsare lowerthan those of other pri-
vate insurancecompanies.This HMO is considereda suc-
cess, and the number of groupsor companiescontracting
with it increaseseach year.However,most of Medicare's
contracts are with international companies and with
embassies.Thus it is unclearwhether Medicare'sexperi-
ence can be expandedto the national and publicorganiza-
tions that form the bulk of the market.
Professional
organizations,
unions,andcooperatives
Health insuranceplansprovidedthroughprofessionalorga-
nizationsandunionsandcooperativesare commonin Egypt.
'here are abouttwenty-oneprofessionalorganizations(med-
icalpractitioners,engineers,lawyers),most of whichmake
a healthinsuranceplan availableto their members.Health
coveragein many of the union plans can be extended to
cover familymembersand retirees.
MedicalUnion is the largestand most successfulhealth
insurance plan. The plan is open to medical association
physicians,dentists, pharmacists,and veterinarians.Like
other professional-based healthcareplans,the medicalasso-
ciationprovideseligibilityto aliits membersand their fam-
ifies.The plan alsocoversretireesand widowsof deceased
members.MedicalUnionhas about 160,000membersand
coversa wide range of services,induding hospitalization,
outpatient visits, diagnosticprocedures,emergencycare,
and dental care.
The plan uses a number of measuresto control costs
andutilizationof services,includingrequiringcopayments,
setting moderatereimbursementlimitsfor hospitalization
and key procedures,providinginformationto beneficia-
ries about the cost of servicesand proceduresoffered by
different providers, and requiring preauthorization for
hospitalization and major outpatient procedures.
Preauthorizationallowsbeneficiariesto verifytheirremain-
ingcreditwithrespectto the reimbursementlimit.Utilization
of outpatientservicesiscontrolledthroughthe use ofvouch-
PRIVATE HEALTH INSURANCE IN EGYPT
121
ers.Beneficiariesare givenfivevouchersfor outpatientcare
eachyear.Theplanmaintainsa listof 20,000providers,160
hospitals,and 300 laboratoriesaccreditedfor paymentby
the health plan.
Memberspay a fixed annualpremiumto participatein
the plan. Fees are based on a slidingscale that increases
with length of membership.This systemmakesparticipa-
tion in the program more affordableto young physicians
and other union membersat the beginningof their careers.
In additionto premiums,planparticipantsmustmakecopay-
ments each time medicalcare is received.The copayment
is determined by the type of facility and service used.
Specialtycare requireshighercopayments.
The plan has a good reputation amongits beneficiary
population.Mostbeneficiariesuse their participationin the
planto replaceservicesavailablethrough -HO.The Medical
Union plan providesmemberswith the freedom to choose
their providersfrom the accredited list and receivehigh-
qualitycare at a price theycan afford.
Privotehealthinsurance
Egypthas a smallprivatehealthinsurancesectoraccount-
ing forabout 1 percentof annualhealthcare expenditures.
Insurance companiesestablishedprior to the revolution
were nationalizedin 1961.All others were developed by
majorEgyptianbanks.Althoughinsurancecompaniespro-
vide awide range of insurancecoverage,healthplans form
a smallpart of the insurance market. Thus health insur-
anceplansare not preferredby insurers.Onlythree public
organizations and one private company offer health
insurance.
Private healthinsurancemainlycoverspeople who are
ineligiblefor publicinsurance(suchas familymembersof
employeesand the unemployed)and employersand groups
who electnot to use publicinsurancebecauseof its lower
quality.Privatehealthinsuranceservesmostlyprivate,well-
financedcompaniesand high-incomeindividualswho can
affordto paythe highcostof premiums.Privatehealthinsur-
ance is alsoused by the internationalcommunityresiding
in Egypt.
As in many developingcountries,private health insur-
anceis growing.Sincethe 1980sEgypthasreluctantlyimple-
mented structural health sector reforms accompaniedby
shrinking central spending on health care. Today Egypt is
passing through a transitional phase, focusing on reform of
the public health care system to decentralize and provide
greater autonomy to Ministry of Health facilities.The min-
istry is exploring different financing mechanisms-such as
cost recovery for curative care-while maintaining preven-
tive care, primary health care, social health insurance, and
care for the elderly and the indigent within the domain of
the public sector.
Healh system reforms are also aimed at increasing free-
dom in the private health insurance sector. However, it is
unclear whether social,legal, and economic factors willpro-
vide the necessary incentives for expansion of the private
insurance sector. The following sections analyze the fac-
tors affecting the development of private health insurance
and the impediments this industry faces. The market is ana-
lyzed from the perspective of the government, consumers,
and insurers.
Government perspective. Current government policies
do not support the growth of private health insurance, nor
do they encourage beneficiaries to leave the social insur-
ance program. Indeed, for a number of reasons government
policies are intended to maintain the largest possible pool
of beneficiaries in public programs:
* Social concerns.In any health care system, basic values
underlie policydecisions.To understand the health insur-
ance sector, its funding mechanisms, scope of benefits,
and extent of coverage, the social objectives of health
care must be defined. Egypt's constitution guarantees
all citizens access to free health care services. Over the
years the government has pursued this objective using
two parallel strategies. The first included direct govern-
ment financing and service provision. The second
extended coverage using mandatory social insurance.
The Nasser regime strengthened the right to free care.
The government made efforts not only to finance health
care but also to provide free services through a network
of health clinics and hospitals owned and operated by
the Ministry of Health. The government financed med-
ical education and guaranteed all graduating medical
doctors and nurses employment in public facilities.The
government also guaranteed medical coverage to spe-
cific population groups, including the Army, Internal
INNOVATIONS IN HEALTH CARE FINANCING
122
Security Forces, and employees of specific ministries.
Other employees received mandatory health care cov-
erage through the HIO social insurance program.
From a politicalperspective, public financingof health
care services has been important because it has gener-
ated a sense of social solidarity and equity and guaran-
teed universal access to care. But these social objectives
cannot be achieved without paying a price. Funding for
the Ministry of Health has been limited to less than 2
percent of the government budget, and these funds are
insufficient to meet the government's commitments to
health care. The provision of health care services is not
equitable, with far worse access for the poor and the
unemployed. Low-wageworkers must accept poor-qual-
ity care and pay many expenses out of pocket.
* Economic concerns. Significant portions of the Ministry
of Health's budget go to fixed costs such as salaries.
The government policy guaranteeing employment to all
medical school graduates results in increasingsalarycosts
regardless of the number of beneficiaries of public pro-
grams. Thus the government must maintain high levels
of participation to cover these fixed costs.
The government is also forced to continue requiring com-
panies opting out of the public health insurance program
to pay a 1 percent salarytax to support HIO. However, the
companies that opt out of HIO are likely to have employ-
ees in good health. Thus an increasingnumber of low-income
and less-healthy individuals (such as widows and pension-
ers) are participating in HIO and increasing costs.
It is difficult to achieve an equitable public health care
delivery system with a stable share of spending while rely-
ing predominantly on private insurance. The combination
of finance, regulation, and mandates needed to achieve such
an outcome may be theoretically feasible but are politically
and operationally impossible.
Consumerperspective.A large portion of health care in
Egypt is financed out of pocket, which explains why there
is a potential market.for private insurance. However, three
consumer-related factors limit the growth of Egypt's pri-
vate health insurance market. First, the affordability of pri-
vate health insurance. Most Egyptians cannot afford private
insurance premiums, and most public sector companies lack
sufficientcapital to seek group policiesthrough private insur-
ance.Thus private insuranceis likelyto serveonlyprivate,
high-incomeemployersand wealthyindividuals.For most
of the populationprivatehealth insurancepremiumswith-
out public supportis stillbeyondreach.
Second,the availabilityof private insurance.For most
Egyptiansprivate health insuranceis not provided as an
optionfor healthcare coverage.Insurancecoverageis not
universal,and those who are not providedwith coverage
through their employeror through another affiliationare
lesslikelyto seekprivatehealthinsurance.The unemployed,
the self-employed,andworkersin smallcompaniesare likely
to seek coveragethrough Ministryof Health facilitiesor
throughprivateproviderswho are paid out of pocket.
Finally,the acceptabilityof private health insurance.
Culturaland socialfactorsplayan importantrolein deter-
miningthe acceptabilityof privatehealthinsurance.Although
the public recognizesthe protectionagainstrisk that insur-
ance can provide,there is a lack of confidencein private
insuranceplansto providecoveragewhenneeded.For pri-
vate insuranceto become a major sourceof coverage,the
design and implementationof insurance plans must be
simplified.
Insurers'perspective.Although the insurancemarket is
welldeveloped,private health insurancestill lagsbehind.
For insurers,the current marketplaceprovidesfew incen-
tivesto investin privatehealthinsuranceplans.Legal,man-
agerial,andeconomicconstraintshinderthe abilityof private
healthinsuranceto be profitableand to grow.
There are legal constraintsto startinga new insurance
company,domestic or foreign. Government regulations
require capitalreservesto enter the healthinsurancemar-
ket that are beyondthe capacityof manyinvestors.These
regulationsserve as barriers to market entry and reduce
competition.
Becauseprivatehealthinsuranceis not legallymandated
and there is no obligatorymechanismto insure the self-
employed,private insurance is less important and many
Egyptiansare uninsured.Furthermore,the lawguarantees
employeesthe right to opt out of insuranceplans if they
obtainsimilarinsurancethroughotherproviders.Thisoption
diminishesthe size of the risk pool and reduces the eco-
nomic advantageto the insurer.Governmentlawsalsodis-
couragebusinessesfrom optingout of socialinsuranceplans
PRIVATE HEALTH INSURANCE IN EGYPT
123
(suchasHIO) to seekmore efficientprivateinsuranceplans.
As noted, companiesthat opt out of HIO are stillrequired
to pay a 1 percent premiumto HIO.
Froman economicperspective,health insuranceoffers
limitedprofitabilityto insurersbecausepremiumsare too
lowrelativeto the costsof providinghealthcare. Evidence
suggestsa minimal,if not negative,return on investment
for most private insuranceplans.Insurancepremiumsare
low because companiestry to make them affordable or
becausecompaniescannotaccuratelyassessrisks.Moreover,
healthinsuranceplans createmany time-consumingprob-
lemsand contribute little or nothing to companies'profit
margins(Kemprecose1993).
Healthinsurancemanagementandadministrationposes
yetanotherconstrainton the insuranceindustry.The indus-
try has poor informationtechnologyand weak underwrit-
ing skills and lacks expertise in processing claims and
designingpoliciesto meet the needs of Egyptianclients.
Conclusion
Healthcarefinancingin Egyptandthe roleof privatehealth
insurancemust be viewedin perspectiveto the overallecon-
omy.For yearsEgypt's centralizedhealth care systemhas
facedsignificantproblems.The governmenthas beenunable
to providehigh-qualityand efficientpublic servicesto its
citizens,while privateinitiativeshavebeen constrainedby
legalroadblocks,lackof a skilledworkforce,poor technol-
ogy and infrastructuredevelopment, and a complexand
burdensomeregulatoryenvironment.
In the past few yearsEgypt has embarkedon substan-
tialeconomicand politicalreformthat is likelyto affectthe
healthcare sectorin generaland the health insurancemar-
ket in particular.These reforms have encouragedprivati-
zation, increased international trade, and loosenedlegal
constraintson foreigninvestment.Thegovernmentrecently
signaled its commitment to world trade by signing the
GeneralAgreementonTariffandTrade(GATT).It alsohas
loosenedlawson-privateinsurance.Alawallowingthe devel-
opment of new insurance companieshas been proposed
andis beingstudied.The governmentalsohas loweredcap-
ital reserverequirementsfor insurance companiesto LE
60 millionto complywith GATTregulations(Kemprecos
1993).
 Economicreform is likelyto increasethe demand for
privatehealth insuranceas incomesincrease,employment
rises, and the number of national and internationalcom-
panies operating in the country increases.These compa-
niesare expectedto be financially
strongandbetter informed
about health insuranceplans, benefit packages,and the
valuereceivedfortheir money.The largermarket willstim-
ulate competition between health insurance companies and
providers and reduce the cost of care. As noted earlier,many
private organizations opt out of HIO and contract with the
CCO, private hospitals, and private health insurance plans.
This trend is likely to continue, especially if the quality of
HIO services continues to be poor.
Still, private health insurance coverage will continue to
be limited to small segments of the population, mainly
employeesof financiallystrong companies and high-income
individuals, and will likely remain beyond the reach of
most Egyptians. Most will continue to rely on public orga-
nizations and public financing to meet health care needs.
However, for the government to meet its social objectives
in health care, significant reform is needed in public sector
finance, management, and quality of care. As things stand,
public health care coverage is likelyto take two approaches.
First, HIO will likely play an expanding role in health
insurance coverage. The government is considering increas-
ing the number of HIO beneficiaries to cover more seg-
ments of the population (such as children under the age of
five). HIO will also remain the primary health insurance
option forworkers and schoolchildren.However, itis unclear
whether HIO has the capability to manage an increase in
the number of beneficiaries. Significant reform is needed
to improve the efficiencyof HIO services.HIO should con-
sider different service delivery models and strengthen its
ability to design and regulate benefit plans. HIO has a staff
model managed care structure that provides a fertile set-
ting for adopting contemporary managed care approaches.
Through reform, HIO may gain the expertise needed to
design and implement comprehensivegroup coverageunder
a unified social policy.
Second, the goverrnmentwil continue to provide health
Scond,athegrovarnent wicinue torprOovierhea
coverage to groups that are not eligible for HIO coverage.
Ministry of Health medical facilities must continue to meet
the social objectives of providing free access to health care
servicesand serving as a safetynet, especiallyfor low-income
INNOVATIONS
IN HEALTII CAREFINANCING
124
people.However,the ministryis underfunded and cannot
hope to achievethisgoal.Availableservicesin ministryfacil-
ities are likelyto remainscarce, and the qualityof care in
these facilitieswillremain poor.Twoeffortsare needed to
improvecoveragein ministryfacilities:
* The populationserved by the ministryshouldbe better
definedand the basic benefitspackageprovidedto this
population should be specified. Continued and increased
government funding should be given to ministry facili-
ties that are able to meet government targets for specific
measures of health status, utilization, and quality of care.
* Decentralization of ministry facilities is essential to
improve the efficiency of public services. Through cost
recovery efforts, public hospitals today are testing dif-
ferent financing mechanisms, including contracting with
employers and HIO for group coverage. Although pub-
lic hospitals lack the expertise to design benefit plans,
pool risks, and calculate premiums, these attempts are
likely to raise awareness, understanding, and demand
for health insurance.
Finally, the development of knowledgeable consumers
is imperative for the development of a private health insur-
ance market. Many people do not understand their health
risks, the importance of risk sharing, and prepayment for
unforeseen illnesses. Cultural acceptability of the basic
values of health insurance also remains weak. Unless soci-
etyperceives the need for insurance coverage, private health
insurance will continue to serve a limited segment of the
market in Egypt.
References
Badran,Ahmed.1992."Expenditureon Health in Egypt."Paper
presentedto the NationalDemocraticParty.Cairo.
-_____
1993. "The Economics of Health Services in Egypt."
Translatedfrom Arabic.Cairo,
Berman,Peter.1995."Egypt:StrategiesforHealth SectorChange."
HarvardUniversity,The Financingof Health ServicesinEgypt
Data for DecisionMakingProject,Cambridge,Mass.
Boutros, Samir G. 1992. "The Health Delivery System in Egypt:
The CrucialRoleof the Private Sector."Paperprepared for
the Fourth National Conferencefor Junior Physicians,April
13-15, Cairo.
CAPMAS(CentralAgencyfor PublicMobilizationand Statistics).
1992. StatisticalYearbook1992. Cairo.
. 1995. "Egypt Household Health Care Use and
 ExpenditureSurvey."Cairo.
Donabedian, Avedis. 1976. Benefits in Medical Programs.
Cambridge,Mass.:Harvard UniversityPress.
Jeffers, James. 1982. "Health Policy Review."U.S. Agencyfor
InternationalDevelopment,Health SectorAssessmentPhase
I, Cairo.
Kemprecos,Louise. 1993. "Health Care Financingin Egypt."
Ministry of Health and U.S. Agency for International
Development,Cairo.
MOH-IDC(Ministryof Health Informationand Documentation
Center).Variousyears."BasicHealth StatisticalData." Cairo.
Musgrove,Philip.1996.PublicandPrivateRolesin Health:Theory
and FinancingPatterns.World Bank DiscussionPaper 339.
Washington,D.C.
WorldBank. 1992.WorldTables.Baltimore,Md.:JohnsHopkins
UniversityPress.
. 1993.WorldDevelopment Report1993:InvestinginHealth.
NewYork:OxfordUniversityPress.

PRIVATE HEALTII INSURANCE IN EGYPT

125
Strategies for Pricing Publicly
Provided Health Services
PaulJ. GertlerandJeffreyS. Hammer
M ost governmentsspend a lot on health. How
these public expendituresare financedis a cru-
cialelementof successful
healthpolicies,because
it determinesthe budget availablefor public activitiesand
has implicationsfor howexpendituresare allocated.Public
expendituresarefinancedbybothpublicandprivatesources,
with public subsidiesfrom the generalgovernmentbudget
supplementedby privaterevenuefromuser fees.Thiscom-
binationaffectshowpublicsubsidiesare allocatedanddeter-
mines who gets them. Subsidy allocationdecisions also
determinethe extentto whichthe poorare cross-subsidized.
The structureof feescreatesfinancialincentivesthat affect
utilizationpatterns and healthoutcomes,and affectshow
wellindividualsare insured againstthe risk of large eco-
nomic lossesassociatedwith unexpectedillness.
This paper examinesthe way governmentsfinanceand
allocatepublic expenditureson health. Much of the policy
debate has focused on the extent to which governments
are able to mobilizeprivateresourcesto supplementpub-
licsubsidiesin financingthese expenditures.Proponentsof
privateresourcemobilizationarguethat individualsarewill-
ingto payformedicalcare andthat additionalfinancingwill
allowgovernmentsto expandand improvecrucialprograms
(WorldBank 1987;Jimenez 1996).Opponents argue that
the poor areunableto payformedicalcareandwillbe worse
off if governmentsexpand private resource mobilization
(Cornia,Jolly,and Stewart 1987;Gilson 1988).
Most of the literature contributingto this debate has
focusedon the technicalissuesof howmuch moneycan be
mobilizedand what impactit has on health outcomesand
finance,andpublicpolicyat the University
PaulJ. Gertleris professorof economics,
appointments in the HaasSchoolof Businessand Schoolof PublicHealth.JeffreyS. Hammeris principaleconomistin the Policy
ResearchDepartmentat the WorldBank.
127
accessto medicalcare. While informative,this literature
doesnot fullyprescribeoptimalpolicy.Optimalpolicyneeds
to be based on the benefit that the policywouldhave for
societyabove and beyond what would have happened in
the absence of public intervention.The benefit of a pro-
posed policyis the extent to whichit amelioratesindivid-
ual and sociallossesfromprivatemarket failures.Priorities
shouldbe based not onlyon the effectivenessof the policy,
but alsoon the importancegovernmentsplaceon the types
of lossesand the individualswho incur the losses.
Weexaminethe implicationsof currentpoliciesandpos-
sibilitiesforpolicyreformin the context of competinggov-
ernmentpriorities.Governmentscan interveneto correct
private market failuresthat cause health outcomesto be
lowerthan they otherwisecould be, to cross-subsidizethe
poor's accessto medicalcare, and to correctinsuranceand
medicalcare marketfailures.Sincegovernmentshavebud-
get constraints,theycannotfullysubsidizeallprogramsand
activities.This paper arguesthat:
* Publicspendingon healthcan improvehealthoutcomes,
promotenon-healthaspectsofwell-being(forexample,by
insuringthat individualsarenot at riskfor the largeunex-
pected economiclossesassociatedwith random adverse
healthevents),and redistributepurchasingpowerto the
poor. Optimalsubsidyand fee policywilldifferdepend-
ing on how much weight governmentplaces on these
competingobjectives.Subsidiesneed to be reallocated
toward the poor and toward public health programs.
However,increasingpublicsubsidiescan financeonlya
fractionof theresourcesneededto expandthe healthsector.
of Califomiaat Berkeley,
wherehe holdsjoint
*Prices for curativeservices(userfees) havetwo distinct
roles.They can raise revenue, freeingpublic resources
that can be reallocatedto public healthactivitiesand to
limitedcofinancingof qualityimprovementsin curative
care.Butperhapsmore important,theycan increaseeffi-
ciencyin the use of publicfacilitiesandin the healthcare
systemas awhole.However,these gainsmustbe weighed
against evidencethat higher fees can compromisethe
objectivesmentionedabove.The literatureon user fees
has tended to focuson raisingrevenue (and its conse-
quences for the poor), but their more importanteffect
is likelyto be in the allocationof resources.In general,
user fees at the point of servicecan playan important
role in cofinancinghealth care, but not as the primary
meansof finance.
* Revenuefrom user fees is sometimesused to finance
improvementsin the quality of and accessto curative
medicalcare.Individualsare willingto payat leastsome
of the costof improvingqualityand access,especiallyfor
dirugs.However,the rich are willing(and able)to paya
lot more thanthe poor.Thusifgovernmentsuse the aver-
age "willingnessto pay" amount to finance quality
improvements,the rich willuse more servicesand the
poor willuse less.
* Optimal policiesalso depend on the behavior of con-
sumers,privateproviders,and civilservants.Consumers
andprivateprovidersdeterminethe marketenvironment
in whichpoliciesoperate-defining limitsto or in some
casesadditionalopportunitiesforwhat canbe achieved.
Civilservantsdetermine the abilityof governmentsto
implementpolicies.Policiesthat are optimalin onecon-
text should not be generalizedto all. Countries differ
significantlyin the relativeweightsthey placeon policy
goals,and in the constraintstheyfacein their resources
and in the reactionsof markets. Seriouspolicyforma-
tion requiresconsiderablymore analysisrelativeto ide-
ologythan has characterizeddebateson the topic.
* Socialinsurance plans, which enable governmentsto
mobilizeprivate resourcesfor healthby collectingpre-
paymentsand chargingfor the healthservicesprovided
to beneficiaries,hold promise,particularlyfor middle-
andhigh-incomecountries.Theseschemesmobilizepri-
vate resourceswith no lossin the insurancevalueof the
publichealth care system.Price discriminationis easier
INNOVATIONS IN HEALTII
128
becauseit can be centralizedand needsto be done only
periodically,outside the pressure of havingto treat an
illness.Despitetheirpromise,however,prepaymentplans
often introduce inefficientmedicalcare cost inflation
that developingcountriesmaybe unable to afford.
The next sectiondiscussesthe role feesplayin the bud-
getaryprocess,consideringhow fees can stretch the gov-
emient's budget for variousprograms and how they can
be used to allocatepublic subsidies.From this discussion
comes a set of conditionsthat determinehow fees affect
budgetaryflexibility-most ofwhichconcernhowpriceand
qualityaffectutilization.The thirdsectionreviewsthe empir-
ical literature on utilization.The fourth section uses the
informationpresented in the previoussectionsto recom-
mend optimalfee policy,reflectinggovernmentobjectives.
The final sectionpresents conclusions.
Role of Fees in the Budgetary Process
Public expenditures on health are financed by revenues
fromprivate sourcesand allocationsfromthe generalgov-
ermnentbudget(generaltax revenuesand donorassistance).
In manycountries,especiallydevelopingcountries,the pub-
lic sectorcollectsprivaterevenuesthrough feeschargedat
the point of service.
Much of the literaturejustifiesincreasinguser fees in
terms of mobilizingresources(or achievingcost recovery)
and in terms of creatingincentivesfor more efficientuse
ofpublicmedicalservices(WorldBank 1987).Butan equally
importantrole for fees is in determiningthe allocationof
subsidesfrom the general governmentbudget across ser-
vices(hospitalization,primarycare,vectorcontrol)andtypes
of individuals(the poor, the elderly,children).The alloca-
tion of subsidiesis one of the mainpolicyinstrumentsgov-
ernments haveto correct health care market failuresand
improvewelfare.
This sectiondescribesthe role of user feesin determin-
ing the government'sbudget constraint.The structure of
feesdeterminesnot onlythe amountof resourcesavailable
andthe amountspent on eachprogram,but alsothe extent
to which a particularprogram'sexpendituresare publicly
subsidized.It is important to remember that this discus-
sion is limited to howfees affect the budget. It saysnoth-
ing about which programs should be funded, how much
CARE FINANCING
shouldbe spent on them, and how much of the expendi-
ture shouldbe financedby public subsidies.That discus-
sion,coveredlater in the paper,requiresinformationon the
benefitsof such allocationsand the objectivesof govern-
ment intervention.
Thebudgetconstraint
Mostof the resourceallocationdecisionsthat publichealth
caresystemsmust makeare relatedto one anotherthrough
the government'sbudget constraint.The two main types
of decisionsare: What servicesshould be offered, and of
what quality?Andwhat shouldthe user fee or copayment
be for each service?
Thesedecisionsare relevantto alllevelsof government
where officialshave to make finance and resource alloca-
tion decisions.In manycountriessuchdecisionsare made
at high levelsof government-either the centralor provin-
cial level. Other countries are devolvingresource alloca-
tion and financingdecisionsto local levels.The analysis
belowappliesto localofficialsand publicfacility-levelman-
agers as wellas centraland provincialofficials.The extent
to whichit appliesto the localleveldepends on the degree
of autonomyin the system.
Thbelevelof servicesand the fee structurecannot be set
independentof one another,but rather must be set to sat-
isfythe budgetconstraint-that is,total expendituresmust
be lessthan or equalto total revenues.Revenuescomefrom
public subsidies and general tax revenue and from user
chargesforservicesprovided.Thisbudgetis spenton admin-
istrativecosts,inpatient services,outpatient services,and
publichealth diseasepreventionand controlactivities.
Governmentspendingon healthcoversa wide rangeof
services,from publichealth activities(suchas the preven-
tion and treatment of communicablediseases)to curative
servicesthat benefitonlythe individual.Spendingon some
programs (hospitalization,primary care, prenatal care)
depends on the number of people who demand care.
Althoughthere maybe short-termrationingof these ser-
vices,in most cases the public sectoris obligedto provide
these servicesto all who requestthem. The costs of other
programs,such as massinformationcampaignsand vector
controlprograms,do not depend on the number of users.
In manycasesthe amountof funds availableforthese pro-
STRATEGIES FOR PRICING PUBLICLY PROVIDED HEALTH SERVICES
129
gramsis determinedby the residualamount left over from
spendingon preventionand treatment services.
The government'sbudgetconstraint,whichsets expen-
ditures equalto revenues,can be expressedas:
A + Xc1 Uj + -cjoUjo + CDC = XfUi + -fioUio + G
i i i i
whereA is administrativecosts,cij is costof inpatient ser-
vicei, Uigis utilizationof inpatientservicei, cio is cost of
outpatientservicei, Uio is utilizationof outpatientservice
i, CDCis expenditureson programsthat are not utilization
driven (vector control, research, sanitation, water treat-
ment),fij is user fee chargedfor inpatientservicei,fio is
user fee chargedfor outpatientservicei, and G is subsidies
from generaltax revenues.
Manypolicymakerspromote user feesas a wayof mobi-
lizingprivateresourcesfor publicexpenditures.It is impor-
tant to note,however,that charginguser feesis not the same
as forcingindividualsto payout of pocket.Here we define
user fees as the price receivedby a facilityor program-
not necessarilywhatindividualspay at the point of service.
Indeed,individualscouldcontributeto prepaymentor insur-
anceplans to financetheir paymentof fees at the time of
treatment. Moreover, fees are not necessarilypaid to
providersin the form of fee-for-service.Prepaymentand
insuranceplans could just as easilypay providersby capi-
tation.But sinceinsuranceaffectspeople'sutilizationdeci-
sions and the form of paymentaffectsproviderbehavior,
the source and form of payment need to be taken into
accountwhen decidingon the structure of fees.
Althoughthe abovecharacterizationof the publicbud-
get constraint is described in the context of centralized
decisionmaking,it can easilybe generalizedto a less cen-
tralized structure. The simplest and most efficient bud-
getarymodel is one in which the entity that collectsfees
keeps them and is free to use them as it sees fit. In this
casefee revenuesexpandavailableresourcesand localman-
agers, if competent and publiclymotivated,can use the
resourcesto improvewelfare.In this casethe centralgov-
erinent must decide how to allocate subsidesfrom the
centralbudget amongnationalprogramsand to lowerlev-
els of government(provinces,districts, states,and so on).
These lower levels of government then combine these
allocationswith subsidiesfrom the localbudgetand decide
how to allocatethe combined resources among facilities
and localprograms.Then each facilityand program com-
binesthese allocationswith fee revenueand makesexpen-
diture allocationdecisionsamongprogramsand services.
Thus each decisionmakerin the processreceivesan allo-
cationfrom a higherlevelof government-G in the above
equation-and combinesit with localresourcesto finance
expenditures.
Resource
mobilization
The classicapproach to resource mobilizationis to raise
prices (userfees) to generateprivate revenuesthat can be
used to finance a serviceor finance improvementsin the
quality of that service.The practiceof charginguser fees
for medicalservicesat public facilitieshas been adopted
throughout much of the world (Griffin 1987; Nolan and
Turbat 1995;Jimenez 1996).However,fee structuresand
controlover revenuesvarygreatlyacrosssettings.
Much ofthe evaluationofresourcemobilizationfocuses
on cost recovery-that is, the percentageof costscovered
by fees. However,it is not clearhow to evaluateresource
mobilizationeffortswhen cost recoveryis limited.In par-
ticular, the value of private resources needs to be mea-
sured in terms of freeingup scarce public subsidiesthat
can be reallocatedto high-priorityprograms(suchas con-
tagiousdiseasepreventionandvectorcontrol)and in terms
of providing facility managers with enough additional
resources(suchas drugs)to makeup the differencebetween
effectiveand ineffectivetreatment. The value of private
resourcesdepends on the extent to whichthe funds pro-
videneededbudgetaryflexibilityatthe margin,wheresmall
amounts of moneygo a long way.
CreeseandKutzin(1995)examinednationalcostrecov-
eryratiosfromfifteen(mostlyAfrican) countriesand found
that elevenhavefee revenuesthat financeless than 5 per-
cent of public expenditures. While such levels are sub-
stantiallyless than the 10 to 20 percent potentialcited in
WorldBank (1987), it is hard to evaluate these data out-
side the countries' institutionaland policyenvironments.
A number of questionscome to mind. For example,how
do these cost recoveryratios compare with government
targets?And howmuch potentialrevenueis not being col-
lected, and why?Many countriesdo not allowlocal units
INNOVATIONS IN HEALTIH CARE FINANCING
130
to retain and spend the feesthat theycollect,and dedicate
few resourcesto administeringand monitoringcollection.
In addition,these data usuallycomefrom national infor-
mation systemsthat suffer from seriousunderreporting.
Thesedata do not implythat well-designedand -admin-
isteredprogramscannotmobilizeresources-just that many
of the countriesstudied lackedthe politicalwill to do so.
Bycontrast,China has been extremelysuccessfulin mobi-
lizingresourcesthroughfees (WorldBank 1996).Even in
1978, before recentreformswere initiated,subsidiesfrom
generaltax revenuesfinancedjust28 percentofpublichealth
expenditures.By1993publicsubsidiesaccountedfor even
less-14 percent of public health expenditures.The rest
wasfinancedthroughfeeschargedto both insuredand unin-
suredpatients.Indeed, costrecoveryratios are alsohigh in
localinitiatives,wherethe revenueis typicallyretained and
whereit is easierto implementand evaluateresourcemobi-
lizationefforts.Forexample,in 1993the revenuesfromfees
chargedto insured and uninsuredpatients accounted for
91 percentof hospitalexpendituresand 84percent ofhealth
centerexpendituresin China(WorldBank 1996).In Senegal
private revenues amounted to 127 percent of recurrent
expenditures in health centers (UNICEF/BIMU 1995).
Similarlevelshavebeen reportedin LatinAmericaand else-
where(Richardsonandothers1992;Olaveandothers1992;
BarnumandKutzin1993;Lewis1993).In addition,McPake,
Hanson, and Mills (1993)found that a number of African
countriesuseddrugfee revenuesto obtaintangibleimprove-
ments in health services.
Still,the evidencethat somelocalesare mobilizingsub-
stantialresourcesdoes not makeup for the administrative
costsof collectingthe fees, includingthe time (opportunity)
costsof administeringfee exemptionpolicies.There is lit-
tle ifanycredibledata on thisimportantissue.Most assess-
mentstake placein the contextof schemesthat havebeen
fundedthroughexternalassistance,whichbiasesdownward
estimatesof real-worldadministrativecosts (Creese and
Kutzin1995).
The extent to which raising fees mobilizes private
resourcesdepends on the extent to whichindividualsare
willingto paythe higherprice for services.Patientsare not
willingto pay any amount for curative care. As fees rise,
utilization will fall. The question is, by how much? The
less sensitive demand is to price increases-that is, the
more price inelasticl-the more revenue is mobilized
through price increases.This is because a price increase
has two effectson revenues.It increasesrevenuesby rais-
ing the revenueper patientvisit,but it lowersrevenuesby
reducing the number of visits.If the reduction in visitsis
great enough, price increases actuallyreduce revenues.
Similarly,the less sensitiveis demand, the less prices will
changeserviceuse.
The storyis somewhatmore complicatedwith respect
to increasinguser feesto financequalityimprovements.In
this casethere are two effectson utilization-the negative
effectsof the price increaseand the positiveeffectof the
qualityincrease.Both need to be measuredto assessthe
amountof resourcesthat can be mobilized.The less price
elasticand themore qualityelasticis the demand,the greater
are the resourcesthat can be mobilizedfroma fee increase
used to financequalityimprovements.
Allocating
publicsubsidies
In additionto mobilizingprivateresources,feesdetermine
the allocationof public subsidies.This point is extremely
importantbecauseit is throughthe allocationof publicsub-
sidiesthat governmentis able to pursue its objectivesand
correct market failures.Increasesin feesfree up subsidies
that can be reallocatedto other programs.The more sub-
sidies a given fee increasefrees up, the greater is the gov-
ernment's budgetary flexibilityin allocating subsidies.
Interestingly,the conditionthat increasesbudgetaryflexi-
bilityis exactlythe oppositeof the conditionthat mobilizes
more private resources-namely, the more price elasticis
demand,the greateris the amountof subsidiesthat are freed
up.
Tosee this, we rewritethe budget constraintas follows:
A + X (c1I - fj1)Uij+ A (c1i - f1I )Ui0 + CDC = G,
where (cij -fii) is the publicsubsidyrate per unit of a ser-
vice.Then the amountof publicsubsidiesspent on a pub-
licprogramis the subsidyrate timesthe amountof services
provided. In this formulation of the budget constraint,
administrativecosts plus the sum of subsidiesto each of
the servicesand publichealthprogramcostscannotexceed
totalsubsidiesallocatedfromthe generalgovemmentbudget.
STRATEGIES FOR PRICING PUBLICLY PROVIDED HEALTH SERVICES
131
There are a number of waysthe governmentmightwant
to reallocateits public subsidies.Consideran increasein
public spendingon public health activitiessuch as vector
control or sanitation-that is, CDCspending.To increase
CDCexpenditures,the governmentmust reduce subsidies
to other programs;otherwiseit would spend more than its
availableresources and violate the budget constraint.To
do so, it raisesthe fee forthose services-thereby lowering
the subsidyfor beneficiariesof the program and inducing
someto stop usingthe service.
Theamountof subsidiesthat can be reallocateddepends
onthe amountfreedup bythe priceincrease,whichdepends
on how sensitive utilization is to price. The more price
elasticis demand, the greater is the drop in utilizationfor
a given price increase.Thus the greater is the amount of
subsidiesthat canbe reallocatedthroughreductionsin both
unit subsidiesand volumeprovided.In essence,the more
price elastic is demand, the more easilythe government
canreallocatesubsidies-that is,the greaterisits budgetary
flexibility.However,the more price elasticis demand, the
fewerthe amountof privateresourcesthat canbe mobilized.
Anotherreallocationprioritymay be to shift subsidies
from a lower-prioritypatient care programto a higher-pri-
orityprogram.To increasepublic subsidiesto a care pro-
gram, the government lowers the fee charged, thereby
increasingthe subsidy rate. The amount of public subsi-
diesgoingto that programincreasesfor two reasons.First,
users of the programreceivea highersubsidy.Second,the
lowerfee attractsnew userswho otherwisewouldnot have
receivedthe subsidy.This discussionimpliesthat reallo-
cating public subsides across care programs is a careful
balance of raisingand loweringuser fees.
Revenueretention
An importantassumptionin this discussionis that anyrev-
enuesraisedfromprivatesourcesare kept in the healthsec-
tor. If the fee revenue must be returned to the general
treasury,then the fee increaseeffectivelydoes not increase
resourcesfor health care; the same outcome holds when
local healthunits are forcedto return revenuesto central
ministriesof health. It is as if the governmentloweredpub-
lic subsidiesby one dollarfor everydollarraised privately.
This approachimpliesthat no resourcesare mobilizedand
providesno incentivefor health care providers to collect
fees, resultingin substantiallyless revenuebeing collected
than couldbe raised-an importantadministrativeissuein
implementinga successfuluser fee strategy.
In most casesthe claimon fee revenuescollectedat the
locallevel is likelyto come from higher levels of govern-
ment. In the worst case fee revenuesleavethe health sec-
tor andare returnedto centralor localtreasuries,andhealth
sectorresourcesdo not expand.This is the case in African
countriessuchasEritrea,Ethiopia,Namibia,andZimbabwe
(Creeseand Kutzin1995).
Amoresubtleoutcomethat is harderto documentoccurs
whenfee revenuesmerelydisplacepublicsubsidiesby one
dollar foreverydollar of revenueraised.Indeed,ministries
of healthin Cambodia,Chile, China,Iran, Jordan, Nepal,
andThailandhavecut budgetsas feerevenueshaveincreased
(Leighton 1996).In contrast, healthcare providersin the
CentralAfricanRepublicand Kenyarefusedto restructure
fees until theyhad receivedexplicitassurancesfrom their
ministriesof finance that public subsidieswould not be
reduced.
In both the direct and the subtlecases,fee revenuesdo
not augmentthe abilitiesof policymakersand facilityman-
agersto expand and improveprograms.Rather,user fees
simplybecome a way to cut public subsidiesto the health
sector.
Even when fee revenues remain in the health sector,
manygovernmentsdo not allowlocalfacilitiesandprograms
to control the funds. In many countries this is because
local facilityand programmanagershave no authorityon
how to spend resources.For example,staffingand infra-
structuredecisionsare usuallyhandled at higherlevels.At
best local managerscan alter the mix of drugs and sup-
plies;few are allowedto shift resourcesbetweenbudgetary
items. The reasons for this lack of local control include
lack of budgetaryskillsand concernabout potentialgraft.
However,asthe currentdecentralization trendgatherssteam,
such rationalesmaydiminish.
Indeed, one reason to speed decentralizationis that it
increasesthe efficiencywith whichfees are collected.The
fact that a governmentmandatesthat feeswillbe charged
does not mean that facilitiesand programswillcollectthe
fees and remit them to the government. Indeed, if rev-
enuesare simplypassedon to higherauthorities,localfacil-
INNOVATIONS IN HEALTIH CARE FINANCING
132
itiesand programshavelittleorno incentiveto collectthem.
Allthat remainsis the threatof auditandpunishment,which
is costlyand rarely credible.But if local facilitiesand pro-
grams are able to keep and use at least some revenues,
they havean incentiveto collectthem.
Despite this argument,there has been little systematic
evaluationof the hypothesisthat fee retention increases
the efficiencyof collection.Afewcasestudies,suchasChisvo
and Munro(1994)in Zimbabwe,claimthat the lackoflocal
retention of fees explainswhy only a smallpercentage of
public health expendituresare financedthrough fee rev-
enues.In a studyof Africa,Nolanand Turbat (1995)com-
pare percentagesof public expendituresfinancedthrough
fee revenues and find no correlation with fee retention
authority.But the comparisonsare not clear,and manyother
factors(includingdifferentfee policies)couldconfoundthe
observedcross-nationalrelationship.
Thereis,however,evidenceon the expendituresidethat
local controlis important.Parker and Kippenberg(1991)
examinedfourteencountries'experiences underthe Bamako
Initiative and found that stronger local management of
resourcesincreasedthe availabilityof essentialmedicalcare
inputsandimprovedserviceutilization.Mwabu,Mwanzia,
and Liambila(1995)found that whilemost revenuesfrom
a 1989fee increasewere locallyretained, Kenya'scentral
treasury retained authorityover spending.As a result 40
percentof Kenyanfacilitiesreportedthat theydidnot spend
the revenues.
What Effect Do Prices Have on the Budget
and on Individual Welfare?
Thissectionexaminesthe extentto whichpublichealthcare
systemsare able to mobilizeprivateresourcesand maintain
budgetaryflexibilityto reallocatepublicsubsidies,and the
consequencesofthesepoliciesonhealthoutcomesandaccess
to medicalcare.The firsttwo issuesare essentiallya ques-
tion of the extent to whichraisinga fee lowersutilization.
Thefirstsubsectionreviewsempiricalevidenceon howprice
increasesaffectthe utilizationof the institutionthat raises
the price-that is, ownpriceelasticitiesof demand.
Whether inelasticdemand is good or bad depends on
governmentobjectives.When subsidiesare reallocatedfrom
oneprogramto another,themore agivenfeeincreasereduces
utilization,the greateris the amount of subsidiesthat can
be reallocated.When fees are increasedto financea pro-
gram,the more priceinelasticis demand, the greateris the
amountof resourcesthat can be mobilized.When feesare
increasedto financeimprovementsin quality,the less the
fee increasereduces utilizationand the more the quality
improvementraisesutilization,the greater is the amount
of resourcesmobilized.In this casethe question is essen-
tiallyhowmuchpeoplearewillingto payforqualityimprove-
ments,whichis examinedin the second subsection.
The discussionsin the next two subsectionsshould be
interpreted only in terms of implicationsfor the budget
and in terms of welfare.A decreasein utilizationof public
facilitiesdoes not necessarilymean that health outcomes
or accessto all medicalcare wasreduced. For example,if
a price increase induces individualsto substitute private
care forpublic care,accessis not reduced.In addition,indi-
viduals who decide not to seek care because of a price
increasemayhave onlyminorillnessesthat do not require
treatmentor that can be treated at home.Own priceeffects
tellusnothingabout accessandhealthoutcomes,onlyabout
resourcemobilizationand budgetaryflexibilityin allocat-
ingpublicsubsidies.Theeffectsof priceincreaseson access
and healthoutcomesare analyzedin the third subsection.
None of the discussionin this sectionis devotedto what
governmentsshoulddo.This discussionprovidesinforma-
tion aboutthe budgetaryconstraintsongovernmentactions
and implicationsof actionson outcomes.To take the dis-
cussiona step further,we must understandthe objectives
of governmentpolicy-the subject of the next section of
the paper.
Priceelasticityofdemandformedicalcare
Many studieshaveused cross-sectionalhouseholdsurveys
to estimatethe priceelasticityofdemandforoutpatientser-
vices (table 1). Although a few earlystudies using ques-
tionabledata found completelyinelasticdemand-that is,
priceshadno effectonutilization-most studieshavefound
that higherpricesare associatedwith lowerutilization,but
that overallown price elasticitiesare low and well below
unity.Thisfindingsuggeststhat increasesin feeswillmobi-
lizesubstantialprivaterevenues,but that flexibilityin real-
locatingsubsidiesis limited.
STRATEGIESFOR PRICING PUBLICLY PROVIDED HEALTH SERVICES
133
Anotherimportantresult of a number of these studies
is that pricesensitivitydiffersby economicanddemographic
group.The poor appearto be more pricesensitivethan the
rich, and children'sutilizationseemsto be price sensitive
than adults'. Thus increasingfees mayreduce the utiliza-
tion of the poor and childrenby more than the utilization
of the rich and adults.
Policymakers shouldviewtheseresultswithcaution,how-
ever,sincethe studies suffer from severalmethodological
problems.First,the countriesstudiedtypicallyhad public
institutionsthat chargedverylowfees,withlittlegeographical
variation in those fees. In some studies the investigators
used travel costs to measure price elasticities,since time
costs ration the market when fees are low.These studies
used the estimated modelsof demand to conduct policy
simulationsthat forecasthowincreasesin feeswere likely
to affect utilizationand revenues.However,the forecasts
were based on price changes that were far outside the
observedrangeof the pricedata.Thustheyare highlyunre-
liable.In particular,while an individual'sutilizationdeci-
sionsmaybe insensitiveto priceswhenprices are low,they
mightbe very sensitiveto priceswhenprices are high.
Severalstudieshaveevaluatedactualprice increasesby
comparingutilizationbefore and aftera fee increaseat pub-
lic facilities. In the Ashanti-Akim region of Ghana,
Waddingtonand Enyimayew(1990) found that after an
increasein user fees,long-termutilizationfellat clinicsserv-
ing the poor but did not fall at clinicsservingthe nonpoor.
Mwabu, Mwanzia,and Liambila(1995) found a 52 per-
cent drop in outpatientvisitsto governmenthealthcenters
after Kenyaintroduced fees in 1989.After the fees were
suspendedin 1990,visitsincreasedby 41 percent-almost
to the originallevel. Yoder (1989) reports a 32 percent
drop in visits to government health care facilitiesafter
Swazilandincreasedfees.Kahenyaand Lake (1994)found
that attendanceat elevenclinicsin Zambiafell by an aver-
age of 64 percent after feeswere raised,with largerreduc-
tions occurringat facilitiesin poorerareas.Finally,Bennett
(1989)found that the utilizationof governmentfacilities
dropped by 40 to 51 percent after feeswere increasedin
Lesotho,and that the dropswere greatestamongchildren.
Asnoted, however,these studiesshouldbe viewedwith
extreme caution.Althoughthe resultsare consistentwith
econometric findings that demand is sensitive to price,
TABLEI
Econometricestimatesof own price elasticitiesof the demand for medicalcare in developingcountries
Own priceelasticity
Country Dataperiod/type Servicetype Overall Low income High income Source

BurkinaFaso 1985 Publicprovider Sauerbom,Nougtara,and Latimer

Allages -0.79 -1.44 -0.12 1994
Age0- 1 -3.64
Age 1-14 -1.73
Age 15+ -0.27

C6te d'lvoire 1985 Healthclinic -0.61 -0.38 Gertler andvander Gaag(1990)

Hospitaloutpatient -0.47 -0.29

C6te d'lvoire 1985-87 Healthclinic -0.37 Dow (1996)

Hospitaloutpaient -0.15

Ghana 1987 Hospitalinpatient -1.82 LavyandQuigley(1993)

Hospitaloutpatient -0.25
Dispensary -0.34
Pharmacy -0.20
Healthclinic -0.22

Kenya 1980-81 Govemmentprovider -0.10 Mwabu,Ainsworth,and Nyamete

Missionprovider -1.57 (1993)
Privateprovider -1.94

Indonesia 1991-93 GerderandMolyneaux(1997)

Children Healthcenter -1.07
Healthsubcenter -0.35
Adults Healthcenter -1.04
Healthsubcenter -0.47
Elderly Healthcenter -0.47
Health subcenter -0.11

Mali 1982 -0.98 Birdsallandothers( 1983)

Pakistan 1986 Aldermanand Gertler(1997)

Female Traditionalhealer -0.43 -0.24
Children Publicclinic -0.43 -0.23
Pharmacist -0.44 -0.25
Privatedoctor -0.17 -0.09
Male Traditionalhealer -0.60 -0.26
Children Publicdinic -0.61 -0.27
Pharmacist -0.63 -0.27
Privatedoctor -0.25 -0.10

Peru 1985 Privatedoctor -0.44 -0.12 Gertlerandvander Gaag(1990)

Hospitaloutpatient -0.67 -0.33
Healthclinic -0.76 -0.30

Philippines 1981 Publicproviders -2.26 -1.28 Ching(1995)

Privateproviders -3.93 -2.23

Philippines 1981 Prenatalcare -0.01 Akin andGriffin(1986)

Philippines 1983-84 Urbanmatemity -0.24 Schwartzandothers(1988)

Ruralmatemity -0.05

INNOVATIONS IN HEALTH CARE FINANCING

134
theyreportonlythe fallin utilization,and saylittleabout
thedegreeofpriceelasticity. Largedropsinutilizationcould
be associatedwithbig priceincreases,and demandcould
stillbe relatively
priceinsensitive.Moreover, becausethese
studieslackedcontrolgroupsforwhichfeesdidnotchange,
thereisno wayofknowinghowmuchofthe fallin utiliza-
tionwasdueto thepriceincreaseandhowmuchwasdue
to factorssuch as changesin quality in the disease
environment.
Mostof the resultsdiscussedso far analyzedatathat
reflecttherulesthatgovernments useto setpricesandlocate
facilities(Rosenzweig andWolpin1986;Pitt,Rosenzweig,
and Gibbons 1993; Gertler and Molyneaux1993;
Frankenberg1995).Sincegovernmentpolicyis tryingto
achievesomeobjective,thevariationin feesis unlikelyto
be random.Rather,it reflectsgovernment policyIn many
casesgovernmentsset fees and locatefacilitiesbasedon
population characteristics
suchaseconomic statusandhealth
problems.If the multivariateanalysisdoes not explicitly
accountforgovernment policiesforsettingfeesanddeter-
miningthelocationsoffacilities, the estimates oftheeffects
of thefeeonutilization willbe confounded withtheeffects
of utilizationon government policy 2 In addition,mostof
thesestudieshaveonlyrudimentary controlsforqualityof
care.Although theyusuallydistinguish betweenlevels(hos-
pital,healthcenter,and so on) and sector(publicor pri-
vate)ofcare,theydonotcontrolforqualityvariations (such
as drugavailabilityand providertraining)withinprovider
types.
Threestudiesthatarenot subjecttothesecriticisms ana-
lyzethe effectexperimentally designedfeeincreaseshave
onindividualutilization in experimental andcontrolareas.
Thefirst,Gertlerand Molyneaux(1997),estimatedprice
elasticities
of demandfor outpatientservicesin Indonesia
usinglongitudinal paneldata in whichpublicsectoruser
feeswerevariedexperimentally in twoofIndonesia's twenty-
sevenprovinces. Thestudydesignwasintegratedwithlocal
politicaldecisionmaking authority,
whichwasalreadyplan-
ningto increaseuserfees.Ratherthanraisingfees every-
where,feechanges werestaggered to generatepricevariation
basedon an explicitexperimental design.Userfeeswere
increasedin somedistricts(treatmentareas)but not in
others(controls)and in both governmenthealthcenters
andhealthsubcenters.
STRATEGIESFOR PRICING PUBLICLY PROVIDED HEALTH SERVICES
135
TheestimatedIndonesianpriceelasticitiesof demand
are shownin table 1. Althougha priceincreasesignifi-
cantlylowersutilization,the effecton childrenis greater
thanon adults,andthe effecton adultsisgreaterthanon
the elderly.Demandfor healthcentercareis moreprice
elasticthanfor healthsubcenters.Thisfindingis not sur-
prising,sincethe healthsubcentersserveruralpopulations
thathavefeweralternative (publicor private)providersthan
in urbanareas,wherehealthcenterstendto be located.In
fact,thepriceelasticity ofthedemandforhealthcentercare
iscloseto unity,whereasitiswellbelow1forsubcenters-
suggesting that littlerevenuewillbe mobilizedby raising
healthcenterfees,but that a lotwillbe mobilizedby rais-
ingsubcenterfees.
Sirilarly,
Cretinandothers(1992and1996)reportresults
froma ruralhealthinsurancestudyin tworuralChinese
countiesin whichcopayments, anotherformof userfee,
wereexperimentally variedto estimatepriceelasticitiesof
demand.During1989and 1990twenty-six villagesin two
ruralcountiesof Sichuanprovince,China,participatedin
an experimental longitudinal studyto providean analytic
basisfordevelopingsoundhealthcarefinancingmecha-
nisms.The experimentassignedeachvillagetwo health
insuranceplans,one for 1989and onefor 1990.Eightdif-
ferentplanswereassigned,withoutpatientand inpatient
coinsurance ratesrangingfrom30 to 75percent.Threeof
theplansemphasized thecoverage ofoutpatientcare,three
emphasized thecoverageofinpatientcare,andtwooffered
balancedcoverage ofoutpatientandinpatientcare.Although
participationin the insuranceplanswasvoluntary,each
householdhad to enrollas a unit,and participationrates
weremorethan95percent.
Asexpected,highercoinsurance ratesforoutpatientser-
vices(usedbyabouttwo-thirds ofthepopulationeachyear)
wereassociated withsignificantly lowerprobability of use
andsignificantly lowerexpenditures. Forinpatientservices
(usedbyjust3 percentofthepopulationeachyear)higher
coinsurance alsoledtolessutilization andlessexpenditures,
althoughthedecreasewasnot asstrong.Theseresultssug-
gestthatdemandforoutpatientservicesismorepriceelas-
ticthandemandforinpatientservices.Sincemoreserious
illnesses
aretreatedbyinpatienthospitalization, theseresults
areconsistentwiththe hypothesis that demandfor med-
icalcareislesspricesensitiveformoreseriousillnesses. This
finding is somewhat reassuring because it suggests that
reductions in utilization resulting from price increases are
likelyto be lower for less serious illnesses.
Manning and others (1987) and Newhouse (1995) report
the results of a large health insurance experiment in the
United States conducted in the late 1970s. More than 20,000
individuals in six sites were randomly assigned to one of
fourteen health insurance plans that had different copay-
ment structures. This experiment differed from the
Indonesian and Chinese experiments in that its design was
based on controlled random assignment of a large number
of individuals. Although the Asian experiments were con-
trolled, individuals were not randomly assigned. Rather, the
intervention was at the community level. Randomization
at the individual level provides a better and more robust
design.
The U.S. experiment was one of the first scientifically
valid studies that convincingly documented that individu-
als' medical care utilization decisions were influenced by
prices. The results indicated that prices had a bigger influ-
ence on decisions to initiate treatment that on the amount
obtained once treatment began. Overall, the experiment
found price elasticities of about -0.2, and the price elas-
ticityincreasedwith the coinsurancerate. Moreover,demand
for acute care and inpatient services was less sensitive to
price than chronic care and outpatient care. This is consis-
tent with the Chinese results, and with the hypothesis that
demand for the treatment of more severe illnesses is less
sensitive to price.
Willingness
to pay for better quality and access
There is evidence that people are willing to pay at least a
share of the cost of improving access and quality, especially
for drugs. (See Alderman and Lavy 1996 for a review of
the literature.) For example, studies of cross-sectionalhouse-
hold data have found that individuals are willing to pay at
least some of the cost of improving access to medical care,
as measured by the distance that they have to travel to reach
the closest public facility (Gertler and van der Gaag 1990;
Lavy and Germain 1995). In addition, four studies analyz-
ing cross-sectional data have found that a number of struc-
tural quality indicators, especially drug availability,
significantlyaffected demand in Ghana (Lavyand Germain
INNOVATIONS IN HEALTII CARE FINANCING
136
1995), Nigeria (Akin, Guilky, and Denton 1995), Kenya
(Mwabu, Ainsworth, and Nyamete 1993), and the
Philippines (Hotchkiss 1993).
The estimated quality effects are quite large. For exam-
ple, if thepercentage of Ghanaian public facilitieswith drugs
increased from 66 percent to 100, utilization of public
facilities would increase by nearly 44 percent (Lavy and
Germain 1995). Simultaneous improvements in drugs, infra-
structure, and serviceswould increase the use of public facil-
ities by 127 percent. Much of the increase in utilization,
however, comes from substituting public for private care.
The same quality improvements that increase public uti-
lization by 127 percent would only reduce self-care by 14
percent. Thus the net effect on utilization is quite small.
However, users of public services would have higher-qual-
ity care than before. Since quality improvements mainly
affect the choice among providers rather than whether to
obtain treatment, the net effect on health outcomes is deter-
mined solely by the increase in quality to current public
sector patients, not by increased access.
These studies use estimated parameters from demand
models to calculate the willingness to pay for quality and
access improvements. Methodologically,this is equivalent to
asking how much a fee would have to increase to offset the
increase in utilization from the improvement in quality or
access.3 Gertler and van der Gaag (1990) found that the geo-
graphic distribution of individuals in Peru and C6te d'Ivoire
makes them willingto pay (on average) about 20 percent of
the cost of operating a health facility if it cuts travel time to
public health centers from two hours to zero. The poor,
however, were wiling to pay substantially less than the rich.
In Ghana individuals were willing to pay, on average, half
the cost of improved access, with the poor again wiling to
pay substantiallyless (Lavyand Germain 1994). Thus if the
government asked consumers to paythe "averagewillingness
to pay" for improved access,the nonpoor's utilization would
increase while the poor's utilization would decrease.
Although these studies demonstrate a significant statis-
tical correlation between quality and utilization, they raise
several questions about the direction of causality.First, the
studies used cross-sectional household data to investigate
the effects of price, travel time, and quality on utilization
and then used the estimated models to simulate the effects
of price and quality changes on utilization. Thus they suf-
fer from the same problem as the cross-sectionaldemand
studies discussedearlier:the resultsconfusethe effectsof
pricesand qualityon utilizationwith the effects of utiliza-
tion on governmentgeographicalpricingandqualitypolicy.
Manylongitudinal studiesdo not sufferfromthismethod-
ologicalproblem.For example,Litvackand Bodart (1993)
carriedout a fieldexperimentin the Adamaouaprovinceof
Cameroonthat investigatedpeople's willingnessto payfor
drugs.In one treatment area facilitieschargeduser feesto
finance a revolvingdrug fund, whichincreaseddrug avail-
ability.As a resultutilizationincreasedin the treatmentarea
relativeto utilizationin a controlarea.Thusit appearsthat
consumersin the treatmentareawerewillingto payfordrugs;
hence their utilizationincreased.Usinga similarmethodol-
ogy,Yazbeck and Leighton(1995)investigatedthe effectof
introducingfees to financebetter-qualityprenatal care in
Niger.Theyfoundthat prenatalcareenrollmentsin the treat-
ment area increased relativeto the control area, and that
the increasewas greatestamongthe poor.However,in an
analysiswithouta controlgroup,HaddadandFournier(1995)
found that user feesled to a drop in utilizationin Zaire-
despite the fact that the supplyof drugs and the physical
conditionof the facilitywere increasedat the same time.
Measuringqualityis a problemin allthese studies.Most
of the studies used structuralmeasuresof quality,such as
availabilityof drugs, personnel, physical infrastructure,
and equipment. But it is not what people have that mat-
ters, it is what theydo with it. Indeed, severalstudieshave
shownthat processmeasuresof qualityare better predic-
tors of healthoutcomesthan structuralmeasures(Peabody,
Gertler,and Liebowitz1995).Moreover,the most impor-
tant measure, availabilityof drugs, confoundssupplyand
demandeffects.Facilitiesmayhavedrug shortagesbecause
they provide high-qualityservices,and have high utiliza-
tion that depletesthe drug stocks.
TABLE2
Ownandtotal priceelasticities
in Indonesia
(percentagechangewith a I percent increasein public healthcenter fees)
Children
Urban Rural
Visits
to healthcenters -1.07 -0.63
Visits
to all providers -0.48 -0.49
Source:
GertlerandMolyneaux
1997.
STRATEGIESFOR PRICING PUBLICLY PROVIDED HEALTHI SERVICES
137
Effectoffees on accessandhealth outcomes
The precedingsubsectionfocusedon the budgetaryimpli-
cationsof differentuser fee policiesin terms of measuring
the extentto whichraisingfeesincreasesgovemment'sabil-
ity to mobilizeresourcesand allowsflexibilityin allocating
subsidies.Although knowing the own price elasticityof
demand for public facilities is essential for forecasting
expectedrevenues,it is insufficientforevaluatingthe effect
higheruserfeeshaveon individualwelfare.Asa steptoward
thisgoal,this subsectionreviewsthe empiricalevidenceon
the effect of prices and quality on health outcomes and
accessto medicalcare.
The first question regards the effect of price increases
on overallaccess-that is,did the individualswhochoseto
stop seeking treatment at public facilitiesswitchto self-
treatmentor privatesectortreatment?Tomeasurethe effect
price increaseshaveon access,we examinethe effectthey
have on the utilizationof allproviders,public and private.
In their researchin Indonesia,Gertler and Molyneaux
(1997)examinedthe effectpublic healthcenter feeshave
on total visits, including visits to all public and private
providers(table2).Theyfound that priceelasticitiesof total
demand were less than the elasticities of health center
demand, implyingthat higher fees causedsome individu-
alsto switchto otherprovidersratherthan treat themselves.
Similarly,Bennett (1989)reports that after fees increased
in Lesotho,about half the reduction in public sector uti-
lizationwas a reallocationto privatefacilities.
Returningto Indonesia,in urban areas (wherethere are
more private alternatives)the total visit elasticityis about
half the health center elasticityThis finding impliesthat
about half of the reductionin utilizationswitchedto other
providersand about half to self-treatment.In rural areas
the total elasticityis about two-thirds the health center
Adults Seniors
Urban Rural Urban Rural
-1.04 -0.01 -0.45 -0.47
-0.70 0.0 -0.22 -0.39
elasticity,implyingthat a much larger percentage of patients
switched to self-treatment than in urban areas. These results
suggest that public sector fee increases reduce access more
in rural areas, where there are fewer private alternatives.
Simulatingthe effect of fee increases in public facilitieson
total utilization also requires information on how competing
private providers respond to the increased prices at govern-
ment facilities.When government providersraise their prices,
some patients may shift to the private sector, which may lead
private providers to respond to the increased demand by
raisingtheir prices. The extent of the price response depends
on the extent of the increase in demand and the slope of the
private provider supply curve.The larger is the private sector
price response, the fewer is the number of people who will
switch to the private sector, implyingthat more individuals
will choose self-treatment or remain in the public sector.
Private sector price responses are likelyto be veryimpor-
tant. When public user fees were increased experimentally
in Indonesia, Gertler and Molyneaux (1997) found that
while the fee increases caused some individuals to substi-
tute self-treatment for care at public facilities,others turned
to the privatesector instead.The resultingincreaseindemand
caused private doctors and private nurses and paramedics
to increase their fees in response to the increased demand
(table 3). In general, private sector responses were greater
in semiurban and rural areas, where there is more direct
competition between public and privateproviders. Similarly,
private nurses and paramedics, who are closer substitutes
for public primary care facilities, had larger relative price
responses than private doctors. The price elasticityestimates
shown in table 2 reflect both the increases in public sector
fees and the consequent increases in private sector fees.
Reductions in total utilization that are caused by fee
increases can have negative health effects. In the Indonesia
user fee experiment, Gertler and Molyneaux (1997) found
that the observed reductions in utilization were not only
for minor illnesses, but also for medical problems that mea-
surably affect health status indicators. The fee increase
caused an increasein the duration of illnessfor allage groups
and in illness symptoms associated with infectious dis-
eases. These results indicate that an important channel
through which prices hurt health is by delaying treatment
to the point of reducing the efficacy of medical interven-
tion. In an extreme case this may cause people to delay seek-
ing help for emergency care. In addition, the fee increases
impaired older (50 years and above) Indonesians' ability to
function physically.The increase in fees had a large enough
negative effect on health that it reduced labor force par-
ticipation among women (Dow and others 1997).
In addition, empirical evidence suggests that increases
in access and quality improve health outcomes. In a cross-
sectional analysis of household data, Benefo and Schultz
(1994) found that child mortality was lower among famni-
lies that lived doser to government health facilities in C6te
d'Ivoire and in Ghana. They also found that a doubling of
drug prices was associated with a 50 percent increase in
child mortality. Thomas, Lavy, and Strauss (forthcoming)
found in an analysisof cross-sectional data from Ghana that
improving drug supplies significantly improves the nutri-
tional status of children. In an analysis of cross-sectional
data fromJamaica, Peabody, Gertler, and Liebowitz (1996)
found that the birth weight of babies was 500 grams higher
in communities that offered good prenatal care services
using process measures of quality.

TABLE3
Private providers'price responsesto 100 percent publicsectorfee increasesin Indonesia
(percentagechange)

Urbanarea Semiurbanhealth Ruralhealth

healthcenters centersand subcenters subcenters
Privatedoctors
Healthcenterfees 4.4 18.4
Healthsubcenterfees - 3.5 20.1

Privatenurses
andparamedics
Healthcenterfees 23.8 9.5
Healthsubcenterfees - 16.7 57.9
Source:
GertlerandMolyneaux
1997.

INNOVATIONS IN HEAL-ri CARE FINANCING

138
 Aswith the cross-sectionalstudiesof demand,however,
it is hard to say howmuch the associationsbetweenhealth
outcomesand quality of care reflect the impact of quality
and accesson health outcomes or the effect of outcomes
on governmentpolicytowardthe geographicalallocationof
facilitiesand quality.Frankenberg (1995), in one of the
few explicitattemptsto determinethe directionof causal-
ity,usedlongitudinaldata fromIndonesiato showthat infant
mortalitywas lower in familieslocated near public health
centers.
These resultssuggestthat there are real returns to pub-
lic programsand public subsidiesin terms of health out-
comes.Raisingfees-thereby loweringsubsidies-can have
negativehealthconsequences.Thusif governmentschoose
to raise fees, unless the freed subsidiesare reallocatedto
moreeffectiveprograms,healthoutcomesmaydeteriorate.
Evidence suggeststhat investingthe subsidies in better
accessand qualitycan improvehealth outcomes.
How ShouldGovernmentsSet Fees?
While the above analysissuggeststhat governmentsmay
be able to mobilizeprivate resourcesto cofinancepublic
programs,it does not sayhow governmentsshouldchoose
the optimal combination of user fees and allocation of
publicsubsidiesacrossprograms.Optimalpolicyneeds to
be based on what is best for furtheringsocial objectives,
subjectto the limitsimposedby medical,behavioral,and
economicconstraints.Three groupsof objectivesare often
cited: improvinghealth status,improvingequity in access
to medicalcare,andirnprovingindividuals'insuranceagainst
the risk of largefinanciallossesdue to illhealth (Hammer
and Berman 1996).In manycasesthe policyprescriptions
that best achieveeach of these objectivesare in conflict.
Thus, because resources are limited, governmentsmust
maketradeoffsin financingprogramsand base those deci-
sions on the relativevalueit placeson each objective.
Regardlessof the objectivesa governmentis pursuing
throughits involvementin the healthsector,mostcountries
havelimitedpublicresourcesto investin health.In allocat-
ing their limitedbudgetsgovernmentofficialsand program
managersmust use the resourceswiselyand get as closeas
possibleto their goals within a fixed budget. In previous
sectionswe consideredthe possibilityof charginguser fees
STRATEGIES FOR PRICING PUBLICLY PROVIDED HEALTH SERVICES
139
to supplementthe publicsubsidiesallocatedfrom general
tax revenuesto financepublicprograms.Here we consider
the task of settinguser chargesand allocatingthe totalbud-
get (public subsidiesplus user fee revenues)while maxi-
mizinggovernmentobjectives.Thissectionconsidersoptimal
policywhenhealthoutcomesare the mainobjectiveof pub-
lic policyand discusseshow the policyshouldbe adjusted
whenequityand insuranceconsiderationsare added.
Whatdopricesdo?
One of the mainmessagesof thispaper is that the levelof
feechargeddeterminesthe degreeto whicha particularpro-
gram(orgroup)is subsidized.Much ofthe discussionabout
userfeesis couchedin termsofwhethergovernmentshould
raise fees closer to the cost of providingthe service.To
helpguide optimalfee policy,weturn this questionaround
and askwhengovernmentshouldsubsidizeservicesin order
to lowerpricesbelowthe costof providingthe service.
Formostcommodities, thereisa certain"rightness"about
the level of use (demand) when consumersface a price
reflectingthe true resource cost of producing it. People
ask themselveswhether it is worth buyingthe commodity
givenall the other thingsthey can do with their money.If
the answeris yes,then theyget more valueout of the com-
moditythan it costssocietyto produceit. If not, then they
decideto spend their money on somethingelse that they
think is more worthwhilefor the price.Individualsdo not
purchase good or serviceswhose prices exceed the value
they place on it. Similarly,if someone is in a position to
providea good or service,they willdo so onlyif the price
exceedsthe costof their makingit. So pricestellboth pro-
ducersand consumershowmuch somethingreallycoststo
produce and how much peoplereallyvalueit.
There are severalwaysprices can help guide resources
efficientlyin the health sector.First, people often bypass
lower-levelclinicsto go to hospitalsevenwhen the clinic
could have handled their problem. They do so because
theywillget better care at the hospital,and if both are free
or have the same price, theyhave no incentivenot to use
the hospital.If pricesreflectedservicecosts,hospitalprices
wouldbe higherthan clinicprices,and onlythose whoval-
uedthe hospitalservicemore thanits unit costwouldbypass
the clinic.
 A second example is when prices limit the use of ser-
vices by people who do not think their health problem is
serious enough to be worth the trouble and cost of seeking
help. Seeking help takes up real resources in terms of the
time of trained professionals as well as increased conges-
tion and waiting time for others. If prices are less than the
cost of providing treatment, then treating minor problems
whose value is less than the cost is inefficient. The main
point is that prices make people's choices efficient-putting
resources where they are most valuable to them.
What is wrong with this picture? For any of several mar-
ket failure reasons (see below), the amount demanded or
the value placed on goods may not be "right." The value to
societyfor seeking care or providing a service may be higher
than that privately judged by the individual. If that is the
case, society benefits if more people use the service than
would based on individuals' private benefits. Thus public
subsidies to lower the price are justified. How much of a
subsidy depends on the degree to which the social value
exceeds the privatevalue. The benefits that can be obtained
from subsidies also depend on how responsive to prices con-
sumers are. The more price elastic is demand, the greater
is the social benefit from a given subsidy.
As with government intervention in any sector, the allo-
cation of government subsidies needs to be justifiedin terms
of the benefit the investment has for society above and
beyond what would have happened without public inter-
vention. The way to assess the benefit of a proposed pub-
lic intervention is to identify the failures of private markets
and quantify the loss from these failures. Priorities should
be based on the degree to which the subsidy ameliorates
these losses and the importance government places on those
losses and the individuals who incur them. Important mar-
ket failures in the health sector that justify public subsidy
include public goods, inequityin access to medical care, and
insurance market failure due to asymmetric information.
Improving
health status
Most countries' ministries of health try to improve health
by funding public health activitiesand deliveringhealth care
services in public health centers and hospitals. This is why
policymakers become concerned when fee increases lead
to big reductions in utilization.
INNOVATIONS IN HEALTH CARE FINANCING
140
One of the main ways in which ministries of health can
improve health status is to encourage or discourage uti-
lization by the way they set the price of health care ser-
vices. Ministries may want to stratify their price subsidies
to encourage utilization of specific services (immunizations,
prenatal care) and by specific groups (the poor). But not
all increases in utilization are from new utilization. Some
may be substituting for private sector services (or other, less
public services) that have been substituted for by the sub-
sidized public services. The degree to which the increased
utilization improves health depends on the efficacy of the
additional health care consumed. To determine the amount
of additional care consumed, we have to subtract any reduc-
tion in private sector services that the individual would have
purchased had there been no subsidy.
However, ministries of health do not haveunlimited pub-
lic resources that they can spend on their various activities.
Ministries have fixed budgets that they can relax only by
charging fees for their services. Although price increases
may generate substantial revenue, they also deter people
from seeking care who might have sought the treatment
when it was priced lower.
To translate this discussion into a set of policy rules, we
need to establish the link between policy levers and policy
objectives. Thus, since a ministry of health's main objec-
tive is in terms of health and its policy levers are in terms
of prices that determine the level of private resources and
the allocation of public subsidies, we need to establish
links between health and prices. By altering prices, gov-
ernments affect the utilization of medical care and the
amount of money spent on publichealth activities.Utilization
of medical care and public health programs influenceshealth
outcomes.
With improved health as the objective and the links
between policy and objectives established, we can identify
four pricing principles that need to be balanced for the
government to get the best health outcomes for its subsi-
dies (see Hammer forthcoming for formal derivation and
details of the pricing rules).
* Subsidiesshouldbe higherforservicesforwhichpublic care
is better than private-that is, for services that yield the
best health outcomes compared with people's alterna-
tives. If the alternativeto public care is a traditional healer
of dubious quality, fees should be raised with great cau-
 tion. If the alternativeis a reasonableprivate sector (in
Indonesiathe private sectorconsistsof publicproviders
working duringthe afternoon),raisingfees may make
more sense.If healthisthe objective,itis betterto encour-
age peopleto use the most productiveservicesat subsi-
dizedprices.
* Subsidies shouldbe higherforservicesforwhichtotal(pub-
licandprivate)demandis mostelasticwith respecttofees
inpublicfacilities.Governmentscannotmandatethe use
of healthcare.They can onlyprovideincentivesfor use.
Subsidies encourageuse of a serviceby loweringthe
price.The more price elasticis demand,the largeris the
increasein utilizationfromagivenpricesubsidy.However,
demand may be more price elasticfor less efficacious
services.Thus subsidiesshould be higher for services
that produce the best health outcomes.These services
are mostsuccessfulin producingthe mosthealthbecause
of the combination of efficacyand of the volume of
patientsgeneratedby the introductionof the subsidy.
e Subsidiesshould be higherfor people whose demandis
morepriceelastic.For similarreasonsas in the previous
point, subsidies produce better health outcomes for
groupsfor whomthe subsidyis likelyto encourageuse.
This findingimpliesthat subsidiesshouldbe higherfor
poor individualswhose demand is more price elastic.
An interesting implicationof this pricing principle is
that it is optimalto lowerprices forthe poor evenif the
governmentis concernedwith neither equity nor wel-
fare but with health status per se.
Subsidiesshouldbe higherfor servicesand in areaswhere
there arefew private sectoralternatives(competition).
Subsidieswillproduce substantiallyweakerhealthout-
comesif theyonlycauseindividualsto substituteout of
the privatesectorinto the publicsector.The best health
outcomesare achievedwhen subsidiesencouragenew
utilization,sothat illnessesthat wouldnot otherwisehave
been treated are treated. Thus certain preventiveser-
vices and health care servicesin rural areas should be
more heavilysubsidizedbecausethere are fewerprivate
sector alternatives.
Thefirstthreeprinciplesarguethat settingpricesfor ser-
vices or for particulargroupsmust balancetwo competing
goals:limitingthe adversehealth effect from a reduction
in utilizationand mobilizingresourcesthat can be used to
STRATEGIES
FOR PRICINGPUBLICLYPROVIDEDHEALTHSERVICES
141
subsidizeother activitiesor groups and providemore ser-
vices.Servicesor groupsfor whichpricesdiscouragelarge
numbersof (price elastic)individualsfrom getting treat-
ment shouldhavelowerprices.Conversely,whendemand
is more price inelastic,higher prices affect health status
less and mobilizemore revenuethat can be used to cross-
subsidizeother beneficialactivities.The basic idea in set-
ting pricesis to push public subsidiesas far as they can go
in achievinghealth gains.Thus price subsidiesshouldbe
assessedin terms of their effect on health outcomesand
their impact on the budget, rather than relative to the
resourcecosts of servicedelivery.
The firstand fourthprinciplepoint out that interaction
betweenthe publicand privatesectorsis crucialin setting
prices.If the private sector offers comparablequalityser-
vices and individualsare willingto pay the private sector
price,governmentsubsidieswillnot improvehealth.Allthey
willdo is cause individualsto substitutepublicfor private
care.In this casethe ministryof health shouldnot provide
the care, or it should at leastprice the servicesso that few
subsidiesare absorbed.This is clearlythe case for luxury
roomsin hospitalsbecausethe rich are the onlygroup that
usessuchrooms, andthese servicesare usuallyavailablein
the private sector.
When the public sector lowers its prices because of
subsidiesand drawspatients awayfromthe privatesector,
it is in essencecompetingwith the privatesector.Subsidies
to public providers lower the profitability of private
providers.Public subsidiesaffect the prices that the pri-
vate sector can chargeand raisespeculationon whether it
is profitable for private providers to locate in the same
area as the public provider.The fact that there are no pri-
vate providersin an area does not necessarilyindicatethat
privateproviderswouldnot servethe area if therewere no
public servicesavailable-it merelyindicatesthat the pri-
vate sector does not find the area profitable.As the pub-
lic sector raises its prices, however, the competitive
constraintson the private sector are eased.As a result the
privatesectormayraiseits prices,andnewprivateproviders
may enter the market. These supplyresponseswillaffect
the demandfor publicand private servicesand, therefore,
affect health outcomes and resource mobilization.Thus
these supplyresponsesshouldbe factoredinto the setting
of public sector prices.
 One clear messageis that the governmentshould sub-
sidizeservicesthat the private sectoris unlikelyto provide.
Public goods are the most obviouscandidatesfor public
subsidies.Apure publicgoodis oneforwhichaprivatemar-
ket cannot existbecausebeneficiariescannot be made to
payforbenefits(nonexcludable)and one person'sbenefits
are not reducedwhenothersbenefit aswell(nonrivalrous).
Health sectorexamplesinclude someforms of vectorcon-
trol (for example,draining swamps),some forms of sani-
tation (especiallyin urban areas), and provisionof health
informationand education for activitiessuch as washing
hands, whichhave no product associatedwith them that
advertisingwouldpromote.Research,epidemiologicalsur-
veillance,and food and drug safetyare other examples.A
health servicehas a positiveexternalityif its use generates
benefitsto societyaboveand beyondthe benefit to the pri-
vateindividual.The most commonexternalityin the health
sector comesfrom preventionand treatment of infectious
diseases.In the Gambia,for example,pesticide-treated bed-
nets reducedthe incidenceof malariaevenamongpeople
whohad not used them, suggestingthat the societalbene-
fit from bed-nets was greater than the private benefits
(TropicalDiseaseResearchProgram1995).
Left to their own devices,individualswillprevent and
treat infectiousdiseasesless than is sociallyoptimal.Many
individualsare unwillingto paythe full cost of immuniza-
tionbecausetheyknowthat theywillbe protectedifenough
other people are immunized.4 Even when immunization
offersimportantmedicalbenefits,the costmayimpedeindi-
vidualsfrom seeking treatment soon enough to prevent
the spreadto other individualsor from completingthe full
course of treatment. When drug therapies are not com-
pleted, it maylead to a resurgence of the disease, to an
increasein transmission,and to resistanceto knowndrug
therapies.For example,tuberculosisis a virulent,commu-
nicabledisease,and althoughthe drug therapyis available
and effective,it is expensive.Individualsfeel better after
partialtreatment and tend to want to stop treatment long
beforethe courseof drugsiscompleted.Theyremaina pub-
lichazardbecausetheycan stilltransmitthe disease.Toget
individualsto obtainproperpreventionand treatment,the
governmentneedsto use publicsubsidiesto lowerthe price
of these servicesto encourageutilization.In some cases
the governmentmust fullysubsidizethe activities.
INNOVATIONS
IN HEALTIICAREFINANCING
142
Somecountriesfullysubsidizethe preventionand treat-
ment of communicablediseases.For example,Creese and
Kutzin(1995)report that Ethiopia,Ghana,Jamaica,Mali,
Niger,Papua New Guinea, and Zimbabwedo not charge
for the treatmentof tuberculosis.Moreover,allthese coun-
triesexpectPapuaNewGuineado not chargefor the treat-
ment of sexuallytransmitted diseases.There is evidence
that subsidizingthe use of public goodsprogramsleads to
improvementsin utilization.In China and Zambia child
immunizationrates fell dramaticallyafter user fees were
introduced (Booth and others 1995; Sheng-Lanand oth-
ers 1994).
Taiwan(China)providesan exampleof adramaticimprove-
mentin healthindicatorsthroughpublicinvestmentsat low
incomelevels.In the 1950sTaiwan(China)was extremely
poor,with a per capitaincomeof less than $150in today's
terms.Associatedwith this low livingstandardwere wide-
spreadincidencesofinfectiousandparasiticdiseases.In 1952
the maincausesof deathwere gastritis,duodentitis,enteri-
tis, clotitis,pneumonia,and tuberculosis.About 1.2 million
people (in a populationof 7.8 million)were infectedwith
malaria.In 1962,383casesof cholerawere reported.About
90 of the populationwasinfectedwithhepatitisB byage 40,
and 15 to 20 percentwere hepatitisB carriers.The infant
mortalityrate was 45 per 1,000live births; the maternal
mortalityrate was 197per 100,000livebirths.
Thegovermment respondedto these problemswith exten-
siveimprovementsin water supplyand sanitation,disease
controlprograms,and immunizationcampaigns.Free vac-
cinationsagainstthe main infectiousdiseaseswere made
availableto infantsandpreschoolchildren.Toexpandimmu-
nization,healtheducation,and treatment,the government
alsoset up primarycare facilitiesthroughoutthe country.
These efforts, combinedwith better livingconditions,
were ableto control infectiousdiseasesby the mid-1960s.
No cases of smallpoxor rabies have been reported since
1959.In 1965Taiwan(China)wasdeclaredfree of malaria
bythe WorldHealth Organization.By 1970Taiwan(China)
had health indicators similarto those in most industrial
countriestoday.Lifeexpectancyincreasedfrom55 yearsin
1951to 69 years in 1970.Neonatalmortalityfell by more
than halfbetween1955and 1970.Infantmortalitydropped
by about two-thirdsover the same period, and maternal
mortalityhad similarimprovements.
 The most strikingfeature of Taiwan's(China) achieve-
ments is that they were realizeddespite very low income
levels.In 1970 real per capita income was $389 (in 1993
dollars),whichwould place 1970 Taiwan(China) among
the poorestcountriesin today'sworld.A secondclearpoint
is that governmentsshould not expend resourceswherea
well-functioning marketexists.If the privatesectorprovides
an acceptable and affordable alternativeto a public ser-
vice, there is littlejustificationfor public sector subsidies
for that service.This situationis most likelyin the market
foroutpatientservicesanddrugsfornoncommunicabledis-
eases.In this casethe benefitsof treatment accruemostly
to the individual;thus there shouldbe a privatemarket for
theseservices.Moreover,thissituationjustifiesshiftingmore
subsidiesto rural areaswherethere are fewerprivatealter-
natives.
A possiblerole for governmentin the market for indi-
vidual(non-publicgood)healthcare servicessuch as cura-
tive care is in caseswhereprivate providershave sufficient
market power to set prices above marginal (incremental
unit) costs, as in the case of a monopoly.5When private
prices are higher than marginalcosts, utilizationis lower
than wouldbe warrantedby the cost of providingthe ser-
vice and there is (deadweight)loss in economicefficiency.
In this case the governmentcould either regulateprivate
prices or directlyprovide services(pricedat cost) through
the public sector.
A third messageis that prices shouldbe used to direct
individuals to the most efficient treatment location.
Illnessesand prevention activitiesthat can be treated at
health clinicsshould not be treated at hospitals.Because
demand for the treatment of more serious illnesses is
less price elastic,increasingthe price of hospitalcare rel-
ative to health center care will induce people with less
serious illnessesto not bypasshealth centers in favor of
hospitals. Such cascading systems of charges exist in a
numberof countries,includingKenya,Indonesia,Namibia,
Zambia,and Zimbabwe(Barnumand Kutzin1993).Criel
and Van Balen (1993) found that these price structures
succeeded in moving people out of hospitals and into
health centers in Zaire. In Zambia and Zimbabwe,how-
ever, hospitals are still overcrowded and health centers
still underutilized.This implies that either the price dif-
ferential between health centers and hospitals is too low
STRATEGIES FOR PRICING PUBLICLY PROVIDED HEALTH SERVICES
143
or that the health centers provide few services of real
value-that is, theyhaveno drugs and few qualifiedmed-
ical personnel. In the second case the quality-adjusted
price differentialis too low and, in any event, there is no
health benefit of sending people to health centers. This
situation would require improvementsin quality to jus-
tify keepinghealth centers open.
A fourthmessageis that governmentsshouldnot use the
sameconsultationfeesforeach diagnosisand demographic
group. Fees shouldbe lower-that is, subsidiesshould be
higher-for the preventionand treatment of illnessesthat
havelargepublichealthexternalitiesandforwhichdemand
is most elastic.Across-the-boardfee increasesin Kenyaled
to a 40 percentreductionin the treatment of sexuallytrans-
mitteddiseases;similarresultswerefoundin Zambia(WHO
1994).Similarly, across-the-boardfeeincreasesledto reduc-
tionsin childimmunizationsin China (Sheng-Lanandoth-
ers 1994) and Swaziland(Yoder1989).Moreover,there is
strong econometricevidencethat children's demand for
medicalcare ismore pricesensitivethan adults'.Combined
with the importanceof preventionand treatment earlyin
life, this findingsuggeststhat children's health care ser-
vices shouldreceivehigher subsidies.A similarprice elas-
ticity argumentcan be made for directingmore subsidies
to the poor.However,chargingthe poora lowerfeeis admin-
istrativelydifficult(seebelow).
Adjusting
policiesto increase
equity
Recognizingthat poorindividualsmaynot be ableto afford
health care, most countriessubsidizetheir accessto care.
In countrieswherehealth care is deliveredthrough public
deliverysystems,subsidiesare used to keep user charges
low so that even the poorest familiescan afford medical
care. Supportfor this use of publicsubsidiesis often based
on the idea that nobody,regardlessof income,should be
denied accessto basicminimalhealthcare.Althoughthese
commitmentsare not boundless,theyarepervasivethrough-
out the world.Sucharrangementshaveimportantimplica-
tions in that redistributionpolicies are inseparable from
health care policies.Unlessprivate health care and insur-
ance marketsare able to guaranteeuniversalaccess,gov-
ernmentswillinterveneand subsidizecertain servicesand
groupsto varyingextents.
 The health sector is not an effective vehicle for general
poverty alleviation, however. Studies of the demand for
health care show it to be an income-elastic good-that is,
the rich spend a larger portion of their income on health
care than do the poor (Gertler and van der Gaag 1990;
Baker and van der Gaag 1993).6 Thus health care subsi-
dies accrue more to the rich more than to the poor. Other
goods that are more income inelastic (such as food) would
be better vehicles for general poverty alleviation.
Much of the concern about user fees derives from the
fear that increasing fees mayreduce utilization by the poor-
a reasonable concern given the strong empirical evidence
that poor people's demand for health care is more price
elastic than rich people's. Moreover, this finding extends
to the case where fees are used to generate revenues that
finance improvements in the quality of and access to cura-
tive care. The extent to which this policy improves welfare
depends on how willing individuals are to pay for the qual-
ity and access improvements. If individuals are willing to
pay the full cost of the improvement, the improvements can
be fullyfinanced through increased user fees without reduc-
tions in utilization. But if the rich are willing to pay but the
poor are not, this policycould lead to a reallocation of pub-
lic subsidies from the poor to the rich. Thus equity propo-
nents are concerned that increased user fees would become
a financial barrier to the poor and reduce their access to
care (Cornia,Jolly, and Stewart 1987; Gilson 1988). In this
case there is a tradeoff between using subsidies to pursue
equitable access to medical care and overall improvements
in health.
The current situation needs to be remedied because most
curative public expenditures are used to care for the non-
poor. Moreover, the poor actually pay higher prices than the
nonpoor when transport and other time costs are taken into
account. This is because the current geographic distribu-
tion of public facilities requires the poor to travel much
further than the nonpoor. These higher prices are in part
responsible for the poor having low utilization rates and
obtaining a small share of public subsidies.
This subsection considers pricing policy in the context
of the government expanding its objectives beyond health
to include equity concerns in its objectives for pricing pol-
icy.We begin with the common approach of using across-
the-board subsidies, which are used by many countries, and
INNOVATIONS IN HEALTII CARE FINANCING
144
then move to price discrimination strategies that try to
exempt the poor from paying fees.
Across-the-boardsubsidies.Many governments try to pro-
mote equity by subsidizing the public health care system.
Because low-income countries have trouble implementing
means testing (that is, identifying the poor individually by
examining their financial resources), they keep fees low for
everyone. This approach amounts to across-the-board sub-
sidiesfrom the averagetaxpayer to the averageuser of health
facilities. If the average user is poorer than the average tax-
payer, there is a net redistribution of income.
Many countries, however, allocate most public subsi-
dies to the services used least by the poor-hospital ser-
vices.These servicesare expensiveand are rationed byprice,
travel time, and social status rather than by clinical need.
As a result public subsidies tend to benefit the rich more
than the poor. Governments can better target subsidies to
the poor by more heavily subsidizing services that are used
by the poor.
Indonesia is typical of countries that try to subsidize the
poor's access to medical care through low-fee public health
care systems. The wealthiest 20 percent captures about 29
percent of government health care subsidies; the poorest
20 percent captures just 12 percent. This is partly because
the wealthy use hospital services in much greater rates than
the poor. One reason is that hospitals tend to be located in
urban areas close to the wealthy and far from the rural
poor. Moreover, hospital services are subsidized at much
higher levels than are health centers and health subcenters.7
The situation is similar in Vietnam-the allocation of
public subsidies increaseswith income (WorldBank 1995b).
These results are driven by the fact that the rich capture a
much larger share of both hospital inpatient and outpa-
tient subsidies. This is because they use more hospital ser-
vices,and hospital servicesreceivethe highestunit subsidies.
Although the poor use commune health centers at much
greater rates than the nonpoor, this has little impact on the
benefit incidence distribution because public subsidies to
commune health centers account for a small portion of pub-
lic expenditures.
The subsidies that leak to the nonpoor are a major cost
of subsidizingthe poor with across-the-board subsidies. The
greater is the income elasticity of demand, the higher is
thiscost of targeting.Jamaica,likeIndonesiaand Vietnam,
heavilysubsidizeshospitalcare.In order to targetone dol-
lar to the poor,the governmentmust givethe nonpoorabout
$3.25in subsidies(Gertlerand Sturm1997).Similarly, van
der Gaag(1995)showsthat whileespousingequityasa goal,
countriessuch as China, C6ted'Ivoire,Peru, and Tanzania
alsoprovidehighersubsidiesto servicesused bythe wealthy.
Solonand others (1991)showsthat high-incomeindividu-
als in the Philippinesreceivemuch more in public health
care benefitsthan theypay in taxes,
Pice discrimination. The extentto whichthe government
is able to price discriminateand only raise fees that the
nonpoor pay mitigates the severityof this health-equity
tradeoff.8 Toimproveequity,the governmentmustdevelop
policiesthat lowerthe pricepaid bythe poor relativeto the
price paid by the nonpoor by evenmore than is indicated
by the optimalpricingpoliciesdevelopedin the previous
section.There are a number of waysto do this.
The government'sabilityto implementa pricingpolicy
that maximizeshealthcare outcomesand redistributessub-
sidiestoward the poor depends on its abilityto identify
the poor (in order to price discriminate and target pro-
grams). Here we considerfour common types of target-
ing: individual means testing, geographic targeting,
self-selection,and indicatortargeting.The targetingeffec-
tivenesscriteria:
* Type1 error.failingto exemptsomeonewho shouldbe
exempted. The greater is the type 1 error, the fewer
poor are protectedby the pricediscriminationmethod.
An extremeexampleof type 1 error would be if facili-
tieschargedeveryonethe fullcostof deliveringa service.
In this casetype 1 errorwould be 100percent.
* Type2 error:exempting someone who should not be
exempted.The greateris the type 2 error,the greateris
the leakageof potentialrevenuesfrom the nonpoor and
the lowerare the subsidiesthat reachthe poor.Anextreme
exampleoftype 2 errorwouldbe if everyoneweregiven
free care.In this caseallpotentialrevenuewouldbe lost
and type2 error would be 100 percent.
* Administrativecosts:the costs of identifyingthe poor
and implementingprice discrimnination can swampall
the gains from price discrimination.There are dimin-
ishingreturns to makingprice discriminationmethods
STRATEGIES
FOR PRICINGPUBLICLYPROVIDEDHEALTHSERVICES
145
more precise.Administrativemethods varyfrom inex-
pensive proceduressuch as geographicprice discrimi-
nation and targeting by age and gender to costly
proceduressuchasa slidingfee systemwithsocialworker
verification.The additionalbenefitsof better targeting
methodsneedto be comparedwiththe additionaladmin-
istrativecostsof implementingthem.
Individualpricediscrimination based on meanstestingis
the ideal method for minimizingthe revenue loss from
protectingthe poor.However,administrativecostsand past
experiencesmake means testingineffectivein most coun-
tries.For example,in Indonesiaand Vietnamthe poor can
haveuserfeeswaivedthroughan affidavitofindigence.Few
people,however,seemto take advantageof thismechanism
(WorldBank 1995aand 1995b).It is not clear whythese
systemsare failing.There are severalpossibilities:people
maynot know about the benefit,pricesmaybe solow that
the benefit is not worth the opportunitycost of obtaining
it, local officialsmay be charginga fee to issue the affi-
davit,facilitiesmaychargea fee to acceptthe affidavit,and
there may be a social stigma associated with using the
affidavit.
One of the biggestobstaclesto meanstestingliesin mea-
suringeconomicwell-beingin an economywheremostpeo-
ple pay no income tax and a sizable portion of economic
resourcesare home produced.Withoutaccurate,fast, and
administrativelysimplemethods of identifyingthe poor,
an individualexemptionmechanismmayexempttoo many
peopleand sacrificesubstantialrevenues.Moreimportant,
identifyingthe poorwhenthey seektreatment is costlyand
difficult.It is certainlybeyondthe capacityof health care
providersto do so, and is impracticaloutside the context
of ageneralgovernmentwide meanstestingprogram.Facility-
based individualexemptionprogramsin the health sector
are too costly,very difficultto administer,and typicallyare
not good at identifyingthe poor in all developingregions
(Booth and others 1995; Chaulagai1995;Ensor and San
1995; McPake and others 1993;Mwabu, Mwanzia,and
Liambila1995;NolanandTubat 1995;Richardsonandoth-
ers 1992;Stinson1982;and Vogel1988).
Geographic targetingis an alternativewayof implement-
ing a pricingpolicythat protects the poor. This approach
attempts to tailor the fee structure to the socioeconomic
compositionof the populationservedby each health care
facility.If the poor live in more rural than in urban areas
and facilitiesare locatedin more urban than in ruralareas,
with uniform fees the poor face higher accesscosts than
the nonpoor.Thus geographictargetingattempts to locate
facilitiescloserto where the poor live.Otherwisethe fees
at facilitiesthat serve the poor must be lowerin order to
compensatefor the higher time costs so that the price of
accessis the same for the poor and the nonpoor.
With this in mind, facilitiesthat serve primarilypoor
householdswould chargezero or near-zerofees, and facil-
itiesthat serveprimarilynonpoorhouseholdswouldcharge
higherfees. The fees chargedby a facilitywould risewith
the averageeconomicstatus of the householdsin its ser-
viceregion.Indeed, facilitiesin wealthyareascould charge
fees equalto or in excessof unit costs.A facility-levelfee
scheduleincreasingwith the economicstatusof the house-
holds in the facility'sserviceregionwouldimplythat gov-
emnmentsubsidiesare pro-poor in that they are largestin
the poorest areas.
In principlegeographicprice discriminationis straight-
forward;in practiceit is quitecomplex.Populationswithin
aregionarenot homogeneous.Everyregionhas somehouse-
holdswhoseincomeis belowthe government'spovertyline.
In regionswherea largeportionof the populationis poor,
the governmentcan keep fees low enoughto protect most
of the poor withoutexperiencinghighlevelsof type2 error.
But in regionswith a smallportion of poor residents,the
government must choose between forgoing substantial
revenuesfrom those able to payin order to protect a small
number of poor, or failingto protect the poor in order to
reduce revenuelossfrom the nonpoor.In thiscaseit would
be cost-effectiveto screenthe poor at healthcare facilities
or to use an individualdiscriminationmethod.
TABLE
4 vices for whichdemand is income inelastic-that is, ser-
Femaleadultmortalityratesbycauseof death
andincomegroupin China
(percentagebetween
dlying ages15and60)
Infectious Noncommunicable
Income
quartile diseases diseases Injuries
Richest 0.4 6.7 1.2
2 0.46 7.6 2.4
Poorest 1.4 8.9 2.7
Source:Murray,Yang,and Qiao 1992.
INNOVATIONS
IN HEALTHICAREFINANCING
146
Usinggeographicpricediscrimination in ruralareaswhere
the poor are concentratedis a promisingwayto protect the
poor without sacrificingconsiderablerevenue.Long travel
timespreventpeople livingin wealthierareas from switch-
ing to the lower-feefacilitiesin poorer areas once fees in
the moreaffluentareashavebeen increased.However,geo-
graphicprice discriminationhas limitedpotential in urban
areaswherethe poor livealongsidethe nonpoor and most
facilitiesare easilyaccessedby both groups.
An alternativeapproachto protectingthe urban poor is
through differentialpricingby levelof serviceand self-selec-
tion. The idea is to havelow subsidiesfor servicesvalued
and used mostlyby the nonpoor, and high subsidies for
servicesused mostlyby the poor.These are the servicesfor
whichdemandis incomeinelastic.
One approach is to shift subsidiestoward the preven-
tion and treatment of infectiousdiseases.Since the poor
tend to sufferproportionallymore from infectiousdiseases,
subsidizingtheir treatment and preventionnot only helps
meet publichealthobjectivesbut alsoimprovesthe distrib-
ution of publicsubsidiesacrossincomegroups.For exam-
ple,table4 showsthe distributionof mortalityfromdifferent
causesacross differentincome groupsof adult women in
China.Whilepoorerwomenhavehighermortalityratesfrom
all causes, the poor die from infectiousdiseasespropor-
tionatelymore than do other incomegroups.Poorwomen
are 3.5 timesas likelyas richwomento die frominfectious
diseases,but are only 1.3times as likelyto die from non-
communicablediseases.Basicprinciplesof targeting(Besley
and Kanbur1993)suggestthat, ifthe costsof treatmentare
the same,reallocatingsubsidiesfromnoncommunicable dis-
eases to the prevention and treatment of communicable
diseaseswouldbetter targetpublic subsidiesto the poor.
As a generalrule,the governmentshouldsubsidizeser-
vicesthat are used more bythe poor and forwhichdemand
does not increase much with income. For example, in
Vietnam demand for commune health centers is highly
incomeinelastic,whiledemandfor hospitalcare is income
elastic(Gertlerand Litvack 1996).Thuskeepingsubsidies
highforhealthcentercare'andlowforhospitaloutpatient
serviceswillbettertargetsubsidesto thepoor.
Vietnam'sresults are likelyto be true for most coun-
tries-that is, demand for health center care is the most
income inelastic, especiallyin rural areas. This finding sug-
gests that increasing subsidies for rural health centers best
targets subsidies to the poor. It also suggests a general pric-
ing structure in which fees are lower (and subsidies higher)
if the patient enters the system at the lowest level, and are
progressively higher the further up the system the patient
enters. Thus if an individual first goes to a commune health
center and requires a higher level of care at a hospital, the
registration fees should be waived (or at least lowered) at
the hospital. This pricing structure provides an affordable
portal of entry into the health care system (through the com-
mune health centers) and allows people who are willing to
pay to go directly to higher levels of care. Since the non-
poor are willing to pay to bypass the lower levels, they will
be charged higher prices and receive lower subsidies.
Policies can take advantage of self-selection if a wider
range of instruments is considered. If adequate medical
treatment is maintained, government facilities may delib-
eratelyoffer fewer amenities sothat onlythe poor will choose
to use them. This approach, however, requires accepting
different levels of service (at least from the consumers' sat-
isfaction standpoint) in order to concentrate more resources
on the poor.
Finally, targeting identifiable groups through indicator
targetingis one way to charge those most able to pay at least
the full cost of care. One group that may be able to afford
care is the insured population. Insurance status is a good
indicator of ability to pay because the insured are wealth-
ier than the general population. For example, in Indonesia
civil servants are covered by insurance, and almost all civil
servants are in the top half of the income distribution (World
Bank 1995a). However, the insurance company pays less
than the full cost of care, so the public system still subsi-
dizes the wealthy insured population. This situation exists
in most countries with mixed public-private systems. By
charging the full cost of care to insured patients, public
systems could reduce subsidies to the nonpoor and reallo-
cate them to services used by the poor.
Adjusting allocationto insureagainst
resource financialrisk
The inherent uncertainty in health status is the classic rea-
son most industrial countries intervene in health markets
(Arrow 1963). No one knows what tomorrow will bring.
STRATEGIESFOR PRICING PUBLICLY PROVIDED HEALTH SERVICES
147
Seeminglyhealthyindividualscan be struck by cancer,injured
in accidents, or experience bouts of severe diarrhea. This
uncertainty is compounded the longer one looks into the
future and the less one knows about one's current health.
While most families are able to finance routine care out of
pocket, few are able to finance rare but expensive incidents.
In fact, all countries' health care expenditures are extremely
skewed in that a small portion of the population accounts
for a large portion of total expenditures. Thus, while most
families have only small expenditures in a given year,a small
number have very large expenditures.
Risk-averse individuals prefer to have predictable health
care expenditures. Predictability relieves the worry of how
to finance costly unexpected illnesses and allows families
to better plan other consumption. Thus individuals will seek
to insure themselves against the financial loss associated
with uncertain illness.In the absence of formal health insur-
ance people have to informally finance the losses out of
accumulated savings, transfers from relatives and friends,
credit markets, or help from charities.
However, informal insurance seems inadequate. Using
household panel data from Indonesia, Gertler and Gruber
(1996) show that these informal sources of insurance are
insufficient for Indonesians to fully finance the costs of
severe illnesses. When illnesses are severe enough to affect
labor supply and income, their economic costs are partly
financed by a reduction in the family's consumption of non-
medical goods and services.As a result there is demand for
insurance for both the medical and income costs associ-
ated with illness.
Despite the demand, most people are unable to buy insur-
ance from private sources .because of market failure from
adverse selection (Rothschild and Stiglitz 1976). Adverse
selection arises because insurers are unable to observe het-
erogeneity in a population's health status. People are born
with different genetic makeups that make them more or
less predisposed to illness, and have different life experi-
ences in terms of exposure to environmental contagion
and accidents. For both reasons there is substantial varia-
tion in the propensityto become ill.Because insurersdo
not observe each individual's propensity to become ill,
they cannot write individual contracts. Rather, they are
forced to offer the best community-rated insurance plans.
The terms of these contracts can be quite unfavorable to
healthy individuals.Good risks (healthypeople) tend to
subsidizethe bad (unhealthypeople),andthe valueofinsur-
ance to the good risksdrops significantly.
Good riskshave
an incentive to drop out of the market, leaving the bad
risks to insure among themselves-and substantiallydri-
ving up the cost of insurance,makingit a financiallybad
dealfor both insurersand beneficiaries.In manycasesit is
sucha bad deal that the insurancemarket failsto exist.
The opposite problem is risk rating (or "cream skim-
ming"),whichoccurswhen unhealthy people are observ-
able.Competingon their abilityto select good risksleads
insurersto avoidinsuringindividualswith preexistingcon-
ditions (such as cancer or AIDS) who are "certain" bad
risks-that is, theywillhavepredictablyhigh medicalcare
expenditures.Insurers do not want to provide these indi-
vidualswithcoverageat the community-rated (average)pre-
mium. Instead they either explicitly deny coverage or
effectivelydeny coverageby charginga premiumapproxi-
matelyequal to the cost of care. In many high-riskcases
the actuariallyfaircostof insurance(expectedexpenditures
plus a loadingfactor to cover administrativecosts) maybe
prohibitivelyexpensiveandthese individualsare effectively
uninsured.9
Insurancemarketfailuredueto adverseselectionoccurs
wheninsuranceisvoluntaryrather thancompulsory. Adverse
selectionand cream-skimming do not occurwheneveryone
is in the insurancepool. Most countriescorrect for insur-
ance market failurethrougha universalpublicsystemwith
subsidizedlow prices or through compulsorysocialinsur-
ance in whichthe poor's enrollmentis subsidized.
In publicsystemsheavilysubsidizedpublichospitalspro-
vide insuranceagainstlarge financialloss associatedwith
catastrophicillness.However,publicsystemsprovidelower
levelsof insuranceiftheyprovidelower-quality servicesthan
couldbe bought in the privatesectorwith socialinsurance
funds.
The debate overwhether to increaseuser feesin hospi-
talshasignoredthe crucialrolepublicsubsidiesplayasinsur-
ance. Subsidiescan reducerisk in two ways.First, theycan
makeuncertainhealthcarecostsmorepredictablebyspread-
ingthemacrosshealthyandsicktimes.Taxesthat areincurred
in all statesof health helpfinancemedicalcare that is pur-
chasedwhensick.Thusraisinguserfeesin aworldof imper-
fect consumptioninsurancehas an importantwelfarecost:
INNOVATIONS IN HEALTII CARE FINANCING
148
higheruser fees tax families"whenthey are down,"impos-
ing higher costs at preciselythe point where the marginal
utility (valueof the next unit) of consumptionis highest.
Second,subsidiesmayhelpmitigatethe lossof incomefrom
illnessby financingmedicalcare that improveshealth and
productivity.In essence,public subsidiesrelaxcredit con-
straintson the purchaseof medicalcare that mayhelp peo-
ple get back to work faster.
Public subsidiesfor medicalcare can correct failurein
the insurancemarket,becauseprivatemarketsare unlikely
to supplyadequateinsurancebecauseof adverseselection.
Insuranceprinciplessuggestthat the subsidiesshould go
to the servicesthat provide care for the rare, high-costill-
nessesthat wreakthe most havocon householdbudgets.
Giventhat the rich disproportionatelyuse hospitalser-
vices under current systems, there is a distinct tradeoff
between equity and efficiencyin subsidizedhospitalcare.
This tradeoff can be mitigatedby enforcingstrict referral
rules requiringhigh charges for people enteringhospitals
directly,withgenerousexemptionsforpeoplewhoare prop-
erlyreferred.
Insurance carrieswith it another market failurecalled
moral hazard: insured people may use more servicesthan
they would otherwisebecausetheir price at the point of
serviceis lowerthanwithoutinsurance.In thiscasepatients
tend to consumemedicalcare beyondthe point wherethe
additionalbenefit is greater than or equalto the additional
cost. In this sensetoo many resourcesare being allocated
to treatment.The greateris the priceelasticityof demand,
the largeris the welfareloss from moral hazard.
These market failureshave very differentimplications
forpricingpolicy.Whenadverseselectionpreventsthe emer-
gence of insurancemarkets, publicpolicyshouldpromote
a pricingstructure that protects against catastrophicloss.
Such a structure typicallyincludessignificantcopayments
by the insuredfor smallexpenditureswith stop-lossprovi-
sions (capson out-of-pocketcosts) for largeexpenditures.
Thistakes care of the welfarelossfrom assumingtoo much
risk. But when moral hazard is a seriousproblem, people
shouldfacethe true costson the marginto limitoveruseof
services.The risk problem needs to be handled by rela-
tivelylarge, inframarginalpayments (Zeckhauser 1970).
Thus adverse selectionsuggeststhat an optimal payment
policywould require a large copaymentfor smallexpendi-
tures and none for largeexpenditures,while moralhazard becausefamiliesmustpay higherpremiumsat the expense
arguesfor low (or no) copaymentfor most expectedcosts of other consumptionor savings,but that lossispredictable
with a largeexposure(fullcostburden) beyondthat point. and can be spreadover the year and across individuals.
Sincesuchdifferentpolicyconclusionsfollowfrom differ- The problem of equity in accessto healthcare does not
ent institutionalstructures,a greatdealof knowledgeabout disappearwiththe introductionof socialinsurance.However,
thewaymarketsworkis essentialforgoodpolicyformulation. usinggovernmentsubsidiesto increaseequity in accessis
In most countries the allocationof public subsidiesis mucheasierwith socialinsurance-the governmentsimply
consistentwith efforts to amelioratelosses from private subsidizesthe poor's enrollment in the insuranceplans.
insurancemarketfailure,sincethe bulk of publicsubsidies For thisapproachto be budgetneutral, subsidiesprovided
is spent on hospitals.However,these subsidiesare insuffi- directlyto facilitieshaveto be reducedto financethe poor's
cientto adequatelyinsurefamiliesagainstthe risk of finan- enrollment.Facilitiesrecoup the lost revenuesby provid-
cial loss from unexpectedill health. In fact, despite large ing care forinsured patients andbeing reimbursed.In this
subsidiesto publichospitals,people are stillincurringlarge waypublicsubsidiesare bettertargetedto the poor,and the
out-of-pocketexpenditures.Moreover,GertlerandGruber facilitiesthat get the subsidiesare the ones that carefor the
(1996) show that in Indonesia, a country with a heavily poor. Administeringsuch a program is easier than price
subsidizedpublichealth care system,familiesfinancethe discrimination byfacilitiesat the timecareis neededbecause
economiccostsof illnessby reducingconsumption. it is centralized,onlyneeds to be done periodicallyoutside
the pressureof havingto treat an illness,and can be done
Is socialinsurance the answer? by a trained staff that does not have other responsibilities.
Althoughsocialinsurancecancorrect someof the prob-
Althoughit is possibleto mobilizeresourcesthroughuser lems created by resource mobilizationpolicy,it creates a
feesin waysthat improvewelfare,thereare potentiallylarge host ofother problemsthat, ifnot addressedaspart ofinsur-
coststo such a policy.Specifically,the scopefor mobilizing ance design,could outweighits benefits.The most obvious
privatefinancingfor publicexpendituresis limitedby two problem,mentionedearlier,is that socialinsurancecannot
costs:reducedinsurancecoverageagainstthe riskof finan- be voluntary.Voluntaryinsurancemarketsfail becauseof
cial loss from unexpected illnessand reduced utilization adverseselection.Forsocialinsuranceto be financialviable,
and possibleconsequentadverse health outcomes (espe- enrollmentmust be compulsory.Thisis not to saythat such
ciallyfor the poor). However,these costs are much lower plans must enroll entire populations,but rather segments
in health systemsfinancedthrough socialinsurance. suchas formalsectorworkers.In fact,mostcountriesalready
Withsocialinsurance,individualsare stillinsuredagainst have compulsorysocialinsurancefor civil servants,many
the riskof financiallossfromillnesswhengovernmentsraise haveexpandedcompulsorycoverageto wagesectoremploy-
user fees. Under socialinsurance,individualsprepaytheir ees, and a few have achieveduniversalcoverage.In addi-
medicalcare expenditures(that is, premiums)into a fund tion to adverseselection,socialinsuranceraises a number
that is used to pay for their medicalcare if and whenthey ofimportantdesignandadministration issuesthat arebeyond
become ill or injured. Thus people can avoid payingunex- the scope of this paper.
pectedfeeswhenthey are ill and so are insured againstthe
risk of financialloss from illness.With insurance,individ- Conclusion
uals'medicalcareexpenditures(premiums)arepredictable
andcanbe planned.Governmentsmobilizeprivateresources User fees have strong potential for improvingthe effi-
byraisingthe feeschargedbyinsuranceplansforhealthser- ciencyof healthcare systemsin developingcountries.Still,
vices provided to the plan's beneficiaries.Because fee theyraisemany questionsand concerns:
increasesraisepremiumsandnot the out-of-pocketcharges * Optimalfee structurescan differdramaticallyfrom one
at the time of treatment,raisingfees causesno lossin the institutionalsettingto another.Countriesdiffer signifi-
insurancevalue. Raisingfees still causesa loss in welfare cantlyin terms of the size and performanceof the pri-

STRATEGIES FOR PRICING PUBLICLY PROVIDED HEALTH SERVICES

149
 vate sector,the nature of insurance,credit, and health must comefrom targetedinterventions,as imperfectas
caremarkets,and the administrativecapacityofthe pub- currenttargetingmethodsmaybe. Health careis highly
lic sector to run or regulate healthservices.Thus much incomeelastic,and anyuniformsubsidysystemis likely
moreanalysisbased on accuratedata is neededto define to be regressive.
better policies.
In manycountriespublichealthbudgetsare determined Notes
reactively:demands for curativecare must be honored
and funds for population-basedpublic goodsare deter- 1.If a priceelasticity
issmall-between0 and1-then demand
minedresidually.In suchcasesfeesfor curativecare can issaidto be inelasticbecausethepercentagereduction
in demand
is lessthan the percentageincreasein price.Whendemandis
conserveresourcesforuseon thesepublicgoods,imnrov- inelastic,priceincreasesraiserevenuesbecausethe positiveprice
ing efficiencyand equity and increasingthe effect pub- effectis largerthanthe negativedemandeffect.If a priceelas-
lic spendinghas on health status. ticityis large-lessthan-1-then demandis saidto be elastic,
* Most countrieshavea largeprivatesectorprovidingpri- andpriceincreases reducerevenuesbecausethe negative demand
maryhealthcare.Publicfundsshouldcomplementrather effectoutweighs thepositivepriceeffect.Finally,
ifdemandisuni-
thancrowdout privaeacivityLargsubtary
elastic-equalto -1-the percentage decreasein demandis
thansiv
cerowdouprivateactiityleto
servicesmaydo littletoLarge
pensive subsiesr ineat exactlyequalto the percentage increasein priceandthereis no
Improveoverallhealth changein revenues.
status if they are strongsubstitutesfor private care. 2. Forexample,iffeesare setlowin areaswherepeoplehave
* In the many countriesin whichinsurancemarketsare seriousillnesses,the observed correlationbetweenpricesanduti-
not goingto be fixedin the near future,hospitalsshould lizationreflectsthe factthat sickerindividuals use morehealth
remain an importantitem for publicexpenditures.This careand the effectof priceonutilization.Alternatively, if facili-
helpsprotectpeoplefromcatastrophicdloss in the absence tiesarelocatednearurbanareaswhereindividuals arewealthier,
helpsporotec peoplefrotems.
caTastrophiclossinhe abmncef the correlationbetweentravelcostsandutilizationreflectsthe
relationshipbetweenincomeandutilization andtheeffectoftravel
services,witha smallnumberof peoplerequiringexpen- costsonutilization. Inbothcasesthe priceelasticity estimatesare
sivecare,impliesthat hospitalswillbe a largepart of the biased,sincetheyareconfounded withotheromittedfactorsrelated
budget. to government policychoices.
X Many countrieswill find that the optimal allocationof 3. Strictlyspeaking,the authorscomputethe compensating
public health subsidieswill involvelarge subsidiesfor differential, whichisthe amountofincomethe individual would
will prin argheah
pubclinchealthsubsicgoodi,fees s iesforfl- be willingto giveup forthe qualityimprovement sothat thereis
nonclinicalpublicgoods,feesat primaryhealthcarefacil- lnochangein welfare.
ities to conservepublic resourcesfor these goods, and 4. Transmission of an infectiousdiseaseis affectedmainlyby
largeper unit subsidiesto hospitalserviceswith a proper thenumberofpeoplewhoareimmunized. Thusindividual immu-
referralsystem.Such a systementailspatientsreferred nizationconfersa benefitto peoplewhoare notimmunized.
on the basis of clinical need and high fees for those 5. Anotherreasonforgovernment interventionin the market
usinghospitalsasthe firstpoint of contact.Thisis atlmost forindividual healthcareservicesariseswhentheprivateprovider
usinghospitalsasrthefirsubstpint
ofscontact.This
is aplms b hasmoreinformation aboutthe patient'sillnessthanthe patient
the oppositeprice and subsidyschemeas iS implied by' and the provideris not a perfectagentfor the patient.By perfect
the currentemphasison primaryhealth care. agentwemeanthat the providercaresaboutthingsotherthan
* In countrieswith social insuranceor adequatelyregu- the patient'shealth.In thiscasethe providercouldinducethe
lated privateinsurancemarkets,optimalpolicieswould patientto buymoreservicesthanhe or shemightotherwise buy.
involvefew subsidiesto hospitalservices.These would 6. The bestcandidatesforredistributionthroughsubsidies
be coveredlby actuarialyfair prepayments. aregoodsandservicesthathavelowor negativeincomeelastici-
tiesof demand-thatis, goodsandserviceswhoseuse doesnot
increasewithincome.Thesearethe thingsthat poorpeopletend
ing qualityimprovementsfrom user fees. to consumerelatively moreof thanotherthings.
* Uniformpriceincreasesreduce facilityuse more for the 7.The subsidies were206,000rupiahfora hospitalinpatient
poor than for others. Tothe extentthat this outcomeis visit,8,100rupiahfora hospitaloutpatientvisit,3,400rupiahfor
sociallyundesirable,adequate protection for the poor a healthcentervisit,and2,200rupiahfora healthsubcentervisit

INNOVATIONS
IN HEALTHCAREFINANCING
150
(WorldBank 1993). Subsidiesthroughhealth centers and sub- van Doorslaer,AdamWagstaff,and FransRutten,eds.,Equity
centersare muchmore equitablydistributedthan hospitalsubsi- in the Financeand Deliveryof Health Care:An International
dies because utilizationrates of these facilitiesare more evenly Perspective.NewYork:OxfordUniversityPress.
distributedacrossincomegroups. Barnum,Howard,and Joseph Kutzin.1993.PublicHospitalsin
8. As discussedin the previoussection,even if the govern- Developing Countries:
ResourceUse,Cost,Financing. Baltimore,
ment is not concernedwith equity,there are still good reasons Md.:Johns HopkinsUniversityPress.
(related to the desire to improve overallhealth indicators) to Benefo,KofiDarkwa,and T PaulSchultz. 1994."Determinants
price discriminatein favorof the poor. of Fertilityand Child Mortalityin C6te d'Ivoire and Ghana."
9. Thisproblemis exacerbatedbythe fact that insurancecon- Living Standards Measurement StudyWorkingPaper 103.
tracts arewritten for limitedterms (forexample,one year).Over WorldBank,Washington,D.C.
time, as more high-riskindividualscontract seriousillness,the Bennett, Sara. 1989."The Impact of the Increasein User Fees."
number of high-riskindividualsable to obtaininsurancedeclines. LesothoEpidemiological Bulletin4.
With an agingpopulationincreasingthe number of individuals Bennett, Sara, and Manengu Musambo. 1990. "Report on
withlong-termchronicillnesses,and improvements in genetictest- CommunityFinancingandDistrictManagementStrengthening
ing and long-rangediagnosis,the number of individualsdenied in Zambia."Bamako InitiativeTechnicalReport 5. United
coveragecan onlygrow. NationsChildren'sFund,BamakoInitiativeManagementUnit,
New York.
References Besley,Timothy,and StephenCoate. 1991. "PublicProvisionof
PrivateGoods and the Redistributionof Income."American
Abel-Smith,Brian,andPankajRawal.1992."Canthe PoorAfford EconomicReview81(September):989-84.
'Free' Health Services?A Case Study of Tanzania."Health Besley,Timothy,and MiguelGouveia.1994."Alternative Systems
PolicyandPlanning7(4): 329-41. ofHealth CareProvision."Economic Policy(October):201-58.
Abu-Zied,H., and W Dann. 1985."HealthUtilizationinIsmaila, Besley,Timothy,and RauiKanbur.1993."Principlesof Targeting."
Egypt."SocialScienceandMedicine21: 451-61. In MichaelLipton and Jacques van der Gaag,eds., Including
Akin,JohnS.,and CharlesC.Griffin.1985.TheDemandforPrimary the Poor.Baltimore,Md.:JohnsHopkinsUniversityPress.
HealthServicesin the Third World.Totowa,N.J.: Rowanand Birdsall,Nancy,and P Chuan. 1986. "Client Choice of Health
Allenhead. Treatmentin RuralMali." Population,Nutrition, and Health
- 1986. "Demandfor PrimaryHealth Care in the Bicol Departmentworkingpaper.WorldBank, Washington,D.C.
RegionofthePhilippines."EconomicDevelopmentandCultural Birdsall,Nancy,F. Orivel,M. Ainsworth,and P Chuhau. 1983.
Change34(July):755-82. "Three Studieson Cost Recoveryin SocialSectorProjects."
Akin,JohnS.,D. Guilkey,andH.Douton.1995."Qualityof Services CountryPolicyDepartmentDiscussionPaper 1983-8.World
and Demandfor HealthCareinNigeria:AMultinominal Probit Bank,Washington,D.C.
Estimation."SocialScienceandMedicine40(11): 1527-37. Blakney,R., J. Litvack,and J. Quick. 1989. "FinancingPrimary
Alderman,Harold. 1996."HouseholdResponsesto PublicHealth Health Care: Experiencesin PharmaceuticalCost Recovery."
Services:Costand QualityTradeoffs."TheWorldBankResearch ManagementSciencesfor Health, Arlington,VA.
Observer11(1):3-22. Booth, D., and others. 1995. "Coping with Cost Recovery."
Alderman,Harold, and Paul Gertler.1989."The Substitutability Stockholm University, Department of Anthropology,
of PublicandPrivateHealth Carefor the TreatmentChildren's DevelopmentStudiesUnit, Stockholm.
in Pakistan."LivingStandardsMeasurementStudyWorking Carrin,Guy.1988."CommunityFinancingof Health Care."World
Paper57. WorldBank, Washington,D.C. HealthForum9(4): 601-06.
- 1997. "FamilyResourcesand Gender Differencesin Chaulagai,ChetNath. 1995."CommunityFinancingfor Essential
Human Capital Investments: The Demand for Children's Drugsin Nepal."WorldHealthForum16(1):92-94.
MedicalCareinPakistan."In HaroldAldermanand Lawrence Ching,Panfila.1995."UserFees,Demandfor Children'sHealth
Haddad, eds.,How FamiliesMakeDecisions.Baltimore,Md.: Care and AccessacrossIncomeGroups in the Philippines."
Johns HopkinsUniversityPress. SocialScienceandMedicine41(1):37-46.
Alderman,Harold,andVictorLavy.1996."HouseholdResponses Chisvo,M., and L. Munro.1994.A Reviewof SocialDimensions
to Public Health Services:Cost to QualityTradeoffs."The of Adjustment in Zimbabwe, 1990-94." United Nations
WorldBankResearchObserver11(1): 3-22. Children'sFund,Harare.
Arrow,Kenneth.1963."Uncertaintyand the WelfareEconomics Collins,C., and A. Green. 1994. "Decentralizationof Primary
of MedicalCare." AmericanEconomicReview53: 537-38. Health Care: Some Negative Implicationsin Developing
Baker,J., and L. van der Gaag. 1993. "Equityin Health Care Countries." InternationalJournalof Health Services24(3):
Financing:EvidencefromFiveDevelopingCountries."In Eddy 459-75.

STRATEGIES
FORPRICINGPUBLICLY HEALTHSERVICES
PROVIDED
151
Cornia, A., R. Jolly, and F. Stewart, eds. 1992. Adjustment with a
Human Face.Oxford: Oxford University Press.
Creese, Andrew. 1991. "User Charges for Health Care: A Review
of Recent Experience." Health Policyand Planning6(4): 603-19.
Creese, Andrew, and Joseph Kutzin. 1995. "Lessons From Cost-
Recovery in Health." Discussion Paper 2. World Health
Organization, Division of Strengthening Health Services,
Geneva.
Criel, Bart, and Harrie Van Balen. 1993. "Paying for the Kasongo
Hospital in Zaire: A Conceptual Framework." Health Policy
and Planning 8(1): 61-71.
de Bethune, X., S. Alfani, and J. Lahaye. 1989. "Influence of an
Abrupt Price Increase on Health Service Utilization: Evidence
from Zaire." Health Policyand Planning4(1): 76-81.
de Ferranti,David. 1985. "Payingfor Health Servicesin Developing
Countries: An Overview." Staff Working Paper 721. World
Bank, Washington, D.C.
Diskett, P, and P.Nickson. 1991. "Financing Primary Health Care:
An NGO Perspective." Development in Practice:An Oxfam
Journal 1(1): 43-51.
Dor, Avi, Paul Gertler, and Jacques van der Gaag. 1987. "Non-
Price Rationing and the Demand for Medical Care in Rural
C6te d'Ivoire." Journal of Health Economics 6(December):
291-304.
Dow,Wllliam H. 1996. "Unconditional Demand for Health Care
in Cote d'Ivoire: Does Selection on Health Status Matter?"
Living Standards Measurement Study Working Paper 127.
World Bank, Washington, D.C.
Dow,William,and Paul Gertler. 1997. "Insurance and the Demand
for Public and Private Medical Care in Indonesia." RAND,
Santa Monica, CA.
Dow,William,Paul Gertler, R. Schoeni,J. Strauss, and D. Thomas.
1997. "Experimental Results on the Impact of Medical Care
Prices on Health and Labor Supply" RAND, SantaMonica, CA.
Ellis, R. 1987. "The Revenue Generating Potential of User Fees
in Kenyan Government Health Facilities." Social Scienceand
Medicine 25(9): 995-1002.
Ensor, Tim, and P San. 1995. "Health Charges and Exemptions
in North Viet Nam." United Nations Children's Fund, Bamako
Initiative Research Program, New York.
Fiedler,John. 1993. "IncreasingRelianceon UserFees asa Response
to PublicHealth Financing Crisis:A Case Studyof El Salvador."
Social Scienceand Medicine 36(6): 735-47.
Frankenberg, Elizabeth. 1995. "The Effects of Access to Medical
Care on Infant Mortalityin Indonesia." Health TransitionReview
5(2): 143-62.
Fryatt, R., R. Rai, S. Crowley,andY Gurung. 1994. "Community
Financing of Drug Supplies in Rural Nepal: Evaluating a Fee
Per Item Drug Scheme." Health Policy and Planning 9(2):
193-203.
Gertler, Paul. 1996. "On the Road to Social Insurance: Lessons
from and for Asian Economies." Paper presented at the
INNOVATIONS IN HEALTI-I CARE FINANCING
152
Conference on Financing Human Resource Development,
May,Asian Development Bank, Manila.
Gertler, Paul, andJennieLitvack. Forthcoming. "Accessto Medical
Care during Transition: The Role of the Private Sector in Viet
Nam." Forthcoming in David Dollar, Jennie Litvack, and Paul
Glewwe, eds., Povertyand the Transitionto the Market in Viet
Nam. Oxford: Oxford University Press.
Gertler, Paul, and John Molyneaux. 1993. "How Economic
Development and Family Planning Programs Combined to
Reduce Indonesian Fertility."Labor and Population Program
Working Paper 93-08. RAND, Santa Monica, CA.
- . 1997. "Experimental Evidence on the Effect of Raising
User Fees for Publicly Delivered Health Care Services:
Utilization,Health Outcomes, and Private Provider Response."
RAND, Santa Monica, CA.
Gertler, Paul, and Orville Solon. 1996. "Who Benefits from Social
Health Insurance in Low Income Countries?" RAND, Santa
Monica, CA.
Gertler, Paul, and Roland Sturm. 1997. "Private Health Insurance
and Public Expenditures." Journal of Econometrics 77(1):
237-57.
Gertler, Paul, and Jacques van der Gaag. 1990. The Willingnessto
Payfor Medical Carein Developing Countries.Baltirnore, MD:
Johns Hopkins University Press.
Gertler, Paul, Luis Locay, and Warren Sanderson. 1987. "Are
User Fees Regressive?"Journal of Econometrics36(S): 67-88.
Gertler,Paul, B.YYang,S.W Lee, andJ. Peabody. 1995. "Financing
Health Care in as Aging Society:A Crisis on Korea's Horizon."
RAND, Santa Monica, CA.
Gilson, Lucy.1988. "Government Health Care Charges: Is Equity
Being Abandoned?" EPC Publication 15. London School of
Hygiene and TropicalMedicine,Evaluation and PlanningCenter
for Health Care, London.
Gilson, Lucy, and Anne Mills. 1995. "Health Sector Reform in
Sub-Saharan Africa: Lessons of the Last Ten Years."In Peter
Berman, ed., Health Sector Reform in Developing Countries:
Making Health DevelopmentSustainable.Cambridge: Harvard
University Press.
Griffn, Charles. 1987. "User Charges for Health Care in Principle
and Practice." Economic Development Institute Seminar Paper
37. World Bank, Washington, D.C.
. 1992. Health Carein Asia: A ComparativeStudy of Cost
and Financing. A Regional and Sectoral Studies book.
Washington, D.C.: World Bank.
Haddad, Slirn, and Pierre Fournier. 1995. "Quality, Cost and
Utilization of Health Services in Developing Countries: A
Longitudinal Studyin Zaire." SocialScienceandMedicine40(6):
743-53.
Hammer, Jeffrey. 1997. "Economic Analysisfor Health Projects."
The World Bank ResearchObserver 12(1): 47-72.
. Forthcoming. "Prices and Protocols in Public Health
Care." The World Bank Economic Review.
Hammer, Jeffrey, and Peter Berman. 1996. "Ends and Means in
Public Health Policy in Developing Countries." In Peter
Berman, ed., Health Sector Reform in Developing Countries:
Making Health Development Sustainable. Cambridge, Mass.:
Harvard University Press.
Hotchkiss, D. 1993. "The Role of Quality in the Demand for
Health Care in Cebu, the Philippines." PhD thesis. University
of North Carolina, Department of Economics, Chapel Hill.
Hongoro, Charles, and S.K. Chandiwana. 1994. "The Effects of
Enforcement of User Fees on the Health Care Delivery System
in Zimbabwe." Ministry of Health and Child Welfare, Harare.
Huber, J. 1993. "Ensuring Access to Health Care with the
Introduction of User Fees: AKenyan Example." Social Science
and Medicine 36(4): 485-94.
Jarrett, Stephen, and Samuel Ofosu-Amaah. 1992. "Strengthening
Health Services for MCH in Africa: The First Four Years of
the Bamako Initiative."Health Policyand Planning7(2): 164-76.
Jimenez, Emmanuel. 1987. Pricing Policy in the Social Sectors:
CostRecoveryfor Educationand Health in DevelopingCountries.
Baltimore, MD.: Johns Hopkins UniversityPress for the World
Bank.
11990. "Social Sector Pricing Policy RevisitedA Survey of
Recent Controversies." In Stanley Fischer and Dennis de Tray,
eds., Proceedings of the World Bank Annual Conference on
DevelopmentEconomics 1989.Washington, D.C.: World Bank.
1996. "Human and Physical Infrastructure." In Jere
1.
Behrman andT.N. Srinavasan, eds.,HandbookofDevelopment
Economics.Amsterdam: North Holland Press.
Kahenya, G., and S. Lake. 1994. "User Fees and their Impact on
Utilization of KeyHealth Services."United Nations Children's
Fund, Lusaka, Zambia.
lCani,N. 1989. "Charging for Drugs inAfrica: UNICEF's Bamako
Initiative." Health Policyand Planning4(2): 110-20.
Korte, R., H. Richter;F. Merkle, and H. Gorgen. 1992. "Financing
Health Services in Sub-Saharan Africa: Options for Decision
Makers duringAdjustment." SocialScienceand Medicine34(1):
1-9.
Lake, S. 1994. "User Charges in the Health Sector: Some
Observations on the Zambian Experience" United Nations
Children's Fund, Lusaka, Zambia.
Lavy,Victor, and Jean-Marc Germain. 1995. "Trade-offs in Cost,
QualityandAccessibilityin Utilizationof Health Facilities:Insights
fromnGhana." In R. Paul Shaw and Martha Ainsworth, eds.,
FinancingHealth Servicesthrough UserFeesand Insurance:Case
StudiesFromSub-SaharanAfrica. Washington,D.C.: World Bank.
Lavy,Victor, and John Quigley. 1993. "The Willingnessto Pay for
the Quality and Intensity of Medical Care." Living Standards
Measurement Study Working Paper 94. World Bank,
Washington, D.C.
Lavy,Victor,J. Strauss, D. Thomas, and P de Vreyer.1996. "Quality
of Health Care, Survival and Health Outcomes in Ghana."
Journalof Health Economics 15(3): 333-57.
STRATEGIES FOR PRICING PUBLICLY PROVIDED
153
Lewis,Maureen. 1993."User Fees in Public Hospitals: Comparison
of Three Country Case Studies." Economic Development and
Cultural Change41(3): 513-32.
Light, Donald. 1992. "Equity and Efficiency in Health Care."
Social Science and Medicine 35(4): 465-69.
Litvack,Jennie, and Claude Bodart. 1993. "UserFeesPlus Quality
Equals Improved Access to Health Care: Results From a Field
Experiment in Cameroon." SocialScience and Medicine 37(3):
369-83.
Makinen, M., and S. Block. 1986. "Pricing for Cost Recovery in
Primary Health Care in Guinea." REACH, Arlington, VA.
Manning,Willard,J. Newhouse, N. Duan, B. Keeler,A. Leibowitz,
and M. Marquis. 1987. "Health Insurance and the Demand
for Medical Care."American EconomicReview 77(3): 251-77.
Mbugua,J.Karanja,GeraldBloom,andMalcolmSegall.1995.'impact
of User Charges on VulnerableGroups: The Case of Kibwezi in
RuralKenya."Social Scienceand Medicine41(6): 829-35.
McPake, Barbara. 1993. "User Charges for Health Services in
Developing Countries: A Review of the Economic Literature."
Social Scienceand Medicine 36(11): 1397-1405.
McPake, Barbara, Kara Hanson, and Anne Mills. 1993.
"Community Financingof Health Care in Africa:An Evaluation
of the Bamako Initiative." Social Science and Medicine 36(1):
1383-95.
Murray, Christopher J. L., Gonghuan Yang, and Xinjian Qiao.
1992.'Adult Mortality: Levels,Patterns and Causes." In Richard
G. A. Feachem, Toud Kjellstrom, Christopher J. L. Murray,
Mead Over, and Margaret Phillips, eds., The Health of Adults
in the DevelopingWorld. New York, Oxford University Press.
Mwabu, Germano. 1989."Non-monetaryFactors inthe Household
Choice of Medical Facilities." Economic Development and
Cultural Change37(2): 383-92.
. 1990. "FinancingHealth ServicesinAfrica:AnAssessment
of Alternative Approaches." Population, Health, and Nutrition
Working Paper 457. World Bank, Washington, D.C.
Mwabu, Germano, andJosephWang'ombe. 1995. "User.Charges
in Kenya." World Bank, International Health Policy Program,
Washington, D.C.
Mwabu, Germano, Martha Ainsworth, and Andrew Nyamete.
1993. "Quality of Medical Care and Choice of Medical
Treatment in Kenya:An Empirical Analysis."Journal Of Human
Resources28(4): 838-62.
Mwabu, Germano, James Mwanzia, and Wilson Liambila. 1995.
User Charges in Government Health Facilitiesin Kenya:Effect
on Attendance and Revenue." Health Policyand Planning10(2):
164-70.
Newhouse, Joseph P., and the Rand Corporation Insurance
Experiment Group. 1995. FreeforAll? Lessonsfrom the Rand
Health Insurance Experiment. Cambridge, Mass.: Harvard
University Press.
Nolan, Brian, and Vincent Turbat. 1995. "Cost Recovery in Public
Health Services in Sub-SaharanAfrica." EDITechnical Series.
HEALTIH SERVICES
 WorldBank, EconomicDevelopmentInstitute, Washington, Solon,Orville,and others. 1991."HealthSectorFinancingin the
D.C. Philippines." ResearchTriangleInstituteand Universityof the
Olave,M., D. Gimenez,R.Weerhoffand others.1992."Studyof PhilippinesSchoolof Economics.Manila.
the RecurrentUnit Costs of PrimaryHealth CareServicesin Stanton,Bonita,and John Clemens.1989."UserFees for Health
a Sampleof BolivianMinistryof Health Facilities,1990-91." Care in DevelopingCountries:A Case Studyof Bangladesh."
U.S.Agencyfor InternationalDevelopment,Bolivia. SocialScienceandMedicine29(1): 1199-1205.
Parker,David,and RudolfKippenberg.1991."CommunityCost- Stinson,Wayne.1982. CommunityFinancingof PrimaryHealth
Sharingand Participation:A Reviewof the Issues."Bamako Care.Washington,D.C.: AmericanPublic HealthAssociation.
InitiativeTechnicalReportSeries9. UnitedNationsChildren's Thomas D., Victor Lavy,and J. Strauss. Forthcoming."Public
Fund, NewYork. Policy and AnthropometricOutcomes." Journalof Public
Peabody,John, PaulGertler,and A.Liebowitz.1995."The Effect Economics.
of Qualityof Prenatal Care on Birthweight."RAND,Santa TropicalDiseaseResearchProgram.1995.TDRNews (February).
Monica,CA. vander Gaag,Jacques.1995.PrivateandPublicInitiatives: Working
Pitt, Mark, M. Rosenzweig, and D. Gibbons. 1993. "The TogetberforHealthandEducation. ADirectionsinDevelopment
Determinants and Consequences of the Placement of book. Washington,D.C.: WorldBank.
GovernmentProgramsinIndonesia."TheWorldBankEconomic Vogel,RonaldJ. 1988. CostRecoveryin the HealthCareSector:
Review7(3): 319-48. SelectedCountryStudiesin WestAfrica.WorldBankTechnical
Richardson,P, J. Galloway,and M. Olave. 1992."Quality,Costs Paper 82.Washington,D.C.
and Cost-Recovery:A Comparative Study of the Unidad .1991."CostRecoveryinthe HealthSectorin Sub-Saharan
Sanitariaofthe Ministryof HealthandNutritionSustainability." Africa."InternationalJournal
ofHealthPlanning andManagement
URC,Washington,D.C. 6:167-91.
Rosenzweig,Mark, and KennethWolpin.1986. "Evaluatingthe Waddington,Cj., and K.A.Enyimayew.1990. 'A Price to Pay:
Effectsof OptimallyDistributedPublicPrograms:ChildHealth The Impact of User Chargesin the Voltaregion of Ghana."
andFamnilyPlanningInterventions."AmericanEconomicReviewv International Journalof HeathPlanningandManagement 5(4):
76: 470-82. 287-312.
Rothschild,Michael,and Joseph Stiglitz.1976. "Equilibriumin Weaver,Marcia,HollyWong,AmadouSeseSako,Robert Simon,
CompetitiveInsuranceMarkets:An Essayon the Economics and FelixLee. 1994. "Prospectsfor Reformof HospitalFees
of ImperfectCompetition." QuarterlyJournalof Economics in Sub-SaharanAfrica:A Case Study of Niamey National
80(November):629-49. HospitalinNiger."SocialScienceandMedicine38(4):565-74.
Sauerborn, Rainer,Adrien Nougtara, and Eric Latimer. 1994. WorldBank.1987.FinancingHealthServicesin Developing Countries:
"Elasticity of Demand for Health Care in Burkina Faso: AnAgendaforReform. AWorldBankPolicyStudy.Washington,
DifferencesacrossAgeand IncomeGroups."HealthPolicyand D.C.
Planning9(2): 185-92. . 1993.WorldDevelopmentReport 1993:InvestinginHealth.
Schwartz,J.Brad,JohnS.Alin, and BarryM.Popkin.1988."Price NewYork:OxfordUniversityPress.
and Income Elasticitiesof the Demand for Modern Health . 1995a."Indonesia:Public Expenditure,Prices,and the
Care: The Case of Infant Deliveryin the Philippines."The Poor."Washington,D.C.
WorldBankEconomicReview2(1): 49-76. .1995b."VietNamPovertyAssessment."Washington, D.C.
Shaw,R. Paul,and CharlesC. Griffin.1995.FinancingHealthCare . 1996."China:Issuesand Options in Health Financing."
in Sub-SaharanAfrica through User Feesand Insurance.A Washington,D.C.
Directionsin Developmentbook. Washington,D.C.: World Yoder,Richard. 1989. "ArePeople Willingand Able to Pay for
Bank. Health Services?"SocialScienceandMedicine29(1): 35-42.
Sheng-Lan,T, and others. 1994. "FinancingHealth Servicesin Zeckhauser,Richard.1970."MedicalInsurance:A Case Studyof
China:Adaptingto EconomicReform."ResearchReport26. the Tradeoff between Risk Spreading and Appropriate
Instituteof DevelopmentStudies,Sussex. Incentives."JournalofEconomicTheory77(1): 10-24.

INNOVATIONS
IN HEALTHCAREFINANCING
154
Cost Recovery Strategies
in Sub-Saharan Africa
JosephWang'ombe

C ost recoveryis an essentialcomponentof health * Increaseefficiencyin the provisionof healthcarebymak-

care reform-one that requires a major overhaul ingproviderscostconsciousand encouragingcost-effec-
of health care financingsystems.For the past ten tive provisionof care.
yearsmostcountriesin Sub-SaharanAfricahavebeenactively
reformingtheir healthsectors.Beforethat, these countries The General Model
had deterioratinghealthindicatorsand weak health infra-
structure, partly becauseof persistentshortfallsin health Manystudieshaveevaluatedthe effectsofcostrecoveryand
care financing. othermeasures(suchasdecentralization) on healthcarerev-
Most Sub-Saharancountries share a common agenda enue,quality,equity,andutilization(CreeseandKutzin1995;
forhealthreformthat includesincreasingeconomicgrowth, Shaw and Griffin 1995;Nolan and Turbat 1995; Berman
expandingbasic schooling(especiallyfor girls),reallocat- 1995;McPake1993).Yetmoststudiesare hesitantto assess
ing governmentspendingon health from tertiarycare and the effectsof costrecoveryon healthoutcomesfor tworea-
specialtytrainingto public health measuresand essential sons.First,healh outcomes(especially in the long run) are
carepackages,encouragingdiversityand competitionin the the resultof inputsfrommanyother sectors-includingpol-
provision of care and the developmentof cost contain- itics,agriculture,the environment,and nationalsecurity.
ment approachesto insurance,increasingthe efficiencyof Second,it is difficultto isolatethe effectsof healthout-
government health services,and fosteringthe participa- comeswithinthe healthsector.A countrythat makesgood
tion of communitiesand householdsin promotinghealth- policychoicesand adoptsappropriatestrategiesin the health
ier behaviorandin managinglocalhealthcareservices(World sector maynot generatebetter health outcomesif policies
Bank 1993). areweakin othersectors.Mwabu(1996)demonstrateshow
In most countriescostrecoveryeffortsare intendedto: countriesthat pursued structuraladjustmentpoliciesdur-
* Raiserevenuefor healthcare by imposinguser charges ing 1980-93recordedgainsin the healthstatusoftheirpop-
for publichealth servicesthat used to be providedfree ulations.The convergenceof cross-sectoralcompensatory
of charge. effectson healthis givenas the explanation.Adjustmentin
* Improvethe coverageand qualityof care by increasing agriculture,education, and other sectors,combinedwith
resourcesfor the health sector. adjustmentsin the health sector,increaselife expectancy
* Enhance equity in the provisionof health care by tar- andreduceinfantmortality.Thispaper reviewsSub-Saharan
getingspendingtoward servicesfor the poor and other Africancountries'cost recovery,cost sharing,and user fee
vulnerablegroups. reformsin the healthsector in the context of generalstruc-
* Improve serviceutilizationpatterns and controlfrivo- turaladjustmentefforts,bearingin mindthe complexinter-
lous demand. actionsamongdifferentsectors as they affecthealth.

JosephWang'ombe
is associate
professorandchairmanof the Department
of Community
Healthat the University
ofNairobi.

155
 Most recent surveysevaluating Sub-Saharan Africa's per-
formance in cost recovery rely on data from 1992 and, in a
few cases, from 1993 (Shaw and Griffin 1995; Gilson and
Mills 1995;Creese and Kutzin 1995;Bennett and Ngalande-
Banda 1994;McPake 1993;Nolan andTurbat 1993).During
1993-95, however, many countries in the region imple-
mented important health sector reforms. These efforts
may reflect the release of the World Bank's World
DevelopmentReport 1993:Investingin Health, and the cam-
paign that ensued thereafter. Thus attempts should be made
to compile more recent survey data.
Recent Reform Efforts
ManyAfrican countries have adopted health sector reforms,
including:
* Charging user fees at public facilities to generate rev-
enue from out-of-pocket payments, socialinsurance, pri-
vate insurance, and community financing (for example,
the Bamako Initiative).
* Increasing funds for the health sector and raising the
level of spending for public health care for the poor and
other vulnerable groups.
* Adopting innovative financing arrangements in the orga-
nization and management of health care delivery sys-
tems. The most common strategy has been to decentralize
health care delivery systems and privatize public insti-
tutions.
* Providing incentives to expand nongovernmental deliv-
ery of health care (for example, tax incentives for pri-
vate health care providers and subsidies to religious
groups, programs, and institutions).
User Fees As a Source of Additional
Revenue
A recent World Bank review of user fee schemes in Sub-
SaharanAfricafoundthatin 1993onlyfivecountries-Angola,
Botswana, Malawi,Sao Tome and Principe, and Tanzania-
did not have any apparent form of user fees or cost recovery
in the public health care system (Shaw 1995).By 1994 these
countries (except Botswana)had started to implement some
form of cost recovery,and by now have acquired some expe-
riencewithcost recovery(Bennettand Ngalande-Banda1994).
INNOVATIONS IN HEALTH CARE FINANCING
156
Country progress with cost recovery can be classified in
four categories: a national systemof user fees, some national
system of fees but with minimal enforcement, some facili-
ties and communities collecting fees, and no apparent form
of user fees (table 1). The countries in the first category are
implementing clear policies on cost recovery and are run-
ning a unitary health care delivery system; that is, the cen-
ter controls or oversees regional efforts. Countries in the
second category also have a unitary system, but cost recov-
ery policies are less evolved, and institutions have not been
reoriented to implement cost recovery policies. Countries
in the third category do not have a national system, but indi-
vidual communities and health facilities are implementing
cost recovery schemes. (In these countries the Bamako
Initiative is implemented separate from the general policy
of cost recovery.) Countries in the fourth category are still
formulating policies and establishing structures and insti-
tutions (at this point onlyBotswana occupies this category).
Cost recovery for public health care services was ini-
tially viewed as a panacea for inadequate revenue genera-
tion in the health sector. However, it has become apparent
that although they can increase revenue in absolute terms,
direct fees have little proportionate impact on the public
health system's financial requirements (Waddington and
Enyimayew 1989; Shaw and Griffin 1995). Most countries
in Sub-Saharan Africa recover 3-5 percent of ministry of
health recurrent expenditures through user fees (table 2).
The literature espousing user fees, however, claimsthat they
should recover as much as 20 percent of recurrent costs
(World Bank 1993).
Such levels-and some much higher-have been
achieved, but only in small-scale projects and community-
based schemes similar to the Bamako Initiative. In Ghana,
for example, user fees under some programs have recov-
ered 52 percent of costs, in Guinea-Bissau 32 percent, in
Mali 55 percent, in Senegal 50 percent, and in Uganda 19
percent.
These high levels of cost recovery are partly explained
by the special circumstances of these initiatives. Such pro-
jects have access to essential (often foreign) management
skillsand can finance necessaryinfrastructure.Inthe Bamako
Initiative projects, external funds from the United Nations
Children's Fund (UNICEF) are in some places used to
buy the initial stocks of drugs that establish the base for
TABLEI
Costrecoveryin publichealthfacilitiesin Sub-Saharan
Africa,1996
CategoryI Category2 Category3
Costrecoveryin placeand Somenationalsystemof user No nationalsystemof user Category4
dominatedby nationalsystem fee but minimalor poor feesbut somefacilities
and No apparentformof userfees
of userfees enforcement communities collectfees or costrecoveryin place
Anglophone,usophone
countnies
Angola Equatorial
Guinea Botswana
The Gambia Ethiopia
Ghana Guinea-Bissau
Kenya Malawi
Lesotho Nigeria
Malawi SaoTom6and Principe
Mozambique SierraLeone
Namibia Sudan
Swaziland Tanzania
Francophone countries
Benin BurkinaFaso CentralAfricanRepublic
Burundi Mauritania Congo
Cameroon Togo Madagascar
C6te d'lvoire Rwanda Niger
Guinea Zaire
Mali
Senegal
Source:Derivedfrom Shaw 1995.

the revolving drug fund. Moreover, the projects are cush-
ioned against inadequacies in the general health care system.
The potential for revenue generation from direct user
fees has not been realized. Shortfalls may be explained by
inefficiencies, implementation costs, lack of skills, an inabil-
ity to pay for full service cost, and incomplete institutional
adjustment and realignment (McPake 1993).Yet even if
fullyrealized, the potential of user fees would not result in
full cost recovery. Hence work continues on alternative
sources of finance, includingsocial insurance, private insur-
ance, and community-based schemes (Shaw and Griffin
1995).
User Fees and Equity
Discussions of user fees and equity in Sub-Saharan Africa
generally focus on the utilization, accessibility, and avail-
ability of services, considerations for indigent and low-
income groups, initiatives to deal with health problems
specific to the region (for example, childhood diseases such
as diarrhea, measles, and whooping cough), and support for
public health care approaches in general. Proponents of user
fees claim that the increased revenue and price effects on
demand that such fees generate can improve equity by
increasingthe availabilityof servicesand encouraging appro-
priate utilization (World Bank 1987; de Ferranti 1985;
Mwabu and Mwangi 1986). The increased revenue should
be reinvested and targeted toward servicesthat improve the
health of the poor. To ensure that vulnerable groups are
not excluded from public services, exemption schemes
should be developed. Moreover, these groups should be
entitled to public subsidies.
Although recent studies on the effect of user fees on uti-
lization rates have found negative consequences for equity,
this conclusion is somewhat controversial given earlier stud-
ies by Heller (1982) and Akin and others (1985), which
found that demand was inelastic with respect to price and
income. Yet later work, using different econometric for-
mulations from Heller and Akin and others, found that
users are sensitive to price changes (Gertler and van der
Gaag 1990; Waddington and Enyimayew 1990; Mwabu
and Wang'ombe 1995;Bennett and Ngalande-Banda 1994).
Even in countries where price elasticities are low, large
drops in utilization have been demonstrated-in Kenya
by up to 38 percent (Mwabu and Wang'ombe 1995). Huge
drops in utilization in countries where price elasticities are

COST RECOVERY STRATEGIES IN SUB-SAHARAN AFRICA

157
TABLE
2 Meanstestingis onemethod that is used to identifythe
Ministryof healthrecurrentexpenditures
recov- poor (see Levine and others 1992 forexamplesof means
eredthroughuserfees,variouscountries testingin Ethiopia,Ghana, and Senegal).Most countries,
oftotal)
(percentage however,lacksufficientskillsto use meanstestingto deter-
Country Share Year mine the extent of poverty and implement appropriate
Botswana i1.3-2.8 1983 exemptionpolicies.As a resultfee exemptionprograms
are often exploitedbythe nonpoor (civilservants,medical
C8ted'lvoire 3.1-7.0 1986
Ghana 7.9 1986 workers,the military,and so on). In Kenya,for example,
7.8 1992 largeamountsof potentialrevenuewere lostwheenexemp-
Guinea-Bissau 0.5 1988 tions were granted to civil servantsand healthworkers
Kenya 2.1 1993 (Collins and others 1995). This practice has since been
replaced with a medical allowanceto civil servants with
Lesotho 5.8 1986/87
9.0 1991/92 whichit is hoped they willpayfor medicalcare. Civilser-
Mali 1.2-7.0 1986 vants in Ghana, Mali, Niger, St. Lucia, and Yemenalso
Mozambique 8.0 1985 receive non-income-related exemptions (Bennett and
Lessthan1.0 1992 Ngalande-Banda1994).
Senegal 4.4-7.0 1986 The increasedrevenuefrom user feeswas expected to
Swaziland 2.2 1985 help deliver servicesof public health importance (immu-
4.6 1988/89 nizations,communicablediseasecontrol) and to support
Memo
items publichealth care in general. As noted above, however,
China(excludeinsurancereimbursements) 24.0 1980
36.0 1988 revenuepotentialhas not been realized.No Africancoun-
try has managedto use funds from cost sharingto finance
public utility health care services; public health care is still
Salvador 4.0 1990 financedby the budget. Retainingfees in the institutions
Yemen 3.3 1983 and districtswhere they are collectedmight help address
Source:
Creese
andKutzin1995.
this problem,so long as districtsare able to decideon the
lowcan be explainedby the fact that feeswere introduced servicesmixforfunding.Retainingfeeswherethey are col-
where they previouslydid not exist or were very low.In lectedis alsoanecessaryconditionof decentralization, which
some cases declines in utilization reversed slightlyafter is becomning popular. This approachis now operationalin
users adjusted to the initialpriceshocks(Collinsand oth- Uganda(Tmdyebwa 1997),Kenya(Collinsandothers1995),
ers 1995). and Mozambique, Nigeria, and Zambia (Bennett and
In most casestwo groupscut back on their use of pub- Ngalande-Banda1994).Other countries-Burundi,Congo,
lic serviceswhenuser fees are introduced:people who are Central AfricanRepublic,Ghana, Guinea, Malawi,Mali,
completelyexcludedfrommodernhealthcarebecausethey Niger,Togo,Zaire-split revenuesbetweenhealthfacilities
depend entirelyon governmentservices,and peoplewho and the ministryof health (Bennettand Ngalande-Banda
are picked up by other providers. Equity considerations 1994).
require that the firstgroup be protected,usuallywith fee In these countriesonlyhealth centersretain alltheir fee
exemptions.AllSub-SaharanAfricacountriesthat are pur- revenue;hospitalsremit some of their incometo the cen-
suingcostrecoveryhave sometypeof exemptionprogram. tral treasury.In Ethiopiaand Namibiaallthe moneyreverts
These programs attempt to identify people who cannot to the treasury.Retentionof feesin districtsandinstitutions
affordthe servicesthey need and to make surethat people could also increase equity in another area-namely, by
whocan affordthem are chargedfor the servicestheyuse. improvingthe qualityof care.Revenuesare expectedto be
In manycountriesthe taskof identifyingthe deservingpoor reinvestednot onlyto increase(ormaintain)the volumeof
isleft to communitiesor managersofhealthcareinstitutions. servicesbut alsoto improvetheir quality.Moreover,users

IN HEALTHCAREFINANCING
INNOVATIONS
158
willdemandqualityimprovementsiftheyhaveto continue lichealthcareservices(suchas familyplanning).Nationally,
paying.The slowrecovery of serviceutilization rates in user chargesare being implementedon a gradatedscale.
GhanaandKenyaafteruserfeeswereimposedcanbe partly Gate fees are higher at provincialhospitalsand lowerat
attributed to the fact that facilitiescouldnot retain feesto
healthcenters.There are no user chargesat villagedispen-
investin qualityirnprovements (Waddington and Enyimayew saries(Collinsandothers1995).Cbted'Ivoire,Ghana,Mali,
1990;Collinsand others 1995). Namibia, Zambia, and Zimbabwealsouse gradated user
Policieshave since been changed in Kenya,however. charges(BarnumandKutzin1993;Bennett and Ngalande-
Facilitiesnowretain 100percent of fees and are supposed Banda1994).Namibiais encouragingproperuse of arefer-
to use 75 percent for qualitymaintenanceand 25 percent ral systemby exempting referral clients from charges at
forpublichealthcare.Commnunity projectsandthe Bamako higher-levelinstitutions(Creeseand Kutzin1995).The rev-
Initiativetend to be successfulat increasingquality,partic-enue collectedat variouslevelsis used to improvequality
ularly in francophone countries such as Cameroon and and to encourageclientconformitywiththe referralsystem.
Guinea(LitvackandBodart1993;Nolanand Turbat1993). Mwabuandothers (1995)showthe importanceof avail-
The revolvingfund feature of these projectsallowsfunds abilityof drugs on demand of services.BamakoInitiative
to be used to maintaindrug supplies.Thesesuccesseshelp projectsare credited with maintainingqualityby making
legitimizedecentralizationpolicies. drugs available.If the selectionof drugs coverslocaldis-
eases,clientsonlyneedgoto higher-levelfacilitiesforrefer-
User Fees and Efficiency of Service ral services.Drugsmaybecheaperatthe communityfacility
Delivery than at the referralfacility.But evenif they are not, having
drugs availableat the communitylevel allowsclients to
It is often claimedthat user fees increaseefficiencyin the escapeother consumptioncosts(suchas transportandtravel
deliveryof healthcare, for severalreasons.First,user fees and waitingtime).
causeserviceprovidersandusersto behavemoreefficiently. Asnoted above,user feescan alsolowerutilizationrates.
Second, providershave an incentiveto allocaterevenues Presumably,part of that reductionis due to a reduction in
to produce appropriateservicesat the appropriatelevels, frivolousdemand. Although lower utilization rates may
and to choose appropriateproduction techniques.Finally, increaseefficiency,there is no evidenceshowingwhat por-
userfeessendpricesignalsto whichclientsrespondbyusing tion of the reduction is due to frivolousdemand.
onlythe servicesthat theyneed.Asa resultclientswillrespect
the referral systemand seek onlyprimarycare at low-level User Fee Implementation
institutions(healthcenters and dispensaries)and onlyter-
tiarycare at high-levelinstitutions(districtand provincial The contextin whichcost recoveryis implemented-that
hospitalsand other referralinstitutions;WorldBank 1987; is,the institutional,administrative,andmanagementframe-
de Ferranti 1985). Until cost sharing was adopted, man- work-is as importantfor the successof the policyas the
agersof public serviceswere not concernedwith allocative abilityof users to pay for servicesor the quality of those
and technical efficiency.It was common to see tertiary services.Sincehealthreformdependson successfulreforms
institutionsprovidingprimarycare. The pyramidreferral in other,related sectors,it must be implementedas part of
systemwas failing. a broader processof structuraladjustment.But successin
Evidenceon cost sharingin Sub-SaharanAfricareveals the health sector also depends on internal structuresand
a dynamicpicture of policyinitiativesaffectingpeople's capabilities.Successin cost recoveryrequiresa publicsec-
choices of levels and types of servicesand gradation of tor that iscapableof admrinisteringand managingthe process.
fees. In Kenya,for instance,the national referralhospital Mostexamplesof administrativeandmanagementprob-
(KenyattaNational Hospital) has in the past four years lems come from Ghana and Kenya, where districts and
stopped providingadultoutpatientcare exceptemergency regionshold funds for long periods in non-interest-earn-
casualtyservices.The hospitalalsono longerprovidespub- ing accounts and spend too little in the face of shortages

COST RECOVERY STRATEGIES IN SUB-SAIIARAN AFRICA

159
and financialneeds (Waddingtonand Enyimayew1990;
Collinsand others 1995).These problemscanbe tracedto
a lack of knowledgeand skillsin financialmanagementor
to bottlenecks in making decisions on expenditures.
Ministriesofhealthin anglophoneAfricatraditionallytrans-
ferred spendingauthorityfromheadquartersto regionsand
districts and imposed strict timeframes and amounts.
Althoughthissystemworkedwellwhenministriesprovided
free servicesand districtsspent moneygranted by head-
quarters, it is no longereffective.
Mostattemptsto respondto theseproblemshavefocused
on two policies.The first frees the provisionof health ser-
vices from the encumbrancesof the civilserviceby creat-
ing autonomousor semiautonomousboards. Zambia has
proposedcreatinghealthserviceboards,and Ghanaiscon-
sideringdevelopingGhana Health Services,whichwould
be similarto Ghana Education Services.Managementof
largehospitalsand medicalcenters-including Muhimbili
Hospitalin Tanzania,MulagoHospitalin Uganda,and the
RoyalVictoriaHospital in the Gambia-also maybe put
under independentboards (Cassels1995).
The second response decentralizes responsibilityfor
the managementor provisionof healthservicesto localgov-
ernmentsor health care agencies.However,this approach
creates many difficulties,mainlydue to lack of resources
and managerialcapacityat the locallevel.For example,in
Nigeriaand Tanzaniahealthbudgets are raided to finance
localadministration.In Zambiathe reformprogramenvis-
agedthat district and hospitalboards would set their own
feesandemploystaff.Thisapproach,however,couldincrease
the inequitythat decentralizationis meantto address.The
center needs to retain some control over the distribution
ofresources,butwithoutimposingtoomanyrulesandreg-
ulationson the civilservice(Cassels1995)
Conclusion
Thedebateonuserfeesforhealthcareserviceshasmoved
from advocacytoward issuesinvolvingthe programs'effi-
ciencyand effectiveness.So far the experiencewith user
pointing. Except in cases of verticalprograms and com-
munity-basedprojects,userfeeshavefailedto providemore
than5 percent of recurrentministryof healthcosts-much
INNOVATIONS IN HEALTH
160
less than the 20 percent anticipated.This potential is not
realizedbecauseof organizationaland managementprob-
lems. Moreover,implementationof cost recoveryefforts
must ensure that quality, equity, and efficiency are
maintained.
Different types of potentiallysustainableand effective
policies and programs for achievingquality,equity, and
efficiencyare being tried in Sub-SaharanAfrica. These
includefeeexemptionschemesthattargetvulnerablegroups,
fee retention at the point of collectionto reinvestin qual-
ity,and reorganizationand decentralizationof health care
deliverysystemsto enhance efficiency
Theseeffortsfacean apparentcatch-22,however.Because
systemsare not raisingenoughrevenuefromuser fees,there
is not enough money to reinvestin quality improvement
and systemadjustments.But these problemsare not insur-
mountable,as evidenceby the BamakoInitiativeandother
community-basedprojects that have reported successat
local levels.Properlymanagedprojectshave been able to
use initialexternal fundingto lay a foundation for a sus-
tainablerevolvingfund. This model should be applied to
national-levelprogramswherepoliciesare clear.An exter-
nal injection of funds targeted at quality improvements
shouldenablehealth care deliverysystemsto attract higher
feesfrom clients.At the same time, however,governments
must demonstratethe politicalwillto pursue these policies
aswellasrestructurepublichealthservices.Structuraladjust-
ment in other sectorsis essentialas well.
References
Akin,J. S., C. C. Griffin,andD. K Guilkey.1985. TheDemand
for Primary HealthServices in the ThirdWorld.Totowa,NJ.:
RowanandAlenheld.
Arhin,Dyna.1992."HealthFinancingin Zambia:AnApproach
to Cost-Sharing."Mission report.U.S.AgencyforInternational
Development, Washington, D.C.
Barnum,Howard,and JosephKutzin.1993.PublicHospitals
in Developing Countries:Resource Use, Cost, Financing.
Baltimore,Md.:Johns HopkinsUniversityPress for the
WorldBank.
Bennett,Sara,andEllisNgalande-Banda.
1994."Public
andPrivate
Rolesin Health:A Review andAnalysisof Experience
in Sub-
SaharanAfrica."StrengtheningHealthServices
Paper6.World
HealthOrganization, Division HealthServices,
ofStrengthening
Geneva.
CARE FINANCING
Berman, Peter, ed. 1995. Health Sector Reform in Developing
Countries:Making Health DevelopmentSustainable.Cambridge,
Mass.: Harvard University Press.
Cassels, Andrew. 1995. "Health Sector Reform: Key Issues in Less
Developed Countries." Discussion Paper 1. World Health
Organization, Geneva.
Collins, D.H.,J. D. Quick, S. N. Musau, and D. L. Kraushaar. 1995.
"Health Financing Reformnin Kenya:The Falland Riseof Cost-
Sharing,1989-1994."ManagementSciencesFor Health. Nairobi.
Creese, Andrew, and Joseph Kutzin. 1995. "Lessons From Cost
Recovery in Health." Discussion Paper 2. World Health
Organization, Geneva.
de Ferranti,David. 1985. "PayingFor Health Servicesin Developing
Countries: An Overview." Staff Working Paper 721. World
Bank, Wshington, D.C.
Gertler, Paul, and Jacques van der Gaag. 1990. The Willingnessto
Payfor Medical Care.Baltimore, Md.: Johns Hopkins Press.
Gilson, Lucy, and Anne Mills. 1995. "Health Sector Reforms in
Sub-Saharan Africa: Lessons of the Last 10 Years." In Peter
Berman, ed., Health Sector Reform in Developing Countries:
Making Health Development Sustainable. Cambridge, Mass.:
Harvard University Press.
Griffin, Charles. 1989. "User Charges for Health Care in Principle
and Practice." EDI Seminar Paper 37. World Bank, Economic
Development Institute, Washington D.C.
HeJler,P S. 1982. 'A Model of theDemand forMedical and Health
Services in Peninsular Malaysia." Social Science and Medicine
16(3): 267-84.
Kenya Government. 1994. "Kenya's Health Policy Framework."
Nairobi.
-.1996. "Kenya'sHealth PolicyFramework:Implementation
and Action Plans." Nairobi.
Korte, R., and others. 1992. "Financing Health Services in Sub-
Saharan Africa: Options for Decision Makers during
Adjustment." Social Science and Medicine 34(1): 1-9.
Lecher, M., R. Morar, B. Mohlala, F. Sithole, E. Semenya, and D.
Mphatsoe. 1995. "Modelling the Effects of National Health
Insurance on a Health District." University of Cape Town,
Department of Community Health, South Africa.
Levine, R., and others. 1992. "Means Testing in Cost Recovery:
A Review of Experience." Technical Note 23. U.S. Agencyfor
InternationalDevelopment, Health Financingand Sustainability
Project, Washington, D.C.
Litvack,Jennie, and Claude Bodart. 1993. "User Fees plus Quality
Equals Improved Access to Health Care: Results of a Field
Experiment in Cameroon." Social Scienceand Medicine 37(3):
369-83.
McPake, Barbara. 1993. "User Charges For Health Services in
Developing Countries: A Review of the Economic Literature."
Social Scienceand Medicine 36(11): 1397-1405.
McPake, Barbara, Kara Hanson, and Anne Mills. 1993.
"Implementing the Bamako Initiative in Africa: A Review and
COST RECOVERY
STRATEGIES
IN SUB-SAHARAN
161
Five Case Studies." London School of Hygiene and Tropical
Medicine, Department of Public Health and Policy, Health
PolicyUnit, London.
Mwabu, Germano. 1996. "Health Effects of Market-Based
Reforms in Developing Countries.' World Institute for
Development Research Working Paper 120. Helsinki.
Mwabu, Germano, and W Mwangi. 1986. "Health Care Financing
in Kenya: A Simulation of the Welfare Effects of User Fees."
Social Science and Medicine 22(7): 763-67.
Mwabu, Germano, andJosephWang'ombe. 1995. "User Charges
in Kenya:Health ServicePricingReforms in Kenya, 1989-1993.
A Report on Work in Progress with Support from the
International Health PolicyProgram. " IHPP,Washington,D.C.
Mwabu, Germano, Martha Ainsworth, and Andrew Nyamete.
1995. "The Effects of Prices, Service Quality, and Availability
on the Demand For Medical Care. Insights From Kenya." In
R. Paul Shaw and Charles C. Griffin, eds., FinancingHealth
Care in Sub-SabaranAfrica through User Feesand Insurance.
Washington, D.C.: World Bank.
Nolan, B., and V Turbat. 1995. "Cost Recovery in Public Health
Services in Sub-Saharan Africa." World Bank, Economic
Development Institute, Washington, D.C.
Quick,Jonathan. 1995. "Impact of Health CenterFees and Update
on Hospital Utilization." Management Sciences For Health,
Boston.
Shaw, R. Paul. 1995. "User Fees in Sub-Saharan Africa: Aims,
Findings, Policy Implications." In R. Paul Shaw and Charles
C. Griffin, eds., FinancingHealth Care in Sub-SaharanAfrica
throughUserFeesand Insurance.Washington,D.C.: World Bank.
Shaw,R. Paul, and Charles C. Griffin. 1995.FinancingHealth Care
in Sub-Saharan Africa through User Fees and Insurance. A
Directions in Development book. Washington, D.C.: World
Bank.
Tindyebwa, Dennis. 1996. Personal commnunication.Ministry of
Health, Uganda.
Waddington, C., and K. Enyimayew. 1989. "A Price to Pay: The
Impact of User Charges in Ashanti-Akim District, Ghana."
InternationalJournalof Health Planningand Management4(1):
17-47.
. 1990. 'APrice to Pay, Part 2: The Impact of UserCharges
in the Volta Region of Ghana." InternationalJournal of Health
Planningand Management5(4): 287-312.
Waters, H. 1995. "Literature Review: Equity in Health Sector in
Developing Countries, with Lessons Learned for Sub-Saharan
Africa." Basics Trip Report. U.S. Agency for International
Development, Washington, D.C.
WorldBank. 1987.FinancingHealthServicesin DevelopingCountries:
An Agenda For Reform. Washington D.C.
. 1993. WorldDevelopmentReport 1993:Investingin Health.
New York: Oxford University Press.
. 1994. Better Health in Africa: Experience and Lessons
Learned. A Development in Practice book. Washington D.C.
AFRICA
Rural Risk-Sharing Strategies
Andrew Creeseand SaraBennett

R isk-sharing strategies in health care relate not to ill

health, but to its financial consequences. The costs
of health care, like those of other goods and ser-
vices, are commonly assigned to the individuals using the
services, and these costs may be reassigned or shared using
a variety of risk-sharing institutions and mechanisms. Risk
sharing can be defined as the reduction or "elimination of
the uncertain risk of loss for the individual or household by
combining a larger number of similarly exposed individu-
als or households who are included in a common fund that
makes good the loss caused to any one member" (ILO
1996a, p. 3-1). These concepts apply equally to rural and
urban populations. Where rural and urban areas may dif-
fer substantially is in the implementation conditions facili-
tating or impeding risk-sharing efforts. Industrialization,
urbanization, high and rising per capita incomes, and pop-
ulation density-typically urban characteristics-facilitate
the growth of insurance (Ensor 1997).
Writing twenty years ago, Lipton (1976) showed that
earnings and leisure-that is, welfare-are substantially
lower in rural than in urban areas. In advancing his theory
of "urban bias," Lipton also argued that these gaps are clear-
est in the field of medicine:
The townsman has nine times as good a prospect of
medical attention as the villager in India, eleven times
in Ghana, thirty-three times in Ethiopia. The poorer,
the larger in area, and the less densely populated a coun-
try is, the greater in general is this disparity... Most rural
physicianscluster in the provincial capitals ... the qual-
ity of rural specialistsis lower ... the types of health facil-
ities provided almost rule out adequate emphasis on
villages ... referrals of the rural sick to urban hospitals
do not happen on any substantial scale. (pp. 265-66)
Of potential significanceto risk-sharingpotential in rural
areas, Lipton also established that welfare disparities were
typically smaller in rural areas than in urban areas.
Despite recent rapid urban growth, rural areas still dom-
inate developing countries. In 1988, 74 percent of Asia's
population (including China and India) and 73 percent of
Sub-SaharanAfrica'swas rural; in the least developed coun-
tries as a group this figure was 80 percent, with 69 percent
of the rural population below the poverty line (Jazairy,
Alamgir,and Panuccio 1993). It will be well into the third
decade of the next century before more than half of Africa's
population is urban (UN 1993).
Rural residents work primarilyin agricultural activities-
much of it seasonal,self,or familyemployment. Cash income
is seasonal and subject to large annual fluctuations. In poor
countries many rural residents face cash liquidityconstraints
for much of the year.In both urban and especiallyrural areas
a large portion of the economically active population is
engaged in informal employment. The size of the formal
sector, and its rate of expansion or contraction, has been
argued to be an important background factor in the suc-
cess or demise of national health insurance schemes (WHO
1995; Preker and Feachem 1995). It remains to be seen

AndrewCreeseis healtheconomistin the DivisionofAnalysis,Research,andAssessmentat theWrld Health Organization.SaraBennett

is appliedresearchdirectorat the Partnershipsfor Health ReformProject in Washington,D.C., and lecturer in health economicsand
financingat the London Schoolof Hygieneand TropicalMedicine.The authors are gratefulto DavidDunlop,Woutervan Ginneken,
Joseph Kutzin,and RoelandMonaschfor comments,suggestions,and practicalassistancein preparingthis paper.

163
whetherdevelopingcountriescan developmechanismsto
spreadsmall-scaleruralrisk-sharingschemesto the major-
ity of the population.Althoughthe ruralsector in develop-
ing countries has special featuresthat affect risk sharing
for healthcare costs,these issuesare not fundamentallydif-
ferentfromthose facedby informalworkersin urbanareas.
Urbaninformalemploymentis ofincreasingimportance
in developingcountries,andin recentyearshas grownfaster
than formal employment.In Latin Americathe informal
sectornowaccountsfor most urbanemployment(table 1).
Similartrendsarereported forSouthAsiaand Sub-Saharan
Africa.Thus, althoughthis paper's focusis on risk-sharing
arrangementsfor the ruralinformalsector,it alsoanalyzes
someurban and urban-ruralschemes.
Informal employmentcreatessubstantialdifficultiesin
the developmentof healthinsurance:it is difficultto iden-
tifybeneficiaries,to assesstheirincomes,and to collectcon-
tributions. Mandating coverage,which offers substantial
advantagesin terms of the sizeof the risk pool and control
of adverseselection,is alsomuch harder forinformalthan
formalworkers.
The scope of the concept of risk sharingalsowarrants
consideration. Discussionsabout rural risk sharing-or
about health insurancemore generally-typicallyanalyze
particularschemes(asthispaper does).Inmost discussions
of the health insuranceprospectsof developingand tran-
sitioneconornies,the analysisquicklymovesfrom concep-
tualdefinitionsof risksharingtoanalysisorrecommendations
on design and performance characteristicsof particular
schemes.Seldomis insurance consideredin broad terms
as afunction of health care systems,and as an overallpol-
icyobjective.But the poolingof risksis a policyobjective,
TABLEI
Urbaninformalnonagricultural
employment
in LatinAmerica,1990and 1994
(percent)tieoalexrens
Country 1990 1994
Argentina 47.5 52.5
Boliva 56.9 61.3
Brazil 52.0 56.4
Chile 49.9 51.0
Colombia 59.1 61.6
Ecuador 51.6 54.2
Peru 51.8 56.0
Source:ILO1996b.
and allhealth care systemsare characterizedby somerisk-
sharing arrangements.Thus risk-sharingarrangementsin
ruralareascouldincludepubliclyfinancedhealthcare,com-
munity-basedhealth insurance, and even private health
insurance.
Focusingon individualschemes,or evenbroad typesof
insurance,impedesan overallassessmentofthe roleof insur-
ance and the performanceof all actors.In large part the
debateremainsstraight-jacketed becauseinsuranceis treated
whollyor predominantlyas a sourceof healthcare finance.
Saltman(1995), for example,distinguishesfinance, allo-
cation mechanisms,and production componentsin com-
paring health care systems.But health insurance is not,
strictlyspeaking,a sourceof finance,but rather a type of
allocativemechanism.
Experiencewith ruralor informalrisk sharingis of major
potentialrelevanceto countriesat differentlevelsof devel-
opment.Witha growingrecognitionof centralgovernments'
limitedabilityto financeandmanagehealthcare,newforms
of financeand, perhapsmoreimportant,new formsof orga-
nizationarebeingintroduced.In low-incomecountrieswith
poor growthprospectsand large informalsectors,a major
goalof healthreformis to findnewwaysto organizeacces-
siblecare of goodquality,usinga maximumof nongovern-
ment resourcesin a transparentmanner.
The workpresentedhereis preliminary.It reviewsrecent
evidenceon the organizationand performanceof health
insuranceschemesforthe informalsector.Despiteour argu-
ment that risksharingshouldbe viewedas a functionrather
than as a categoryof schemes,it has proved necessaryat
this point to confineour analysis.At a later date weintend
to developthe analysisto considera wider varietyof risk-
sharingmechanisms.
Thirty-sixrisk-sharingschemesfor the informalsector
were reviewed.These schemesare in no way representa-
with ruraland urban informalhealth
insurance.The schemesincludedhere wereselectedbased
on the availabilityof adequate documentation.Thus our
sampleis likelybiased toward successfulschemes.Failed
schemes,while equallyinstructive,are less likelyto have
been documented.
The schemes examined spread risks to different degrees
and in different ways. The level of risk sharing depends on
the risks that are included in an insurance benefits pack-

INNOVATIONS
IN HEALTIICAREFINANCING
164
ageand on howlargeand diversethe risk pool is.The high- Thusthe relativelysmallset of documentedrecent expe-
est level of risk pooling occursin tax-funded,mandatory riencesin this area shouldbe seen as a subset of this older,
nationalhealth servicesystemsthat are capableof provid- more diverseset of risk-sharingarrangements.Giventhat
ingaccessibleservicesto the entirepopulation.At the oppo- theyhave barelybeen analyzedin terms of their potential
site extreme is out-of-pocketpayment by individualsfor for expandinginto healthinsurancefunds,the lessonsthat
theircare.Anindividual'sriskof incurringthe financialcosts can be drawn from both sets of experiencesare limited.
of health care can alsobe distributedover time. The sim- But history shows that people have organized and man-
plestexampleoccurswhenpricesfor healthcare are set on aged cash-basedrisk-sharingmechanismsfor big expendi-
a "feeper episode"basis,whichallowsthe patientan unlim- ture ("catastrophic")events,oftenwithcomplexcontribution
ited number of visitsuntil a defined illnessepisodeis over. andbenefit arrangementschedules.Membershipis usually
Moreambitiousintertemporalrisksharingisofferedthrough individual(rather than household),voluntary,and tightly
nontransferablehealth care benefitsfor a defined period, limited.Kinshipand trust is usuallyimportant.The impor-
rather like a season ticket for health care. The most sys- tance of socialhomogeneityand group confidencesignal
tematic intertemporalrisk-redistributingmechanismsare the limitationsofsuchschemesasa basisfornationalschemes
offeredin schemessuch as Singapore'sMedisave,an indi- unlessthey are supported as a set of initiativesthat other
vidualearmarkedmedicalcare savingsaccountthat is avail- actors (governmentand nongoverrment)join in a coordi-
able over a lifetime.This program allowspeople to build natedmanner.Of the schemesreviewedbelow,the mutual
up credit for health care whenthey are well,to cushionor funds ofYaounde,Cameroon,andYoffe,Senegal,are clos-
coverthe increasingcostsof care in old age. est to suchtraditionalarrangements.
The German,Japanese, and Koreaninsurancesystems
Long Traditions of Risk Sharing originatedin smallschemesof employedpeoplein the same
craft, town, or industry.Coveragegrewto the wholepop-
The risk-sharingschemesanalyzedbelowsharecertainchar- ulationasthesecountriesmovedtowardfull-orhigh-employ-
acteristics.
Theyinvolve(withoneexception)voluntarymem- ment industrialeconomies.Employmentin agriculturefell
bership, prepaymentof contributionsinto an identifiable sharplyand its productivityand earningsrose, enablinga
fund,some (often loose)notion of entitlementto benefits, largerportionof agriculturalworkersto organizethemselves
and a defined set of serviceproviders.To varyingdegrees into insuranceschemesor to buy into industrialinsurance
theyhave attractedinternationalinterestand support.But funds.Government,at leastin Korea,subsidizedruralpar-
theyrepresentonly a fractionof risk-sharingexperiencein ticipation.In recent yearsa reversetrend has been taking
protectingruralpopulationsagainstthe costsof unexpected placein Central andEasternEurope,wherefallingemploy-
bereavement,disability,and illness.Numeroustraditional ment and shrinkingtax revenue havestymiedattempts to
structuresalso spreadfinancialrisks amonggroupsor are establishemployment-basednational health insurancein
linkedto nonspecificsavingsschemes.Someschemesallow countriessuchas Bulgaria,Kazakstan,the KyrgyzRepublic,
both nonspecificsavingsand indemnifymembers against the SlovakRepublic,and Russia.
the costsofunpredictableevents,suchas marriageor death.
The oldestdocumentedaccountsofvoluntaryprepayment Typology of (Mainly Rural) Risk-Sharing
associationsfor healthcare date from the fifthcenturyB.C. Schemes
in Greece,wheretrade and craftgroupsorganizedmutual
helpschemesbasedonregularpooledsavingsto protectmem- Therearemanywaysto typologize the variousschemesinvolv-
bersincaseof death,ilness,orincapacity(WSM1996).Today ingruralrisksharing.Theycanbedefinedaccordingtowhether
voluntaryprepaymentschemesexistinmanypartsoftheworld, they are voluntaryor mandatory,the kind of benefitspro-
sometimeswith healthcare as a specificclaimon resources. vided,the degree of interpersonalrisk sharing,or the cir-
(See Lukholo 1996 and Alain, Tchente, and Guillerme- cumstancesthat led to their creation,such as government
Dieumegard1991for other descriptionsof suchschemes.) collapse,extemalassistance,or localcommunityinitiative.

RURAL RISK-SHARING STRATEGIES

165
 The niain dimension for categorization used here is fund
ownership and management-that is, the location of con-
trol and decisionmaking regarding the use of resources
and collection of contributions in the scheme (table 2).
The schemesexamined were predominantlyinsurance rather
than personal prepayment schemes, all but one were vol-
untary, most were rural (though five had an urban focus),
and most received external financial or technical support.
Health facility schemes are generally initiated by hospi-
tal staff and cover catastrophic hospital care costs. Such
schemes have geographically defined beneficiary groups
based on the hospital catchment area, often including both
rural and urban communities. Examples include India's
Kasturbahospital scheme and Zaire's Masisihospital scheme.
Community schemes usuallyfocus on primary care, espe-
ciallydrugs, but also may include referral services and often
have a broad community development orientation.Examples
include Guinea-Bissau's Abota, Indonesia's Dana Sehat,
Taiwan's (China) Farmer's health insurance, and Vietnam's
Quang Nam Da Nang (QNDN).
Cooperativeschemes are often linked to local labor mar-
kets and based on individuals' place of work. Contributions
to the health fund may come from the sale of cooperatively
produced goods. Examples include China's former coop-
erative medical system, India's Mathew Milk Cooperative,
and Senegal's Mutuelle de Yoffe.Cooperative schemes are
the successor to the craft-related associations that devel-
oped in ancient Greece, medieval Europe, and seventeenth-
century Chile.
NGO schemes vary,reflecting the origin and purpose of
the sponsoring NGO. Indeed, NGOs may be the motivat-
ing force behind facility,community,or cooperative schemes.
Ghana's Nkoranza, the Philippines's Organization for
TABLE2
Rural risk-sharingschemesby owner and region
Owner Africa
Healthfacility
Community
Cooperative
Govemment
NGO
Total I1
Note:Insomecountriestherearea largenumberof similar
butslightly
different
thesimilarities
between
theseschemes, theyarecournted
onlyonce.
INNOVATIONS IN HEALTHI CARE FINANCING
Education Resources and Training (ORT, an international
NGO), West Bengal's Saheed Shibsankar Sabar Samity
(SSSS, alocalNGO), andZaire's Bwamandaare examples.
Fund ownership and management were used to catego-
rize the schemes for several reasons. First, fund ownership
often reveals a scheme's initial motivation and objectives,
which can differ substantially. For example, some facility-
based schemes are driven primarily by the need to raise
revenue. Others may combine revenue-raising goals with
efforts to increase service utilization levels. Community-
based schemes, on the other hand, tend to have popula-
tions' needs as their principal motivation, although concern
is often focused at one level or on one benefit, such as drugs.
Other schemes have a demonstration motivation, such as
China's Sichuan Rural Health Insurance experiment.
Second, typologizing schemes according to fund own-
ership often reveals a scheme's design details. Design is an
important factor in a scheme's performance. Several risk-
sharing schemes have serious design flaws from a sustain-
ability perspective (see below). Third, ownership may
determine a population's overalltrust and confidence in the
scheme and in the servicesprovided. Finally,ownership may
be an important factor in determining the way a scheme can
be complemented and supported by government, and can
affect approaches to health care financing and organization.
A second key dimension for classifyingschemes relates to
the distinction between schemes that focus on providing
coverage for high-cost,low-frequencyevents and those that
focus on low-cost, high-frequencyevents (designated type 1
and type 2 in table 3). There is no easyway to categorizetype
1 and type 2 schemes, but they represent ends of a spectrum
of different risk-sharingschemes.Schemessuch as Bwamanda
(Zaire),Chogoria (Kenya),and Nkoranza (Ghana) are clearly
Southeast
and LatinAmericaand
SouthAsia EastAsia the Caribbean Total
3 3 0 i 7
4 1 5 1 11
I 1 4 1 7
I 0 4 0 5
2 3 1 0 6
8 14 3 36
schemes,
at operating
underthesameumbrella
name(such
asDana
Sehatin Indonesia).
Because
of
166
type1 schemes. Type2 schemesincludeDanaSehat(Indonesia) clearlydepends on the context, demand for type 1 and
andUnitedMission(Nepal).Someschemesincorporateele- type 2 schemesappearsto be rooted in differentcontex-
mentsofboth types.Forexample,severalschemescoverboth tual factors.Many type 1 schemeshave arisen in circum-
high-cost,low-frequency eventsandlow-cost,high-frequency stanceswhereuser feesfor hospitalserviceswere high.In
ones,oftenwithoutsettingpremiumson an actuarialbasis. this context people were interested in buying insurance
The distinctionbetween type 1 and type 2 schemesis primarilyfor its risk-sharingbenefits.In contrast,demand
important because different conditions are required for for type2 schemesdoes not stemfroma desireforrisk shar-
these types to succeed. In particular,type 1 and type 2 ing.Insteadthese schemeshavefocusedon improvingqual-
schemesdiffer in terms of: ity of care, particularlyby expandingservicesto previously
* The degreeof socialcohesivenessand trust required for unservedcommunities.
their effectiveoperation Much of the debate about risk sharinghas alsofocused
* The levelof demand for insurance on the administrativefeasibilityof implementing such
* Administrativecomplexity. schemes.Again,there is a criticaldifferencebetween type
A cohesivecommunityis much less of an issuefor type 1 andtype2 schemes.Type1 schemesthat aim to covercer-
I schemesthan type 2 schemes.Moreover,beneficiariesin tain variablecosts requireactuarialestimatesof premiums
type 1 schemestend to be distributedover a wide area and and definedpaymentmechanisms(ratherthan simplysup-
to be relativelyheterogeneous.Thusit isunlikelythat strong plementingbudgets)and are more difficultto managethan
feelingsof solidaritycanbe establishedundertheseschemes. type2 schemes.Managementstructuresfor type 1 schemes
Some type 1 schemes appear (mistakenly)to emphasize reflectthis difference.They tend to be more complex,and
communitysolidarity.For example,the Nkoranzaevalua- the managementproblemsencounteredare more substan-
tion report states that "the concept of risk sharingin the tial. Thus capable managementmay be crucialfor type 1
communitymustbe wellexplainedforthe peopleto under- schemes,but is probablynot required fortype 2 schemes.
stand that if you insure and do not benefit directly,your A formalmodel for type 2 schemeshas been proposed
'neighbourwillbenefit fromyourcontribution"'(Somkang by Hsiao and Sen (1995), under the title of cooperative
and others 1994,p. 12). healthcare.Underthismodelstrongsocialbondsandmutual
Such an appeal to social solidaritymaybe effectiveat trust are viewedas the definingelements of the commu-
the villagelevelbut seemsunlikelyto be successfulat the nitythat undertakesthe financingand provisionof health
district level. On the whole type 1 schemes are less con- care.Basedon ruralChina andIndia, servicesare expected
cerned with altruismand communitydevelopment.This to includepreventivecare, immunization,familyplanning,
feature and the nature of the benefitsthey offermean that maternaland childcare,healtheducation,andbasic ambu-
theyare morelikelyto be more susceptibleto adverseselec- latorycurativecare. Communitiesare expectedto contain
tion and moral hazardthan type 2 schemes. about 1,000peopleand to be capableof payingfora health
The levelof effectivedemandfor insurancein develop- workeranda stockof genericessentialdrugs.Amalgamations
ing countrieshas been hotly debated. Although demand of these communitieswould establish a primary health
center for about 15,000people staffed by a doctor,mid-
TABLE3 wife, clinicalnurse, pharmacist, and assistant.A mix of
Twotypesof risk-sharing
schemes annual premiumsand copaymentsis envisaged;member-
TypeI schemes Type2 schemes shipwould idealiybe compulsory,and referralswould be
High-cost,
low-frequency
events Low-cost,
high-frequency
events coveredthroughpackagedfees or capitation.
Tendto behospital
owned Tend
to becommunity owned Althoughnearlyallthe schemesexaminedare ownedby
Tendto coverwholedistrict Tend
to bebasedatthevillage
level a singleparty,autonomyin fund managementvaried.Some
Useactuarialbasisor variablecosts Premiumset mainlyaccordingto
for calculating
premium ability
to pay program,such as the Thai healthcard, are centrallydriven
Committed to meeting
certain Committed onlyto raising
extra schemesand operate withintightgovernmentguidelines.
designatedcosts revenue
forservices In Indonesiaandthe Philippinesguidelinesexiston the use

RURAL RISK-SHARING STRATEGIES

167
of fundsbut arenotverytight.Someschemes,suchasZaire's
hospital-basedschemes,had almostcompleteautonomy.
There waslittlediscussion,in the documentationof the
schemesreviewed,on the accountabilityof fund manage-
ment.Where schemeswere communityowned,fund man-
agers were sometimes held accountable to the local
communitythroughcommunitymeetingsor simpleaccount-
ing procedures (suchas showingreceipts to the commu-
nity; Chabot, Boal, and Da Silva 1991; Mogedal 1984).
Ghana'sNkoranzahospitalschemehad an insuranceadvi-
soryboardthat includedmembersof the hospitalmanage-
mentteam and twenty-five communitymembers(Somkang
and others 1994). On the whole, however, systemsfor
accountabilityto beneficiarycommunitiesappearedweak.
Overview of Schemes
This sectionreviewsthe thirty-sixschemes,focusingon:
The contextin whichthe schemewas developed
* Membershipand coverage
* Financingmechanisms
* Administrationand fund management
* Providerpaymentmechanisms
* Arrangementsfor healthcare provision.
Thecontext
The risk-sharingschemescomparedbeloware drawn from
contexts that sometimesdiffer widely Zaire's Bwamanda
schemeand Guinea-Bissau's Abotaschemeweredeveloped
in responseto thenearcollapseofgovemment-funded health
care.Korea'sclassII ruralschemes,on the other hand,were
initiatedata timeofacceleratingeconomicgrowthandurban-
ization,as a wayto protect ruralresidentsunder an increas-
inglynational insurancesystem.China'sCooperativeMedical
Systemwasdevelopedascommunalagriculturalproduction
wasemphasizedasthe mainmechanismforeconomicdevel-
opment.Whenmacroeconomicpolicyshiftedto the "social-
ist market"and centralandprovincialgovernmentsubsidies
werecut and redirected,the systemcollapsed(Hsiao1995).
Thelocalcontextmaybe shapedby the overalleconomic
situation and the direction of policy.Other factors that
may encourage risk-sharingmechanismsinclude decen-
tralization policy,traditions of communityinitiative and
INNOVATIONS IN HEALTH CARE FINANCING
168
management,NGO activity(suchas religiousmissions)or
other formsof technicalassistance,and monetizationin the
localeconomy(suchascashor subsistencecrops).The exist-
inghealth care systemalsomaylead to risk-sharingmech-
anisms, as well as limit or define the kind of insurance
response.For example,the qualityand availabilityofhealth
services,particularlygovernment services,might define
the needforadditionalor better serviceproviders.Insurance
maybe seenas a mechanismforimprovingor extendingthe
provisionof health services.In addition,the cost of exist-
ing servicesmay enableor.inhibitthe developmentof risk
sharing.Bennett and Ngalande-Banda(1994)observe,in
the context of Sub-SaharanAfrica,that user feesfor gov-
errnent servicesmaybe a prerequisitefor wider popular
demandfor or recognitionofthe potentialroleof insurance.
The availability,price, and qualityof private provision
mayhelp determinewhichservicesare coveredin the ben-
efits package.In Koreaand Thailandthe scale of private
provisionmadeit clearthat ruralinsurance,likeurban insur-
ance,wouldhaveto coverservicesprovidedbyprivateprac-
titioners, as well as by governmentclinicsand hospitals.
Thusthe configuration,quality,and priceof existinghealth
serviceshelpshapehealthinsuranceschemes.The national
or localcontextis importantin understandingboth the pur-
pose andperformanceof risk-sharingschemes,andin iden-
tifyingthe barriers or opportunities to their replicability
nationwideor in other countries
In nearlyall the schemesexamined,people were accus-
tomed to payingfees priorto the introductionof riskshar-
ing, or did not have any real accessto health care (as in
Guinea-Bissauand Lalitpur district, Nepal). Moreover,
feeswere often so high that a largeportionof the popula-
tion could not affordthem. This was particularlythe case
for facility-basedschemes;facilitiesoften faced declining
use becauseof high fees and low revenues.Such arrange-
mentsledto considerationsof insurance.In Zaire,for exam-
ple, health zones were meant to be self-supporting;thus,
other than funds receivedfrom externaldonors,zoneshad
to recoverfull operatingcosts.In other African countries
(notablyGhana and Kenya)schemeswereinitiatedby mis-
sionhospitals.Althoughdetailedinformationon the finances
of these hospitalsis not available,in generalmissionhospi-
talsin Africahavebeenforcedto relyon fee incomeasgov-
emmentsubsidiesandexternalsupporthas dried up (Gilson
and others 1994). InJapan, Korea,and Taiwan(China) health
care is dominated by private providers charging high fees.
The existence of other types of insurance schemes in a
country did not appear to affect the uptake of rural risk
sharing,but many people were able to understand the notion
of risk sharing because they were used to traditional mutual
self-help mechanisms (Wong share in Thailand, Gotong
Royong in Indonesia, Abota in Guinea-Bissau).
Some of the schemes examined were run by NGOs that
appeared to be quite autonomous from government. In
other instances schemes were initiated in the face of a break-
down in government finance for health care (for example,
in Zaire, Guinea-Bissau, and recently in China). Although
some schemes (such as that in Guinea-Bissau) operated in
small, close-knit communities, others covered large districts
with diffuse communities. In Guinea-Bissau the cohesive-
ness and small size of local villages were seen as factors
supporting the success of the schemes (Chabot, Boal, and
Da Silva 1991). However, the importance of this factor
depends on the ownership of the scheme. Where funds are
community owned, trust within the community is critical.
Most schemes covering larger areas were facility or gov-
erminent owned, so accountability within management sys-
tems is more important than trust in the community.
Schemes were located both in areas where most bene-
ficiarieswere subsistence farmers (as in Nepal and Guinea-
Bissau) and in areas where a large number of farmers were
organized in cooperatives. Cooperative and mutual schemes
develop where labor is more organized.
Economic growth often supports the growth of risk shar-
ing. In East and South-East Asia rapid expansion of health
insurance in the nonformal sector coincided with rapid eco-
nomic growth. Even in Bwamanda, Zaire, one of the fac-
tors contributing to the success of the scheme was a buoyant
local economy, attributable to a donor-funded rural devel-
opment project that established the health insurance scheme
(Moens 1990).
Membershipand coverage
Membership. Potential beneficiaries of schemes are
defined both by geographic location (particularlycatchment
areas for hospital-based schemes and village, ward, or dis-
trict of residence for community schemes) and by employ-
RURAL RISK-SIIARING STRATEGIES
169
ment variables (whether occupation, place of work, or how
produce is sold). Only one of the schemes reviewed, India's
Kasturba Hospital scheme, was targeted at the poor.
Most schemes use the household as the unit of mem-
bership.Many schemesthat initiallyallowed individualenroll-
ment often faced problems of adverse selection and switched
to household enrollment. In Nkoranza, Ghana, premiums
were set individuallybut the entire household had to join.
The failure of insurance scheme workers to sign up all mem-
bers of a household, however, contributed to the failure of
the scheme (Somkang and others 1994). In Taiwan(China)
individual enrollment was allowed under both the Farmers
Health insurance scheme and labor insurance, since only
workers (not their dependents) could be enrolled. However,
the Farmers scheme appears to have been compulsory and
the labor insurance scheme probably attracted few self-
employed workers. In Vietnam at least two-thirds of the
household was required to join (Ensor 1995).
Other measures that prevent adverse selection include
requiring that a minimum number or portion of households
in a village or administrative area join a scheme. In the
Kasturba program at least 75 percent of poor households
in a village must join. In Thailand at least 30 percent of
households in a village must join in order for the village to
participate in the health card scheme.
If enrollment in a scheme is allowed over a long period
and there is no waiting period, then people tend to enroll
when they need care. At the VH-Shospital in Madras, India,
where enrollment was allowed throughout the year with-
out a waiting period, less than a quarter of subscribers had
renewed their membership. The remaining three-quarters
probably joined when they got sick, eroding the insurance
effect of the scheme (Dave and Berman 1990). Several
schemes that planned to have a limited enrollment period
later extended it because enrollment rates were low. Users
who enroll only when they get sick create big problems in
hospital-based schemes, where the need for service is most
unpredictable and bears large financial consequences. For
community schemes covering mainly primary care it is eas-
ier for households to predict utilization and decide whether
it is worth joining.
Coverage.Except for the schemes in China, coverage of
the target population tends to be low (table 4). Zaire's
TABLE4 There were exceptionsto this generalpattern.Twohos-
Coverage of schemes pital-basedschemesin India (KasturbaandVHS) and one
Numberof Coverage
of schemein Bangladeshsetpremiumson a slidingscaleaccord-
Scheme Country people covered targetpopulation ing to income. More sophisticated schemes in Japan and
Facility-owned Taiwan(China) generallyset premiumsas a percentageof
Chogoria Kenya 9,000 17% earnings. In Korea premiums were based on a complex
CIMIGen Mexico 600 n.a.
Katurba India 4,390 n.a. assessmentof incomeand assets.
Masisi Zaire 3,500 n.a. Some villagesin Guinea-Bissau, the scheme in Lalitpur,
Raigarh India 75,000 n.a. Nepal, and the Kasturba hospital scheme in India allowpay-
VHS India 3,800 n.a.
ment in kind. Interestingly,very few poor agricultural com-
Community-owned
Abota Guinea-Bissau200,000 n.a. munitiesin Nepal choseto pay in kind (Donaldson1982).
Assaba Guatemala 65,000 40 families The Dana Sehat scheme in Indonesia and the ORT
Boboye Niger 250,000 i00%(compulsory) scheme in the Phiippines were the only insurance schemes
DenaSehat Indonesia 12,000,000 13villages
Healh card Thailand 2,700,000 6.5% that allowed payment of premiums more than once a year.
Lalitpur Nepal 6,000 5-10% Still, monthly payments create difficulties for workers with
RuralCooperative China s46,000 31-100%
MedicalSystemproject seasonal income. Ile ORT scheme adopted a flexible pay-
Sichuanruralhealth China 40,443 >90% ment schedule because it was felt that many households
insurance would not be able to afford the annual premium in one lump
Govemment-owned sum. A number of families, however, dropped out of the
Muyinga Burundi n.a. 23 households
QNDN vietnam 10,000 n.a. scheme because they failed to keep up payments (Ron and
Cooperative-owned Kupferman 1996).
CooperativeMedical China About -100% Few schemes had built-in exemption policies. In Boboye,
Scheme 700,000,000
MallurMilk India 5,000 n.a. Niger, the indigentcould get specialwaivers (Diop,Yazbeck,
MutuelleFamille Cameroon 455 22% and Bitran 1995). In Guinea-Bissau's Abota scheme vil-
de
Babouantou lagers could choose to give the indigent access to drugs
Yaounde
Yoff Senegal 150families n.a. despite the fact that they had not paid (Chabot, Boal, and
NGO-owned Da Silva 1991). In the ORT scheme project staff tried to
Bwamanda Zaire 80,000 66% seek supplementary funds to subsidize the premiums of
G.Kendra Bangladesh 20,000 n.a. poor families (Ron and Kupferman 1996).InLalitpur, Nepal,
Goalpara India 1,000 n.a.
Nkoranza Ghana 37,000 32% poor households could get a free health card if they had a
ORT Philippines 1,300 n.a. letter from a community leader (Donaldson 1982).
ssss India 6,800 n.a. Elsewhere people who could not afford premiums would
n.a.isnotavailable.
simplypay user fees when using the service or not seek care.
Bwamanda hospital scheme stands out as having strikingly East and South-East Asian economies provide cover-
high coverage relative to other schemes. Many commu- age for the poorby issuing speciallow-incomecards ordevel-
nity-owned schemes (even those that were successful, such oping schemes specifically for the poor, rather than
as Thailand's health card scheme) failed to cover more integrating them into the main health insurance system. In
than 10 percent of the target population. Korea, however, the poor are theoretically integrated into
the system through subsidized premiums.
Financing-premiums, copayments,and cost recovery Allthe schemesrelied on funds other than those received
from premiums (table 5). Again,Zaire's Bwamanda scheme
Premiums were generally flat-rate premiums, paid on an stands out, with a cost recovery ratio of about 80 percent.
annual basis. In-kind payments generallywere not accepted. Cost recovery ratios are much lower in other schemes. In
Few schemes took proceeds from cooperative sales. Thailand about 35 percent of recurrent costs are covered.

INNOVATIONS IN HEALTH CARE FINANCING

170
However,thisneedsto be placedin the contextofhigh gov-
ernment cost recovery(on the order of 50-60 percent of
recurrentcosts at district hospitals).Thus less cost recov-
ery is achievedunder the health card schemethan under
the user fee system.
Guinea-Bissau'sAbotaschemeis generallyperceivedto
be successful.Formanyyearscommunityhealth insurance
was the sole sourceof financefor drugs in many commu-
nities. It should be emphasized,however,that the drugs
stockedat the communitylevelwere verybasic(just twelve
essentialdrugs), and that drugs were sold to health posts
at donor-subsidizedprices.About 90 percent of the costof
essentialdrugswascoveredby donors(Eklundand Stavem
1996).
As noted, the VHS schemein Madras,India, charged
premiumson a slidingscale,whichled to verydifferentrates
of cost recovery.Predictably,most of the people (74 per-
cent) that joined the schemewere uninsured and seeking
care at the health center level.Health centers record all
insured persons on separate registers and at the end of
each month present claimsto the fund holders,who pay
them.This procedurecoversno more than about 3 percent
TABLE
S
Cost recoveryunder the schemes
of recurrentcosts,andthe samelevelof revenuewouldoth-
erwisebe collectedthroughuser fees.
In some cases,however,managementstructureswere
verysimple.In Guinea-Bissauone of the mainreasonsfor
implementinga prepayment(asopposedto a user fee) sys-
tem wasits administrativesimplicityA villageleaderwould
simplyvisiteachhouseholdoncea yearand ask forpayment
of a fixed amount.Afteran initiallearningperiod,villagers
managedthe schemewell, alhough increasingeconomic
pressureseventuallyled to misuseof funds that threatened
the credibilityof the scheme.
In Nepal's United Missionschemevillagecommittees
made some very good decisions,particularlyrelating to
exemptions,butwere poorat accountingforfunds andcom-
municatingtheir plans to the rest of the village.
Fundmanagement. Fewdetailswereavailableabout man-
agementstructureor managementcapacityforgovernment-
run, mutual, and cooperative schemes. However, more
formaladministrativestructuresare likelyto be in placefor
mutual and cooperative funds, which may facilitate
management.

Scheme Costscovered Costrecoveryratio

Abota,Guinea-Bissau Fullcost recoveryon drugs n.a.
Boboye,Niger Not specified 149%of drugcosts
Bwamanda,
Zaire Aimingto coveroperatingcosts(thatis,excludingcapital 80% of operatingcosts(from patientrevenue,including
andstaffbenefitsin kind) 20% copayments)
CAM, Burundi Not specified(fundsgo to localgovemment) 34% of drugcostsfor cardholders
Farmershealthinsurance, Costsfullycoveredby premiums,government Copaymentis 10-50%of fee (govemmentpays70% of
Taiwan(China) contributions,andcopayments premium,beneficiarypays30%)
Healthcard scheme,Thailand Not specified Estimates
differ;about35% of recurrentcosts
Kasturba,
India Drugcosts,villagehealthworker remuneration, n.a.
andtravelcostsof hospitalteam
Healthinsurance
for informal Costsfullycoveredby premiums,govemment Copaymentis40-60% of fee (govemmentpays50%of
sector,Korea contributions,andcopayments premium,beneficiary
pays50%)
Lalitpur,Nepal Aimedto coveranyhealthpostexpenditureon drugs Lessthan 10%of healthpostrecurrentcosts
over Rs8,500a year(thestandardgovemmentdrugs
budget)
ORT scheme,Philippines Self-supporting
exceptORTfinanceof one doctor n.a.
andtwo nurses
VHS, India Unclear Rangefrom 1I.1%of recurrentcostsfor low-incometo
59.9% for high-incomesubscribers
n.a.isnotavailable.

RURALRISK-SHARINGSTRATEGIES
171
 Where premiumsare collectedat one point in time and
must meet financialcommitmentsfor an entire year,it is
essential(particularlyin high-inflationenvironments)that
the funds are invested.In its firstyearof operationGhana's
Nkoranzaschemeranintodifficultybecause ithad noinvest-
mentpolicyandhighinflationratesrapidlyerodedthe value
ofthe fund.In lateryears,however,the schemeboughttrea-
surybonds.Severalinnovativewaysto combatinflationhave
been found. In Zaire's Masisischemefunds were held by
the district pharmacy,which immediatelyconverted rev-
enuesinto drugs.In Zaire'sBwamandaschemefunds were
capitalizedby the NGO. This problem does not occur in
more stable economicenvironments,particularlywhere
investmentopportunities are safe and accessible(Japan,
Korea, and Taiwan,China).Under the Thai schemefund
holders did not haveto pay providersuntil the end of the
yearand couldinvestthe fundsduringthe year.Investments
often took the form of interest-bearingloans to commu-
nity members(Myers1989;Supachutikul1996).
Few of the studies reviewedprovided much informa-
tion on management information systems. In general
schemesfocusedfirston developingadequatefinancialman-
agementsystemsin orderto accountforfinancesandensure
that onlyinsured people could accessbenefits.Protecting
the schemeagainst fraudulent claimsoften proved diffi-
cult becausein manycontextsit wasdifficultcheckthe iden-
tityof the personseekingcare (McFarlane1996;Somkang
and others 1994).
Afterthe basic systemswerein place,theymightexpand
to coverutilization,whichis usefulin settingpremiums.Only
afterthese mechanismswere operationaldid attentionturn
to morecomplexissues,such as monitoringqualityof care.
Still,only the most developedof the schemesexamined,
suchas the one in Korea,had informationon these aspects.
Some community-ownedschemes (the Thai health card
scheme,the prepaymentschemeinBoboye,Niger)hadinfor-
mationonqualityfromspecialevaluations,but not fromrou-
tine data. In Taiwan(China), under the FarmersHealth
Insurancescheme,routine patient data now include fees
chargedby category,diagnosis,surgery,and lengthof stay.
Marketingandinformationand educationstrategiesvar-
ied immensely,dependingpartlyon a scheme'scatchment
area.In Nkoranzasubstantialeffortsweredevotedto mar-
keting and informationand education,with a seriesof dis-
INNOVATIONS
IN HEALTIHCAREFINANCING
172
trictwide meetings, campaigns,and informationsheets.
Under the ORT scheme in the Philippinesa registration
campaignwas held. In Chogoria(Kenya)a major market-
ing campaign was planned to launch the new scheme
(McFarlane1996).
Provider
paymentmechonisms
Allthe hospital-basedschemespaid the hospitalon a case-
based or fee-for-servicebasis.For most of the primarycare
schemesall funds collectedwere allocatedto the nearest
provideron a lump-sumbasis.TheThaischemeis an inter-
esting exception.The schemeprovidesaccessto all levels
of care if patients are referred. Initially,therefore, there
was a fixed formulafor allocatingfundsbetween different
levelsof the system.For example,in ChiangMai 15 per-
cent of funds were ultimatelyretained by the villagecom-
mittee, 20 percent by the health center,33 percent by the
communityhospital,and 32 percent by the provincialhos-
pital (Supachutikuland Sirinirund1993).Theseallocations
variedsomewhatbetweenprovinces.Sincetherewerestrict
referralproceduresin order to be eligiblefor fund cover-
age, these fixed proportions could be estimated to some
degree.However,underthispaymentsystemhigherservice
levels, particularlydistrict hospitals, often felt that they
receivedan unfairshare of funds. In more recent yearsthe
referral procedures have become more lax and there is
now discussionabout linkingallocationto actual utiliza-
tion of facilities.In order to institutesucha paymentmech-
anism,however,strongerinformationsystemsare required.
of health care
Provision
Beneftspackage.On the whole,benefitspackageswere
poorlydefined. Some schemeshad exclusions(for exam-
ple, sexuallytransmitted diseasesin Vietnam, dental ser-
vices in the Philippines's ORT scheme), but otherwise
schemestended to coverallthe servicesavailableat the par-
ticipatingfacilities.The main problemwith this approach
was high enrollmentrates amongpeople with preexisting
conditions,particularlychronicillnesses.Kenya'sChogoria
hospitalschemeinitiallyhad a verybroad benefit package
coveringall such conditions. But when the scheme was
recentlyreviewedthe benefitspackagewas definedmuch
more tightlyin order to excludeboth the elderlyand those
withpreexistingconditions(box1).Veryfewofthe schemes BoxI Exclusions andlimitsunderKenya's
used revenueto provide non-personalservices. ChogoraHospitalHealthInsuanceScheme
* Diseasesthatwerediagnosed priorto joining
the scheme
Integration
ofhealthcareservices.
Hospital-basedschemes or withinthewaitingperiodare notcovered.
tended to focus exclusivelyon the hospitalleveland have * Coverage isdiscontinued
whena personturns65.
limitedconnectionswith primarycare. There were, how- * Treatment ofpatientswithAIDSisprovided upto a maxi-
ever, some exceptions. The Chogoria scheme covered pri- mum of Sh36.00 a year.
everyprovidersandusedThemasgoatekeepers tovered
hospial * Treatment illnessis limited
of psychiatric to Sh 68.00per
maryprovidersandused themas gatekeepersto the hospital policyperyear.
level.In Zaire's Bwamandaschemeprimaryclinicsacted * Expensesthatareassociated withnormalorabnormal preg-
as gatekeepersto the hospitalscheme,but feesfor services nancyarenotcovered(although theoperation feefora first
at thislevelwerenot coveredbythe scheme(Moens 1990). Caesariansectioniscovered).
Community-ownedschemes tended to have clearer and * Readingglasses,eyeandeartests,andhearing aidsare not
strongerreferral structures.In Taiwan(China) and Korea covered.
* Self-inflicted
injuries
arenotcovered.
providers are predominantlyprivate and referral systems * Birthdefectsandcosmetic surgery are notcovered.
are extremelyweak (if they existat all).Thelack of a gate- * Dentalprocedures are notcovered.
keeper has contributed to rapid cost escalationin these * Medicalexaminations arenotcovered.
economies. * Procedurescarriedoutfornonmedical reasonsarenotcov-
ered(circumcision,
forexample).
Quality of care.Severalcommunity-ownedand NGO McFarlane
Source: 1996.
schemes used revenues to expand access to health ser-
vices,but fewmadeeffortsto improveother aspectsof qual- Efficiency
ityof care (the exceptionsbeingNiger's Boboye,Vietnam's
QNDN, and Mexico'sprepaymentscheme).None of the Administrative efficiency.In somecases(suchas Guinea-
facility-owned schemeshadexplicitlinksbetweenthe intro- Bissau'sAbota) the prepaymentroute was taken because
ductionof the schemeand attemptsto improvequalityof it was seen as being administrativelymore efficient than
care. user fees. In other instances (suchas Ghana's Nkoranza)
Under Vietnam'sQNDN schemespecial effortswere there appear to be high administrativecosts (particularly
made to improve the quality of care because there were the use of time of scarceskilledpersonnel),yet substantial
(officially)no feesat hospitals.Thus specialeffortshad to problemsremain in the administrationof the scheme.
be madeto attract people to the scheme.In both Mexico
and Vietnamqualityimprovementsfocused on the hotel Allocativeand technicalefficiency.Fee-for-servicereim-
aspectsof care. bursementhas a number of well-knownshortcomings.It
Few schemesadopted specialpharmaceuticalpolicies, provideslittleincentivefor efficiencyon behalf of the hos-
alhough the ORT schemein the Philippinesmanagedto pital.It doesnot guardagainstproblemsof costescalation.
negotiate favorableprices for essential drugs purchased Andit is administratively
complex.Fee-for-servicepayment
from local suppliers. givesprovidersincentivesto overserviceand overprescribe.
In most of the cases examinedoverprovisionwas unlikely
Assessment of Performance to be an issue becausethe providerwas a governmentor
missionfacility,
andstaffwerepaidona salarybasis.However,
Healthstatusimprovement it was a concern in Masisi,Zaire, where part of hospital
revenue was used as incentive payments for doctors
No studieshavebeenmade evaluatingthe impactof these (Notermanand others1996).At Chogoriahospitalthe med-
schemeson health outcomes. icalofficerin chargenoted that "it has taken some time to

RURALRISK-SHARINGSTRATEGIES
173
educateourprescribersto treatpatientsonthe schemein the insuranceschemesexaminedset premiumson a com-
a similarmannerto otherpatients,keepinginviewthecost munity-ratedbasisand thus entaileda subsidyfromthe
oftreatment"(McFarlane1996,p. 7). healthyto thesick.Fewschemes,however, adoptedsliding
Allthe schemesexaminedwerefairlyweakpurchasing scales,andthe useofflat-ratepremiumsimpliesregressiv-
agents;fewdefinedcost-effective packagesof care,fewityinfinancing.AslidingscalesysteminBwamanda allowed
implemented strongreferralandutilizationcontrolsystems
lowercopayments formoredistantresidentsbut wasaban-
to optimizeefficientuseofdifferentlevelsofthehealthcare
donedbecauseoftheextraadministrative costandbecause
system,and fewimplemented a management itdidnotappearto affectutilization.
information It did,however,improve
systemthat monitoredcost-effectiveness or appropriate-
enrollment andthuswouldhaveeffectively reducedadverse
nessof caredelivered. selection.
Not onlyweretheschemesweakpurchasers,theyalso Moreimportantis thequestionof whetherinsuranceis
sometimesintroducedinefficiencies intothe system.For
moreor lessregressivethan otherfinancingalternatives.
example,manyhospital-based schemeslargely ignoredpri-
Althoughflat-ratepremiumsarelikelyto be lessregressive
marycare.Althoughthe documentation doesnot explore
thanuserfees,theymaybe moreregressivethangeneral
theimpactofthis,it wouldseemlikelyto resultin under- taxrevenuefinancing. Empirical workin anumberofindus-
utilizationof healthcenters(leadingto facilities
operating
trialcountrieshasestablished that socialinsuranceis more
atlowcapacity andrisingunitcosts)whilepatientsaretreated
regressive.
lessefficiently
atthehospitallevel.Again,Zaire'sBwamanda Affordabilityisprobablythekeyissueintermsofequity.
schemediffers,becauseaccessunderinsuranceto thehos- Fewschemesmadespecialallowances forpeoplewhocould
pitalrequiresa referralfroma healthcenter. not affordto paythepremiums.In mostcasespeoplewho
Thehospital-based schemesin Nkoranza,Ghanaand couldnot affordpremiumswererequiredto payuserfees
MasisiandBwamanda, Zaire,experienced rapidcostesca-
instead.Thusthe effectiveness of insurancein protecting
lation,at leastin their earlyyears.In Koreaand Taiwan the poorraisesthequestionoftheeffectiveness of exemp-
(China),wherefee-for-service paymentsalsoexist,suchtionmechanisms.
problemshavepersisteddespiteeffortsto containcosts Severalschemesthat examinedtheissueof affordabil-
throughcopayments. Thelackofa gatekeeper hasalsocon-
ityacknowledged thatit couldbe a problem,althoughthe
tributedto rapidcostescalationin theseeconomies. evidencewasnot alwaysclearcut.Formoderateto large
Presumably, if fundsareheld at the community level,
low-income households in Nkoranza,Ghana,theestimated
thenit is in the interestof the fund holderto makesurecostofpremiumsamountedto 5-10 percentoftheannual
that unnecessary utilizationof expensivesecondary-level
householdbudget,whichmaywellconstitutea financial
servicesdoesnot occur. barrierto membership(Somkangand others 1994).In
Muyinga,Burundi,27 percentof respondentsto a house-
Financial
efficiency.
Overtimemostoftheschemeswere hold surveystatedthat financialinabilityto purchasea
ableto devisewaysto investthe revenueraisedfromthe cardwasone of the mainreasonstheydid not participate
insurancescheme.However,schememanagershad not in the scheme(Arhin1994).In Mexicoabout20 percent
alwaysthoughtout this aspectadequatelypriorto imple- of enrolleesin theperinatalprepaymentschemedropped
mentation,andheavyfinanciallossesin thefirstyearcould out, andthiswasmainlyattributedto financialinabilityto
adverselyaffectthe financialefficiencyof anyschemefor keepup payments(Ensor1995).
a longperiod.
In utilization.
Onlythe pilotprojectin Boboye,Niger,
Equity analyzedhowutilizationpatternsvaried by incomegroup.
In Boboye
itwasfoundthatutilization
ratesamongthepoor
Infinancing.
Risksharinghasbeenpromotedas ameans hadgoneup sinceimplementation
ofthescheme.Moreover,
ofencouraging
moreequitablefinancingofhealthcare.All whenpaymentsby the poorwhousedgovernmentfacili-

INNOVATIONS IN HEALTH CARE FINANCING

174
tieswere comparedbefore and after the scheme,total pay-
ments had gone down (Diop,Yazbeck,and Bitran 1995).
The schemein Boboyewas, however,very differentfrom
those implementedelsewhere.And although the Boboye
experiencesuggeststhat it is possibleto designan insur-
ance scheme that has positive equity effects in terms of
whobenefitsfromthe service,it byno meansdemonstrates
that this is alwaysthe case. The compulsory(tax-based)
systemin Boboyeperformedbetter than a user fee exper-
iment in a comparisondistrict.
Manyofthe casestudiesdiscussedutilizationandenroll-
ment patterns by geographicallocationof the household.
There is fairlysubstantialevidencethat utilizationincreases
far more amonginsured householdslocatednear a health
care facilityand that these householdsare alsomore likely
to join such a scheme (Criel 1992; Donaldson 1982;
Noterman and others 1996).The private costsof seeking
care form a considerablebarrier to accessingcare. Under
most schemespeoplepaythe samepremiumwhereverthey
live; thus those located far from the facility (who might
be part of poorer, more remote rural communities)end
up cross-subsidizingthose who live close. This situation
might also lead to a form of adverse selection in which
remote householdsdrop out of the scheme because the
premiums are so high that it is not worthwhilefor them
to join. This can lead to a viciouscirclein which average
utilization rates rise even higher and more people drop
out.
The only documentedattempt to implementa sliding
scalebased on geographicalproximityto the facilitytook
placein Bwamanda,wherefor one yearthe levelof copay-
ment was based on a slidingscale dependingon distance
from the hospital (a 20 percent copaymentwasleviedfor
the nearest group, down to 5 percent for the most remote
group;Criel 1992).Under this systemenrollmentwent up
among the remote group but their utilization of the hos-
pital did not. Furthermore, the implementation team
thought that enrollmentamongthis groupmighthavegone
up in any casebecausethe schemebecame better known.
The slidingscalewasdropped becauseit appearedto have
no impact on utilization and was administrativelymore
complex.
Mostof the schemesin Sub-SaharanAfrica allowedaccess
only to public health care facilities.Thus there is unlikely
RURAL RISK-SHARING STRATEGIES
175
to be a "marketresponse"to the establishmentofinsurance
in remoteareas.In Korea,however,whereprivateproviders
were the mainmode of servicedelivery,the establishment
ofinsuranceschemeswasfoundto be insufficientto encour-
age providersto relocate to rural areas.The government
had to initiateseparateprogramsto financeremote health
carecenters,subsidizeinsurancesocietiesin ruralareas,and
use tax incentivesto encouragethe developmentof clinics
andhospitalsin these areas(Peabody,Lee,andBickel1995).
Finally,exclusionsand limitson the benefitsof schemes
have implicationsfor equity in the utilizationof services.
In schemesthat covercatastrophiccostsof care,settingcer-
tain exclusionsmaybe essentialto guard againstadverse
selection.However,these exdusionsarelikelyto affectmore
vulnerablegroups,suchasthe elderlyandpeoplewithAIDS
(seebox 1).
Equitybetweenschemes.In the fewcountrieswherethere
is substantialcoverageby rural risk-sharingschemes (for
example,in Chinapriorto the breakdownof the Cooperative
MedicalSystemandin Thailand)equitybetween schemes
has becomea majorissue.In Chinathe typeof careto which
peoplehadaccessvariedsubstantially accordingto the wealth
of the community;poor communitiescouldoften onlyafford
to coverprimarycare servicesand did not cover inpatient
servicesat countyhospitals(WorldBank 1996).
Koreanowhas more than 600 insurancefunds.There is
substantialgovernmentsubsidyandregulationofthe funds
but it is not clear that this intervention promotes equity
between schemes.The governmentregulatesreimburse-
ment rates andprovidesa 50 percent subsidyto premiums
charged.However,premiumsaresetbythe individualinsur-
ance society,which implies that schemes coveringmore
affluent groups can buy bigger and better benefit pack-
ages,and that the government'ssubsidyofthisbenefitpack-
age willbe higher in absolute terms (Peabody,Lee, and
Bickel1995;Yu and Anderson1992).
Few of the countries examinedprovided evidence on
equity issues between formal and informal schemes. In
Japan, Korea,and Taiwan(China) the schemesthat cover
the informalsectorarenowfullyintegratedwiththe national
health insurancescheme.In manyother countriesformal
schemes are nonexistentor very limited. Thailand is an
exception,and substantialquestionsare beingposed about
the differentbenefits and governmentsubsidiesprovided
to participantsin formal and informalschemes(Khoman
in thisvolume).In 1994it was estimatedthat total expen-
diture per capitaunder the CivilServantsMedicalBenefit
Schemewas 9.4 times higher than that under the Health
Card Scheme,and expenditureunder the SocialSecurity
Schemewas3.7timesthat under the Health Card Scheme.
Differencesare even more marked if the level of govern-
ment subsidyto the schemesis considered.Government
subsidyper capita to the CivilServantsMedicalBenefit
Schemeand the SocialSecuritySchemewas27.0timesand
4.4 times that to the Health Card Scheme(Supachutikul
1996).
Consumer
satisfaction
Most of the schemespaid little attentionto consumersat-
isfaction,or even to what consumerswanted, duringthe
designphase of the scheme.None of the studies reported
surveysof consumersatisfaction,and few had carried out
marketingsurveysprior to implementation.As judged by
demandfor the schemes,consumersatisfactionwas often
low.
Sustainability
andreplicability
The evidencefrom these experience suggeststhat there
are severalthreatsto the scopefor raisingrevenuethrough
rural risk-sharingschemes:
* The smallscale of the majorityof the schemesexam-
ined
* Adverseselection,leadingto progressivelysmallerrisk
poolsand highercosts
* Heavy administrative structures and costs in some
schemes.
These constraintsoften led to low levelsof cost recov-
ery.Furthermore,schemesoften receivedsubstantialinput,
particularlytechnicalinput, from donors and expatriates,
which suggeststhat they might otherwisebe unsustain-
able. Most of the schemesexaminedwere relativelyshort
livedor,if still operating,were recentlyinitiated.The main
exceptionsare someof the IndianNGO schemesand those
schemesin EastAsiathat havebecomepart of government
nationalhealth insuranceschemes.
INNOVATIONS IN HEALTH CARE FINANCING
176
Schemesthat allowedindividualmembershipoften faced
problemsof adverseselection.In Nkoranza, Ghana, pre-
miumswere set on an individualbasisbut the entirehouse-
hold had to join.However,the failureof insurancescheme
workersto signup allmembersof a householdcontributed
to the failureof the scheme(Somkangand others 1994).
Further prevention measures against adverse selection
include requiringthat a minimumnumber or portion of
householdsin a villageor administrativeareajoina scheme.
If enrollmentin a schemeis allowedover a long period
and there is no waitingperiod,then people tend to enroll
whentheyneed care.Thisis a bigproblemin hospital-based
schemes,wherethe need for serviceis most unpredictable
and bears higher financialconsequences.For community
schemescoveringmainlyprimarycareit is easierfor house-
holdsto predict utilizationand decidewhether it is worth-
whileto join.
The main problem with benefit packages that do not
excludepreexistingconditionswas high enrollment rates
amongthosewithpreexistingconditions,particularlychronic
illnesses.
Twoof the schemes,the ORT schemein the Philippines
andthe RCMSschemein China,claimto fullyrecovercosts,
but they did not set premiums on an actuarial basis.
Presumablyin these casesinput levelsor copaymentswere
adjusted to reflect the revenue received.The advantages
of this approach over a traditionalactuarialapproachare
not clear.
With one exception, the schemes examinedwere vol-
untaryschemes.Manyofthe problems(particularlyadverse
selection)associatedwiththe schemesstemmedfrom their
voluntarynature.It hasbeen arguedthat in developingcoun-
try contextsmandatoryschemesfor informalsector work-
ers are unlikelyto be feasiblebecausethere is insufficient
knowledgeabout the number and locationof ruralhouse-
holds. Identification, income assessment, and contribu-
tion collectioncan rapidlybecomean expensiveprocessin
rural areas. However,authoritiesin Boboye,Niger,man-
agedto implementa mandatoryinsuranceschemethrough
an earmarkedtax. More investigationof the prospects for
implementingmandatoryschemesin developingcountries
is needed.Clearly,mandatoryforms of risk sharingwillbe
easier to implementin areaswherelocalgovernmenttaxa-
tion systemsare extensiveand welldeveloped.
 The inclusioncriteria for schemesin this revieweffec-
tivelyexcludedmany of the better-knownLatinAmerican
schemes.Forexample,the Igualasschemein the Dominican
Republicwas excludedbecausethe insurancecarrier (or
health maintenance organization)is essentiallya private
for-profitbusiness(La Forgia1990).Ecuador and Mexico
both have schemes for informal sector workers that are
subsidized entirely by premiums paid by formal sector
workers (DeRoeck and others 1996). It would be inter-
esting to explore whether special conditions in Latin
Americamade the typicalrisk-sharingscheme discussed
here inappropriate.
Lessonsfrom Rural Risk-Sharing
Experience
Context
The first lesson is that context matters. Some schemes
havecomeinto existencein responseto economicandpolit-
ical crisis (Bwamanda,Zaire). The successof others has
been facilitatedby economicfactors(Korea)or terminated
by them (China). The externalenvironmenthas an influ-
ence on whether and what type of risk-sharinginitiative
mightbe propitious,and on the replicabilitywithina coun-
try or transferabilitybetween countriesof suchexperience.
A considerationof contextprovidesa wideropportunityto
reviewhow the overall risk-sharingfunction in health is
being implemented,includingthe use of tax-basedhealth
expenditureby governmentand other public and private
schemes.
Design
Designis critical.It is temptingto concludethat the expe-
riencewith ruralrisk sharinghas been a litanyof disasters.
Manyofthe schemesexaminedherehadfundamentaldesign
flaws.Althoughmany of these flawsappear obvious,they
clearlywerenot to the people designingand implementing
the schemes.Thus a number of points need to be stated
clearlyand disseminatedwidely.
Protectagainstadverseselection.Waysto ensure against
adverseselectioninclude:
RURALRISK-SHARINGSTRATEGIES
177
* The minimum action that should be taken is making
the householdor eventhe villagethe unit of member-
ship and implementingthis policystrictly.
* Makingthe schemecompulsory(althoughthis does not
seem to be feasiblein manycontexts).
* Stipulatingthat a certainportion of householdsin the
villagemust jointhe schemebeforethe villageis allowed
to enter the scheme(as is done under Thailand'shealth
care project and India's Kasturbahospitalscheme).
* Preventingpeoplewith preexistingconditionsfrom reg-
istering,or limitingthe benefits availablefor suchcon-
ditions(althoughthe advantagesof suchmeasuresneed
to be weighedagainstthe equityimplications).
Ifenrollmentthroughout theyearisallowed,establish
a wait-
ingperiodbeforeservicescanbeaccessed. Householdsshould
not be able to join the schemewhentheyget illand decide
to seekcare.Althoughmostschemesrecognizedthisin their
initialdesign, in many cases few people were enrolledat
the end of the registration period and the period was
extended in the hope that more people would join.
Administrative capacityis requiredto managerequiredwait-
ing periods.
Supportthe referralsystemand definea primarypoint of
contact.Manyschemes,particularly hospital-owned schemes,
have paid little attention to the effect the schemehas on
other levelsof the health care system.Such a segmented
approachadverselyaffectsnot onlythe providersexcluded
from the scheme but also the providers in the scheme.
Referralsystemsexistto ensurethat patients are treated in
the most appropriate and cost-effectivemanner and to
protectthe financialviabilityof hospitals.Insuranceschemes
cannot affordto ignorethese issues.
Developan investmentstrategyforfundspriorto receiving
them.In high-inflationenvironmentsdelaysof even a few
monthscanquicklydepleteinsurancefunds.Aninvestment
strategyis essentialto guard againsterosionof funds.
Becoming
an activepurchaser
The schemesexaminedhave largelyfailedto:
* Negotiatespecialpriceswith providers
 * Definebenefit packagesto ensuredeliveryof onlycost-
effectiveservices
* Monitorthe qualityand appropriatenessof care
* Use paymentmechanismsto encourageefficient,qual-
ity service
* Developstrong essentialdrugspolicies.
Recenthealth sectorreformliteraturefrom both indus-
trial and developingcountriesemphasizesthe importance
of informedpurchasersin the healthcare sector (Saltman
1995).But most rural insurancefunds remain pure finan-
cialintermediaries-that is, collectorsof contributionsand
payersto a singleprovider.Considerablescope existsfor
the developmentof a more activerole,includingusingthe
payment system to change incentives, to include new
providers(suchas primarycareor NGO providers),andto
rewardqualityimprovement.Risk-sharingschemesshould
not be seen simplyas a source of finance, but rather as
ways to organize health servicesfinancing and delivery.
Thepotentialthat risk-sharingmechanismshaveforimprov-
ing systemperformanceis often untapped.
Risksharingand the ruralpoor
Most risk-sharingschemesappear not to be targeted at the
ruralpoor, but at the rural middle classes:
* They seldomallowpaymentin kind
* Theyhave flat-ratepremiums
* Theyhave no exemptionpolicy
* They mayrequire substantialcopayments.
Geographicaland incomeinequitiesin financingcould
be reduced throughthe use of a slidingscale.In principle
it shouldbe easierto implementa slidingscalefor payment
of an annual premium than for multiple user fee-type
charges.However,the user fee literature has highlighted
the problemsinvolvedin successfullytargetingexemptions
basedon income(Parkerand Knippenberg1991;Willisand
Leighton1995).Slidingscalesbased on geographicalloca-
tion would be much less susceptibleto targeting errors,
and appear to be worth trying.
Questioning
focility-owned
schemes

Listening
to consumers About 30 percent of the schemesidentified were owned
and initiatedby healthfacilities,mainlyhospitals.Hospital
A substantialamount of the literatureon rural insurance managementoften has a strong incentive to implement
focuseson whether demand existsamonginformalsector insurance;if financialsupportfrom governmentis limited
workersandruralpeopleforhealthinsurance.In caseswhere and the population cannot afford to pay cost-recovering
healthfacilitiesare chargingsubstantialamountsfor careit fees, then health insuranceschemes mayappear to solve
would seem likelythat demand for healthinsurancedoes manyproblems.However,hospitalmanagershavefewincen-
exist.Yetfew schemeshaverooted their designin a survey tivesto design and implementhealth insurance schemes
of consumerdemand or have evaluatedwhether schemes that protect the interestsof beneficiariesin the most cost-
matchpeople's expectations. effectiveway,or that coverthe entire population.
Marketingeffortsare likelyto be weak in this context. The reviewof facility-owned schemessuggeststhat, with
Toomanyprojectdocumentsseem to assumethat market- the exceptionof Bwamanda,hospital-ownedschemes:
ing techniquesexplainingthe principles of health insur- * Have little incentiveto improvethe qualityof care
ancewillconvincepeopleto jointhe schemes.Butconsumers * Tend to overlookprimarycare
reallyneedto be consultedduringthe designphase.More * Tendto seek overlyfavorableremuneration(particularly
important,they need to have confidencethat the scheme throughretrospectivefee-for-servicepayment)
is managedin theirinterest,andthat it willensuretheyhave * Have few incentivesto improveefficiency.
accessto qualityserviceswhentheyneedthem. Confidence Even if a scheme is designed principallyto cover the
in a risk-sharingmechanismis an abstractnotion:people's costsof hospital care, it may be preferable that responsi-
assessmentof the competence,quality,and professionalism bilityformanagingthe schemeand ownershipof the insur-
of a health care provider is usuallyaccurate. If the right ance fund rest outside of the hospital, or that the fund
type of care is not available,insuranceinitiativescannot itselfownor be responsiblefor managingalldistricthealth
expectto succeed. services.

INNOVATIONS IN HEALTH CARE FINANCING

178
The roleof govemment
A number of important roles emerge for governmentin
the developmentof rural risk-sharingschemes.
Financing.Policymakersshould recognizethat the rev-
enue-raisingpotential of rural risk-sharingschemes,par-
ticularlyin verypoor countries,is likelyto be limited.Thus
they should not set ambitiouscost recoverytargets under
such schemes.Someof the literatureappearsoverlyambi-
tious in terms of the potentialrevenue gainsfrom health
insurance.For example,Shaw and Griffin (1995, p. 55)
claim that "health insuranceis virtuallythe onlypractical
instrumentgovernmentscan use to get out of the expen-
sive businessof providingacross-the-boardsubsidiesfor
hospitalcare." And De Ferranti (1985, p. 41-2) saysthat
"there has beenrecognitionof the high costrecoverypoten-
tialof suchschemes,sincerelativelymodest coveragecharges,
when spread across an entire participantpopulation,can
raise substantialrevenue."
Theevidencefrommostof the schemesexamined,indud-
ing those in affluenteconomiessuch as Koreaand Taiwan
(China),suggeststhat costrecoverylevelsunder ruralinsur-
ance are likelyto be limitedin most developingcountries.
Likeuser fees,insuranceshouldbe seen as a wayto top up
government budgetary funding and to introduce or
strengthenmanagementof the health system.In countries
where rural schemesare widespreadand havebeen inte-
grated with nationalhealth insuranceprograms,there are
substantialgovernmentcontributionsto the scheme,and
there are often alsosubstantialcopayments.
Governmentsubsidycanbe madedirectlyto the provider
(for example,in most community-owned schemesthe gov-
emnment continuesto fundthe bulkofserviceproviders'recur-
rent costs,and revenuefrom the insuranceschemeprovides
a top up) or be directed to the insurancefund itself(as in
Japan,Korea,andTaiwan, China,andisproposedinThailand).
The preferredgovemmentstrategydepends principally
on the ownershipof providerunits. In instanceswherethe
privatesector is dominant,the governmentsubsidizesthe
fund or poor householdsbuy into the fund. But by con-
tributingdirectlyto the fund, rather than to the provider,
governmentscan help developeffectivepurchasingpower
and strengthenfund management.
RuRAL RISK-SHARINGSTRATEGIES
179
Governmentgeneral revenuefinancingcan be used to
solve some of the problems associatedwith rural insur-
ance.For example,it can be used to purchasehealth ser-
vicesforthe poororto offsetregionalinequities.Government
capacityto do this dependson whether it is subsidizing15
percent or 65 percent of the population.
Policyframeworkandoperational guidelines.
Bytheirnature
most insurance schemes are independent local efforts.
However,in countries'thathave had the most successin
increasingruralinsurancecoverage(China,Indonesia,Korea,
Thailand)the schemeshavetakenplacewithinclearlydefined
policyframeworksand haveoften benefited from specific
operationalguidelines(box 2).
Theeffectiveness ofnationalguidelinesin aidingthe devel-
opmentof insurancedependsonhowsensiblethe guidelines
are. In the Philippinespilotprojectshavebeen used to help
developnationalpolicyand legislationon ruralinsurance.
Training.Many countrieswith dearly defined policies
and operationalguidelineshave also developed training
packagesto help communitymembers manageinsurance
schemes.Such trainingpackageshelp ensurethat lessons
learnedby onecommunityare passedon to others.Training
packagesshouldcover,amongother things,the designrec-
ommendationsdescribedabove.
Ensuringaccountabilityof fundholders.With increased
decentralization,
particularlytheestablishment ofautonomous
health care facilities,there are critical questionsabout to
whomhealthinsurancefundmanagersareresponsible. Where
funds are ownedby government,cooperatives,or commu-
nitiesthese issuesmaybe lesspressing,but forfunds owned
byNGOs and facilitiesthe linesof accountability to benefi-
ciariesmaybeextremelyweak.Governmentmustensurethat
fund managersare accountable.In few of the casesexam-
inedwasthisresponsibilitycarriedout effectively.
Conclusion
Debateaboutthe potentialforruralinsurancehas provoked
quiteextremepositions.The diversityof schemesfound in
the literature means that many of these positionscan be
supported.Whilesome schemeshaveoperatedwith com-

Box 2 Guidelinesfor Indonesia'sDana Sehat
communityinsurancescheme
* The schemeisrun bythe community for the healthbenefit
of its members.Localinstitutions(familywelfareagencies,
villagecooperatives,
religious canapplytoman-
organizations)
agethe DanaSehat.
* Premiumpaymentsare supportedbylocaleconomicactiv-
itiessuchas cooperativesof crops,handicrafts,
moneylend-
ing
ing,an
andsoon
soon.
* Vision,mission,objectives,
andprogramidentificationshould
be basedon deliberationand agreementamongcommu-
nitymembers.
* Theschemeis controlledprimarily byitsmembers.Thegov-
emmentprovides toolsandguidelines
on howitsshouldoper-
ate;community membersshouldmonitortheseprocedures.
* ThreedifferentlevelsofDanaSehatdevelopment havebeen
identified,risingfrom simplecommunitymanagementof
small-scale schemesto large-scale,
complex,professionally
managedschemes.
Source: Brahim,
Suwandono, andMalik1995.
I Certainly, widely disseminating the lessons of experience
plex administrative structures, others have had very simple
ones. While some schemes have had substantial problems
with adverse selection, others have avoided such problems
almost completely.While some schemes have devised incen-
tivesto promote efficient use of the health care system, oth-
ers have probably increased inefficienciesin the system. To
engage in further discussion about the potential for rural
insurance we need terms that distinguish more carefully
among different types of schemes and their objectives.
The framework presented here is a first step along this path.
Well-designedinsurance schemes may have even greater
potential for improving health systemperformance-partic-
ularlyqualityand efficiency-than forraisingsubstantialaddi-
tional finance.This is particularlylikelyinpoor communities,
where there simply is not a lot of extra money available.
Delicate organizationalchanges-including tighter refer-
ral control, contracting arrangements between purchasers
and providers, accreditation and service quality improve-
ments, and performance-related pay-all might be intro-
duced under the guise of a shift toward a health insurance
system. Insurance should be seen as a supporting strategy,
not as an exclusivefinancing alternative that might dose off
other options and divert attention from the need to improve
IN HEALTHCAREFINANCING
INNOVATIONS
180
servicedeliveryIn the Bwamanda(Zaire)and CAM (Burundi)
schemes, prepayment, subsidy,and fee-for-serviceelements
coexist. Neither user fees nor voluntary prepayment strate-
gies deal adequately with the needs of the poorest people.
Although some rural risk-sharingschemes have been suc-
cessful, several common failings are evident:
* Schemes in low-incomecountries have generallyachieved
limited population coverage
* With few exceptions, cost recovery rates under the
schemes are low
* The schemes examined have a limitedability to protect
the poorest members of society.
These criticisms are damning since they strike at the most
centralreasonsfor promotinginsurance-that is, the sup-
posed ability of such schemes to raise significant amounts
of revenue in an equitable manner.
Many of the schemes examined suffered from poor
design. It could be argued that with better design, some of
the core problems identified above could be resolved.
may alleviatesome of the problems that have been described.
But many of the schemes reviewed had extensive external
technical assistance from well-informed experts. It would
be impossible to widely replicate health insurance schemes
for the informal sector and to provide such intensive tech-
nical support. The few success stories, such as Zaire's
Bwamanda scheme and Nepal's Boboye scheme, demon-
strate that it is possible to design and operate a successful
health insurance scheme for the informal sector. They do
not, however, address the question of whether it is feasible
to do so on a widespread basis. The evidence suggests that
it is probably not feasible.
The schemes reviewed in this paper were mainly volun-
tary schemes run by hospitals, communities, or NGOs.
Other approaches warrant further exploration. In particu-
lar, many of the problems associated with the schemes
stemmed from their voluntary nature. More information is
needed on both the feasibility and desirability of compul-
sory schemes. Cooperative and mutual insurance organi-
zations formed the foundations of social health insurance
systems in Japan and in many Western European coun-
tries. This review unearthed little information on such orga-
nizationsin developingcountries.Finally few schemesappear
to have used exemption mechanisms or premiums that were
graduated according to income or household location. If
insurance is to succeed in protecting the poor, then much
greater use of these mechanisms will be required.
References
Alain, Henry, Guy-Honore Tchente, and Philippe Guillerme-
Dieumegard. 1991. Tontineset Banquesdu Cameroun:Les
Principesde laSocietedesAmis. Paris:Karthala.
Arhin,Dyna. 1994."HealthCard InsuranceSchemein Burundi: Diop, F., A. Yazbeck, and R. Bitran. 1995. "The Impact of
A SocialAsset or A Non-ViableVenture?"SocialScienceand
Medicine39(6): 861-70.
. 1995."HealthInsuranceinRuralAfrica."TheLancet345: Donaldson,D.S.1982.'AnAnalysisof HealthInsuranceSchemes
44-45.
Arrow,Kenneth.1963."Uncertaintyand the WelfareEconomics
of MedicalCare."AmericanEconomicReview53(5):941-73.
Atim,C. 1997."SocialMovementsandHealthlnsurance:ACritical
Evaluationof Voluntary,Non-ProfitInsuranceSchemeswith
Case Studiesfrom Ghana and Cameroon."Paper presented
at a Universityof Antwerpconferenceon the economicsof
healthinsuranceinlow-andmiddle-incomecountries,January
17-18,Antwerp.
Bennett,Sara,andE. Ngalande-Banda.1994."PublicandPrivate
Rolesin Health:A Reviewand Analysisof Experiencein Sub-
SaharanAfrica."StrengtheningHealthServicesPaper6. World
HealthOrganization, Divisionof Strengthening
HealthServices,
Geneva.
Bloom,G., and others. 1995.FinancingHealthServicesin Poor
RuralAreas:Adaptingto EconomicandInstitutionalReformin
China.Brighton:Institutefor DevelopmentStudies.
Bogg,L., and others. 1996."TheCost of Coverage:RuralHealth
Insurance in China." Health Policy and Planning 11(3):
238-52.
Carrin,G., and others. 1996.TheRefornof theRuralCooperative
MedicalSystemin the People'sRepublicof China:InitialDesign
andInterimExperience. Geneva:WorldHealth Organisation. Ikegami,N., and T. Hasegawa.1995."TheJapaneseHealth Care
ChabotJ., M. Boal,andA.Da Silva.1991."NationalCommunity
Health Insuranceat the VillageLevel:The CasefromGuinea-
Bissau."HealthPolicyandPlanning6(1): 46-54.
Creese,Andrew,and Joseph Kutzin.1995. "Lessonsfrom Cost
Recovery in Health." Discussion Paper 2. World Health
Organization,Geneva.
Criel,B. 1992."CommunityFinancingSchemes:GiveThemTume.
Discussionof TwoPrepaymentSchemes Conducted at the
DistrictLevelin Zaire."Universityof Antwerp,Belgium.
Criel,B., and W van Lerberghe.1997."The BwamandaHospital Jajoo, U.N. 1992. "Risk-Sharingin RuralHealth Care." World
InsuranceScheme:Effectivefor Whom?A Studyof itsImpact
onHospitalUtilisationPatterns."PaperpresentedataUniversityJazairy,I., M. Alamgir,and T Panuccio.1993. TheStateof World
of Antwerpconferenceon the economicsof health insurance
inlow-andmiddle-incomecountries,January17-18,Antwerp.
Dave,P.1991."CommunityandSelf-FinancinginVoluntaryHealth
ProgrammesinIndia."Health PolicyandPlanning 6(1):20-31.
Dave,P,andPeterBerman.1990.TheCostsandFinancingofHealth
Care:Experiencesin the VoluntarySector.CaseStudy 1: The
Voluntary HealthServices,
Madras.NewDelhi:FordFoundation.
de Ferranti,David.1985."Payingfor HealthServicesinDeveloping
Countries:ACallfor Realism."WorldHealthForum6:99-105.
DeRoeck,D., and others. 1996.RuralHealthServicesatSeguridad
SocialCampesinoFacilities: Analysesof FacilityandHousehold
Surveys:EcuadorBethesda,Md.: Abt AssociatesInc.
AlternativeCost RecoverySchemeson Accessand Equity in
Niger."HealthPolicyandPlanning10(3):223-40.
intheLalipurDistrict,Nepal."Thesis.Universityof Wshington,
Seattle.
Eklund,P, and K. Stavem.1990."PrepaidFinancingof Primary
Health Care in Guinea-Bissau:An Assessmentof Eighteen
VillageHealth Posts." PolicyResearchWorkingPaper 488.
WorldBank,WashingtonD.C.
Ensor,T.1995."IntroducingHealthInsuranceinVietnam."Health
PolicyandPlanning10(2): 154-63.
. 1997. "Macro Issues in the Development of Health
Insurance:WorldExperienceand LessonsforTransitionalAsia."
Paperpresentedat a Universityof Antwerpconferenceon the
economicsofhealthinsuranceinlow-and middle-income coun-
tries,January 17-18,Antwerp.
Gilson, L., and others. 1994. "The Potential of Health Sector
Non-GovernmentalOrganizations:PolicyOptions." Health
PolicyandPlanning9(1): 14-24.
Hsiao,W C. L. 1995."The ChineseHealth CareSystem:Lessons
for OtherNations."SocialScienceandMedicine41(8):1047-55.
Hsiao,W C. L., and P D. Sen. 1995."CooperativeFinancingfor
HealthCarein RuralIndia."Paperpresentedat an International
Workshopon Health Insurance in India, sponsored by the
Indian Instituteof Management,September,Bangalore.
System:A Step-WiseApproach to UniversalCoverage."In
David W Dunlop and Jo M. Martins,eds., An International
Assessmentof Health CareFinancing:Lessonsfor Developing
Countries. Washington, D.C.: World Bank Economic
DevelopmentInstitute.
ILO (International LabourOffice).1996a."India:SocialProtection
for the UnorganizedSector."SouthAsiaAdvisoryTeam.New
Delhi,India.
. 1996b.WorldEmploymentReport1996/1997.Geneva.
HealthForum13:171-75.
RuralPoverty.NewYork:NewYorkUniversityPress for the
InternationalFund for AgriculturalDevelopment.

RuRALRISK-SHARING
STRATEGIES
181
Kutzin, Joseph. 1995. "Experience with Organizational and
Financing Reform of the Health Sector." Strengthening Health
Services Paper 8. World Health Organization, Geneva.
La Forgia, G. M. 1990. Health Servicesfor Low-Income Families:
ExtendingCoverageThroughPrepaymentPlansin the Dominican
Republic. Bethesda, Md.: Health Financing and Sustainability
Project.
Litpon, Michael. 1976. Why PoorPeopleStay Poor London: Temple
Smith.
Lukholo A. 1996. Stockvels in South Africa. Johannesburg: Amagi
Books.
McFarlane, G. A. 1996. "Chogoria Hospital Insurance Scheme."
Chogoria Hospital, Kenya.
McPake, Barbara, Kara Hanson, and Anne Mills. 1993.
"Implementing the Bamako Initiative in Africa: A Review and
Five Case Studies." London School of Hygiene and Tropical
Medicine, Department of Public Health and Policy, Health
Policy Unit, London.
Mills, Anne. 1983. "Economic Aspects of Health Insurance." In
K. Lee and Anne Mills, eds., The Economics of Health in
Developing Countries.Oxford: Oxford University Press.
Moens F. 1990. "Design, Implementation and Evaluation of a
Community Financing Scheme for Hospital Care in
Developing Countries: A Pre-Paid Health Plan in the
Bwamanda Health Zone, Zaire." Social Scienceand Medicine
30(12): 1319-27.
Mogedal, S. M. 1984. "Local Health Insurance Supplementing
National Efforts-Nepal." Public Health Review 12: 286-93.
Myers, C.N. 1989. "Thailand's CommunityFinance Experiments:
Experience and Prospects." Harvard Institute for International
Development, Cambridge, Mass.
Noterman,J., and others. 1996. 'A PrepaymentScheme for Hospital
Care in the Masisi District in Zaire: A Critical Evaluation."
Social Science and Medicine 40(7): 919-30.
Parker, D., and R. Knippenberg. 1991. Community Cost-Sharing
and Participation:A Review of the Issues. New York: United
Nations Children's Fund.
Peabody, J.W, S. Lee, and S.R. Bickel. 1995. "Health for All in
the Republic of Korea: One Country's Experience with
Implementing UniversalHealth Care." Health Policy3 1:29-42.
Preker,Alexander,and Richard Feachem. 1995. MarketMechanisms
and the Health Sectorin Centraland EasternEurope.World Bank
Technical Paper 293. Washington, D.C.
Ramirez,T., and others. 1997. "The Market for Maternal and Child
Health Care Servicesin Mexico City and the Role of an NGO."
In Sara Bennett and others, eds. PrivateHealth CareProviders
in Developing Countries:Serving the Public Interest? London:
Zed Press.
Ron, A. 1997. "Community Health Insurance Schemes:Experience
inGuatemalaandthePhilippines."PaperpresentedataUniversity
of Antwerp conference on the economics of health insurance in
low- and middle-income countries,January 17-18, Antwerp.
IN HEALTHCAREFINANCING
INNOVATIONS
182
Ron, A., and A. Kupferman. 1996. Community Health Insurance
Scheme in the Philippines: Extension of a Community-Based
IntegratedProject. Geneva: World Health Organization.
Saltman, R. 1995. "ApplyingPlanned Market Logic to Developing
Countries' Health Systems:An Initial Exploration." Discussion
Paper 4. World Health Organization, Geneva.
Shaw,R. Paul, and Charles C. Griffin. 1995. FinancingHealth Care
in Sub-Saharan AfricathroughUserFeesandInsurance.A Directions
in Development book. Washington, D.C.: World Bank.
Sheng-Lan,T., and others. 1994. FinancingHealth Servicesin China:
Adapting to Economic Reform. Brighton: Institute for
Development Studies.
Somkang, E., P. Akanzinge, T. Apau, and F. Moens. 1994.
"Nkoranza Health Insurance Evaluation Report." Child Health
and Development Centre, Uganda.
Stinson, W 1982. Community Financingof Primary Health Care.
Washington, D.C.: American Public Health Association.
Supachutikul, A. 1996. Situation Analysis of Health Insuranceand
its Future Development. Bangkok: Health Systems Research
Institute.
Supachutikul, A., and P. Sirinirund. 1993. "A Report on
Thai-German Health Card Project."
Suwandono, A., Brahim, and R. Malik. 1995. "The Indonesian
Experiences on Rural Health Financing." Ministry of Health,
Indonesia.
UN (United Nations). 1993. World UrbanizationProspects:The
1992 Revision. New York.
UNICEF (United Nations Children's Fund). 1995. The Bamako
Initiative: Rebuilding Health Systems. New York.
Waddington, C., and K Enyimayew. 1990. "A Price to Pay,Part
2: The Impact of User Fees in the Volta Region of Ghana."
InternationalJournalof Health Planningand Management5(4):
287-312.
WHO (World Health Organization). 1995. International
Consultation on Health InsuranceReform. Seoul, Republic of
Korea. WHO/SHS/NHP/95.9.
Wibulpolprasert, S. 1991. "Community Financing: Thailand's
Experience." Health Policyand Planning 6(4): 354-60.
Willis,C.Y, and C. Leighton. 1995. "Protecting the Poor: The Role
of Means Testing."Health Policyand Planning10(3): 241-56.
World Bank. 1987.FinancingHealthServicesin DevelopingCountries:
AnAgendaforReform. AWorld Bank PolicyStudy.Washington,
D.C.
- 1991. Indonesia:Health Planningand Budgeting.A World
Bank Country Study.Washington, D.C.
. 1996. China: Issues and Options in Health Financing.
Washington, D.C.
WSM (World Solidarity Movement). 1996. Mutulles de Sante'en
Afrique. Senegal.
Yu, S., and G. F. Anderson. 1992. "AchievingUniversal Health
Insurance in Korea: A Model for Other Developing Countries?"
Health Policy 20: 289-99.
Rural Health Care Financing
in kThailand
SirilaksanaKboman
H T ealthcarefinancinghasbecomean issueofglobal
interest and concern in recent years, as both
capitalist and socialist countries grapple with
risingcosts,dwindlingresources,poor-qualitycare, ineffi-
cient resource use, and unequal distribution of services,
Transitioneconomiesin Indochina,Eastern Europe, and
China, as well as capitalisticstrongholdslike the United
States,faceformidablechallengesindeterminingwho should
pay for health care and how it should be managed.
In developingcountries rural areas are of particular
concernbecause regionaldisparitiesin income put rural
populationsat a disadvantagein terms of livingstandards
and accessto health care. Moreover,limited administra-
tivecapacityin ruralareasmakesit harder to managewhat-
ever financingschemeis implemented,especiallyin areas
wherelargeportionsof the populationare engagedin sub-
sistenceactivities,cut off from the formalsector.
This paper reviewsThailand'sexperiencein developing,
implementing, andattimesexperimentingwithvarious health
carefinancingschemes,witha focuson projectsaffectingthe
ruralpopulation.The next sectionsummarizesthe disparity
betweenruralandurbanareaswithrespectto healthoutcomes
and availableresources.The thirdsectionassessesthe range
of financingschemesthat are in place.Schemesthat include
and affectruralpopulationsare discussedin the fourthsec-
tion.Finally,condusionsare presentedin the finalsection.
Rural-Urban Differentials
For the past two decades Thailand has been one of the
world'sfastest-growingeconomies.But althoughabsolute
Sirilaksana
Khomanis deanof the facultyofeconomics
atThammasatUniversity
183
poverty has declined,the distributionof income has not
improved(Krongkaew1995). Moreover,rural and urban
populationshaveverydifferenthealthindicators.Ruralpop-
ulations are afflicted with nearly twice as many poverty-
related diseases-such as infectious diseases-per capita
as urban populations.The Noitheast,the country'spoor-
est region,has the highestincidenceof infectiousdiseases-
53.2 casesper 1,000people,comparedwith 25.5per 1,000
in Bangkok.Urbanresidentshaveahigherincidenceof dis-
ease only for endocrine disturbances (such as diabetes)
and circulatorydiseases(suchas heart disease).
Differencesin illnesspatterns are alsoapparentin hos-
pitaladmissionstatistics(ThailandMinistryof PublicHealth
1995). Amonginpatientsat governmenthospitals,diges-
tive disorders are the largest cause of admissionin the
provinces;in Bangkokthe largestcauseis malignanttumors.
Infant and maternal mortalityrates also showmarked
differencesbetween the different regions and Bangkok.
Infant mortalityin the North, Northeast, and Southis esti-
mated to be about twicethe rate in Bangkok(where it is
18 deaths per 1,000 live births). Similarly,maternal mor-
tality in all regions,particularlythe South, is two to four
timeshigherthan in Bangkok(whereit is 0.1 per 1,000live
births). Since the country's urbanization rate is only 27
percent, with 100 percent urbanizationfor Bangkok, the
regionaldifferencesalsoreflect rural-urbandifferences.
The Northeast also records the country'shighestinci-
dence of first-degreemalnutrition,although the situation
has been improvingoverthe past ten years.About 25 per-
centofthe Northeast'spopulationagedfiveyearsandunder
suffersfrom first-degreemalnutrition, compared with a
in Bangkok.
countrywide average of less than 19 percent (8 percent in
the Central region).
Thais prefer physician services to any other form of treat-
ment; 54 percent of sick people seek treatment at health
outlets staffed by physicians. Again,however, a rural-urban
breakdown reveals that a much higher percentage of the
urban population seeks care from physicians.In urban areas
81 percent of sick people have consultations with physi-
cians, compared with 47 percent in rural areas. Moreover,
rural populations are almost twice as likely to rely on self-
treatment as urban populations (32 percent compared with
17percent). The behavior of urban residents is almost iden-
tical, whether they live in Bangkok or other urban areas
(Khoman 1992).
The Ministry of Public Health is the main provider of
health services, particularly primary care. It has a network
of hospitals at the regional, provincial, and conmnunitylev-
els, and health centers at the subdistrict level. Most of the
mninistry'sservices are in rural areas, as well as in urban
centers besides Bangkok.Almost half of the ministry'sannual
budget is spent in rural areas, and an increasing share is
used for primary health care.
Still,rural-urban differentials in the supply of physicians,
hospitals,and hospital beds remainstriking.In Bangkokthere
is one physician for every 958 people; in the Northeast the
rate is more than ten times higher (table 1). Moreover, pri-
vate households, both rural and urban, remain the largest
source of finance for health services.The share of total health
expenditurescomingfrom households and private companies
increased from 63 percent in 1977 to 72 percent in 1986, to
nearly74 percent in 1992 (Khoman and Mongkolsmai1993).
FinancingSchemes
There are four main types of health care financing schemes
in Thailand: voluntaryhealth insurance, mandatory schemes,
social welfare schemes, and fringe benefit schemes such as
health coverage for government officials and state enter-
prise employees (table 2). The distinction between types of
schemes has not always been clear, however, because
Thailand experiments with ways to provide its population
with secure and accessible health care.
Voluntaryhealth insurance consists of private commercial
insurance,whichcoversabout 0.9 millionpeople almostexclu-
sively in the formal sector, and the Health Card Program,
which was implemented in rural areas in 1983 as a voluntary
scheme. Over the years the Health Card Program has evolved
and can now also be considered a kind of socialwelfare pro-
gram, since it receives an explicit contribution from the gov-
ernment equal to the contribution of the card purchaser.
Coverage is still fairly limited, however, and has fluctuated
between 1.3and 2.7 millionpeople.Thesefluctuationsoccurred
mainlybecause of lapses in policydirection, and the often ad
hoc way in which the program is implemented (see below).
Mandatoryschemesindude the Workmen's Compensation
Fund and the Social Security Scheme. The recently formed
Ministryof Labor and Welfaremanagesboth schemes,which
cover workers in firms with ten or more employees. These
schemes require extensiverecord keeping on employment to
verify eligibility, and thus are confined to formal sector
employees.
The Workmen's Compensation Fund covers job-related
injuries and (theoretically) work-induced illnesses. The lat-

TABLEI
Distributionof medicaland public health resources,1992
(population
perunit)

Region Physician Dentist Pharmacist Nurse Hospital

beda
North 6,316 41,176 24,910 964
Northeast 10,970 78,211 45,020 1,606
Central 5,804 29,181 25,854 815
South 6,079 31,574 21,143 806
Bangkok 958 4,599 2,142 363 260
Regional
average
excluding
Bangkok 7,326 42,811 29,608 1,050 800
WholeKingdomn 4,425 25,530 13,076 885 666
a.Bedsin governmenthospitals for generalservices
only.
Source:ThailandMinistryof PublicHealth1995.

INNOVATIONS
IN HEALTHCAREFINANCING
184
ter, however,are difficultto prove-particularly if adverse
health effects are evident only after manyyears (asis the
case with manganesepoisoning,silicosis,lead poisoning,
and others). Even job-relatedinjuries,particularlyif they
are perceivedas being minor,are underreportedif work-
ers are not assertiveor wellinformed,and if the employer
is concernedwith loss of work time or the hasslesof the
required paperwork (Kultap 1983). All employerswith
ten or more employeesare required to contribute to the
fund.
The Social Security Scheme,implemented under the
Social SecurityAct of 1990,provideshealth insuranceas
part of an overallpackageof benefitsdesignedto provide
securityto populationsnot coveredby other benefit pro-
grams. However,implementationis confined to the most
manageablegroup-employees in formalsectorestablish-
ments. In its firstthree yearsthe schemecoveredemploy-
ees in firmswith twentyor more employees.Coveragehas
sincebeen extendedto firmswith ten or more employees,
increasingthe populationcoveredto 4.5 millionpeople,or
7.6 percent of the populationin 1995.Since this scheme
could be expandedto accommodaterural populations,it
is discussedin greater detailin the next section.
The thirdtypeof schemeis socialwelfareprograms,con-
sistingof severalspecificallytargeted schemes.First, the
Low IncomeSupportProgramprovidesfree medicalcare
to poor rural familiesin governmenthealth facilities.The
cutoffpoint for coverageis 2,000baht per person or 2,800
baht per familyper month. In 1992this programcovered
11.7 million people, or about 20 percent of Thailand's
population.
Freemedicalservicesare alsogivento the elderly.In 1992
about 3.5 millionwere covered,and this number is likely
to increaseconsiderablyin the near future because of an
agingpopulation.In addition,the Ministryof Education
provides free medical care to primaryschoolchildrenin
schoolsunder its jurisdiction.This programbenefits chil-
dren in both urban and rural areas, but providespropor-
tionatelyhigherbenefitsforruralpopulations,whoare less
likelyto have other coverage.More than 5 millionchil-
dren, or about 9 percent of the population, are covered
under this scheme.

TABLE
2
Coverageof healthcarefinancingschemes,1992
Share
of
Population population Subsidy
Scheme Target
population covered (percent) Source
offinance percapita
Voluntaryhealthinsurance
Privateinsurance Mainlyurban 0.9 million 1.6 Insurer
Healthcard Mainlyrural 1.3million 2.3 Card holderand Ministryof PublicHealth 63 baht
Mandatoryschemes
Workmen'sCompensation
Fund Formalsectoremployees 1.8million 3.2 EmployersandMinistryof LaborandWelfare -
Socialsecurity Formalsectoremployees 2.5 million(1992) 4.4 Employers,
employees,and Ministryof 541 baht
4.5 million( 1995) 7.6 LaborandWelare
Socialwelfareschemes,
Low-incomesupport Low-income,mainlyrural 11.7million 20.7 Ministryof Intenor 214 baht
Supportfor the elderly Population
over 60 3.5 million 6.2 Ministryof PublicHealth 72 baht
Schoolchildren Primaryschoolchildren 5.1 million 9.0 Ministryof Educaton
Fringebenefitschemes
Govemmentreimbursement Govemmentofficialsand 5.6 million 9.9 Govemment(variousagencies) 916 baht
employeesandtheir families
Stateenterprisebenefits Stateenterpriseemployees 0.8 million 1.4 815 baht
andtheir families
Insuredpopulabon 33.2 million 58.7
Uninsuredpopulation 23.3 million 41.3
a. Otherwefarerecipients
include
veterans,
monks,andthosedeemed needy.
Source:
ThailandMinistry
of Public
Heaith1992;Thailand
Ministry
of LaborandWefaredata;calculated
fromHsiao1994andMongkolsmai
1993.

RURALHEALTHCAREFINANCINGIN THAILAND
185
 Other welfarerecipientsinclude veterans and monks.
Peoplewho are consideredneedy (asdeterminedby social
workersin governmenthospitals)are also routinelygiven
free care.In addition,the Ministryof Interior offers free-
care cardsto low-incomeurban families.
Finally,health coverageis offeredas a fringebenefit in
large private companies,government agencies,and state
enterprises.About5.6milliongovernmentofficialsand their
familiesare covered,and nearly1.0millionstate enterprise
employeesand dependents. Combined,these two groups
make up about 11 percent of the population.
Governmentofficialstendto be inmiddle-to high-income
groups.In fact, theyreceivethe largest subsidyfrom gov-
emnmentexpenditureson health (seetable 2). Suchcover-
age has little effect on rural populations unless they are
sufficientlyeducated to qualifyfor positionsin govern-
ment agenciesor state enterprises.
Altogether,about 59 percent of Thailand'spopulation
is protected by some kind of health care coverage-and
41 percent, or 23 millionpeople, is not covered by any
scheme.This group includessubsistencefarmers,the self-
employed,rural workers,and urban dwellersengagedin
informalsector activitysuch as street vending and small-
scalecommercialundertakings.Thoselivingon the fringes
of society(slum dwellers,homelessurban migrants)typi-
callyalsohaveto fend for themselves.A surveyofBangkok
transientswhodwellin makeshiftaccommodationsunder
bridges and at railwaystations found that none was cov-
ered by a health scheme(Khoman1995).Yettheysay they
are fit and well.
The Ministryof PublicHealth is tryingto expandinsur-
ance coverageto slum dwellers,the self-employed,and
highlymobilegroupssuchas constructionworkers,service
workers,and prostitutes. Somestudies,however,indicate
that constructionworkersand prostituteshavelittle inter-
est in purchasing any form of health care coverage
(Mongkolsmaiand others 1994). Prostitutes alreadyvisit
clinicsthat treat sexuallytransmitteddiseases,a servicepro-
vided by most governmentprovincialhospitalas the need
arises. Construction workers also showed little interest,
partlydueto the complicatednature of healthcoverageand
the lowperceivedneed. Current researchon slumdwellers
andtransientsshouldprovidefurtherinsightintothisgroup's
willingnessto pay for health coverage(Rojvanit1995).
INNOVATIONS
IN HEALTHCAREFINANCING
186
Health SchemesAffecting Rural
Populations
Low-income
support
Governmentpolicyon providingfree medicalcare to low-
incomegroupswas initiatedin 1975.The goalof this pol-
icywasto reducethe prevailinginequityin accessto health
services.Knownas the Low Income Support Program,it
offers free medicalcare at government hospitalsto low-
income groups and has become the main health scheme
for rural populations.
Coverageinitiallywas limited to people with monthly
incomesbelow 1,000baht. Upon implementation,the tar-
get populationwas set at 7-8 millionpeople. Until 1980,
however,no identificationcardswereissued,and free care
wasgiven at the discretionof health facilities'staff. Since
then low-incomecards have been issued to eligibleciti-
zens,now definedas familieswith monthlyincomesbelow
2,800 baht and individualswith monthlyincomes below
2,000baht. The cards entitle holdersto free medicalcare
at allgovernmenthealthfacilitiesoperatedby the Ministry
of PublicHealth,the BangkokMetropolitanAdministration,
the Red Cross Society,Pattaya City,and municipalities.
Cards are validfor three years.
The governmentprovidesblock grants to health facili-
ties based on the expecteddistributionof the eligiblepop-
ulationand pastrecordsof serviceto patients. Specifically,
the budgetaryallocationis based on the number of low-
incomepeoplelivingin less-developedvillages,as defined
bythe NationalEconomicand SocialDevelopmentBoard.
In addition,the number of users of health care facilitiesis
taken into account,and the number of veterans and their
families.
The free medicalcare budget for low-incomegroups
increasedfrom 521 millionbaht in 1982 to 1,911million
bahtin 1993.Asashareofthe healthbudget,the low-income
budgethas ranged from7.7 percent in 1980to 7.9percent
in 1993,down from 12.5percent in 1976.In real terms the
budget per patient dropped from 155 baht in 1976 to 45
baht in 1980,stabilizingat around 50 baht over the past
ten years.The budget allocatedper card holder has been
rising,however,increasingfrom68baht in 1984to 163baht
in 1992.
 Still,the budgetary allocationis invariablyinsufficient
to cover the cost of providing services.Satsanguanand
Leopairote (1992) report a unit cost of 85 baht for an
outpatient visit and 1,200baht per case or 360 baht per
day for inpatient care at district hospitalsin 1991. As a
result hospitalsroutinelycross-subsidizelow-incomecov-
erage usingother sourcesof revenue, such as reimburse-
mentsforgovernmentofficialsandhospitalfees.The extent
of the subsidy varies by hospital. For example, Nan
ProvincialHospital in the North reported that between
October 1992 and March 1993 the costs of providing
care for low-incomepatients totaled 1.4 millionbaht a
month, while the annual budgetary allocation was 10.5
millionbaht, or 0.87 millionbaht a month.Thus the cross-
subsidizationwasabout 6 millionbaht a year(Mongkolsmai
1993). In the province of Samutsakornthe budget allo-
cationwas2.3 millionbaht whileactualexpenditureswere
5.8 millionbaht.
The schemedoes not requirecost sharingon the part of
the eligiblepopulationand has been subjectto much crit-
icismsince its inception,particularlywith respect to the
distributionof the card. A 1980study by the Ministryof
Public Health's RuralHealth Divisionfound that 12 per-
centof the supposedlylow-income beneficiariesusingprovin-
cialanddistricthospitalsand9 percentvisitinghealthcenters
hadmonthlyincomesabove2,000baht.Anotherstudyfound
that about20percentof cardholderswerenotpoor (Thailand
Ministryof PublicHealth and Facultyof Health Services,
Mahidol University 1988). Mongkolsmaiand Khoman
(1993)alsofound that somenonpoorfamilies,possiblywith
connectionsto officials,possessthe card.
The cutoff level used for card eligibilityis almostfour
times the poverty line, defined as the minimumincome
required for subsistence.Yetin 1988/89,when about 29
percent of the ruralpopulationwas belowthe povertyline,
the LowIncome SupportProgramcoveredonly 7.65 mil-
lion people, or just under 20 percent of the rural popula-
tion. The card coveredjust 28 percent of the low-income
group as definedby the income cutoff level, and 49 per-
cent of the poor as defined by the poverty line. In 1990
coverageimprovedas a result of expandedeffortsto reach
targetedgroupsand increasedscreeningof cardrecipients.
Nevertheless,coverageremainslow,with up to 20 percent
of thosebelowsubsistencestillleft out (Mongkolsmai 1993).
Evenwithlimitedcoverageofthe poor,however,Thailand
still has a safety net that protects the needy,since many
low-incomepeopleroutinelyreceivefreemedicalcareeven
without a card.In 1987,13.7 millionpeople receivedfree
medicalcare,but of these only7.6millionwerelow-income
cardholders.The proportionis believedto be roughlythe
same today,so that people seekinghealth care can access
health facilitieswithout incomebeing a constraint.
Moreover,the LowIncomeSupportProgramhas prob-
ablyimproved accessto health servicesamongthe poor.
The Northeast has the highestshare of low-incomecard
coverage,coveringabout30 percentof the population(table
3).Andcoverageof the poor (thoselivingbelowthe poverty
line)increasedin allregionsbetween 1987and 1990,with
the largestincreasein the Northeast.
The problem that remains is the difficultyof properly
identifyingeligibility.
Giventhat much of the ruralpopula-
tionis engagedin agriculturalactivity,with a substantialpro-

TABLE3
Coverageunderthe LowIncomeSupportProgram,1987and 1990
1987 1990
Population Shareof Shareof Population Shareof Shareof
covered population poor' covered population poor'
Region (millions) (percent) (percent) (millions) (percent) (percent)
North 1.8 18.8 73.7 2.4 23.4 100.8
Northeast 3.5 20.4 42.3 5.3 30.4 81.1
Central 1.3 11.3 72.6 1.5 12.1 75.7
South 1.0 12.4 45.6 1.5 18.2 84.7
Bangkok 0.0 0.5 14.3 0.0 0.3 8.7
Whole Kingdom 7.6 14.5 49.2 10.7 19.2 81.0
a.Determinedusingpovertylinefor 1988/89.
Source:
ThailandMinistryof PublicHealth1995.

RuRAL HEALTIHCAREFINANCINGIN THAILAND

187
portion of income received in kind, assessing and imput-
ing income is difficult. Defining a household is also prob-
lematic in cases where family members work in cities and
remit earnings.
Health Card Program
The Health Card Program was started in 1983 to promote
maternal and child health. It was an innovative program
because it involved selling health cards to villagers who
had been accustomed to receiving free care. Card buyers
prepaid a fixed annual premium in return for free services.
Proceeds from card sales went into a health card fund that
was managed by a village committee. Thus the program
familiarized rural populations with concepts of preventive
behavior, insurance, risk pooling, fund management, and
community self-help, though the initial emphasis was on
improving health.
T'he program was initiated in eighteen villages in seven
provinces. At its inception the program's primary objective
was to improve health among rural populations, with an
emphasison primary care-including health education, envi-
ronmental health (sanitation and water supply), maternal
and child health, family planning, nutrition, immunization,
prevention and control of diseases, treatment of common
ailments, and provision of essential drugs. In addition, the
program incorporated a referral system that required card
holders to visit primary care centers instead of tertiary care
institutions, so that care-seeking behavior would be more
rational and efficient in terms of utilization of health
resources. The program was also intended to involve local
villagers in self-help as well as in managing the health card
fund. Self-help and communal participation were also fos-
tered by recruitingvillagehealth volunteers and villagehealth
communicators who worked with health personnel at the
village health center.
The features of the Health Card Program were constantly
adjusted, and policymakerssometimes vacillated over time.
At issue were the number of free episodes allowed, the
coverage ceiling per visit, the number of family members
included, the types of diseasescovered, the price of the card,
and renewability.In 1983 there were three types of cards:
a familycard priced at 200 baht, a maternal and child card
priced at 100 baht, and an individual card priced at 100
INNOVATIONS IN HEALTH CARE FINANCING
188
baht. Each cardwasvalid forayear. The familycard allowed
eight illness episodes a year for four family members; the
individual card allowed four illness episodes. To prevent
against moral hazard behavior, ceilings were imposed.
Coverage was limited to 1,000 baht for accidents, with a
10 percent discount for amounts over 1,000 baht. Chronic
diseases, cancer, and "self-inflicted"diseaseswere excluded.
Card holders who had not used the card in one year were
entitled to renew it without charge for an additional year,
not exceeding two renewals.
In mid-1984 the ceilingwas raised to 2,000 baht per visit.
A strict referral system was also enforced whereby initial
contact had to be with the village drug fund, and subse-
quently the health center in the subdistrict, the community
hospital at the district level,and finallythe general or regional
hospital at the provincial level. This approach was intended
to increase efficiency in the use of different levels of health
services, since over-utilization of high-level hospital care
was occurringbecause villagersroutinelybypassed the health
centers in favor of physician services.
From the outset, the health card fund was designed to
be a village-levelfund in order to foster grassroots partici-
pation and management skills. Of this fund, 15 percent
was allocated to the health center, 30 percent to the com-
munity hospital, and 30 percent to the regional or provin-
cialhospital, reflectingthe pattern of utilization in the referral
system and the severity of illness and cost to the provider.
Of the remaining 25 percent, 10 percent went to person-
nel in the provider institutions and 15 percent was used as
operating expenses of the fund.
The Health Card Program later gave more attention to
medical treatment, and the price of the family card was
increased to 300 baht.
During 1985-90 some health card funds were initiated
at the district and subdistrict levels: a 300 baht family card,
a 100 baht maternal and child card, and a 200 baht indi-
vidual card. In addition, the number of episodes allowed
was cut from eight to six.
Confusion arose about the different types of cards, the
terms and conditions of use, the losses incurred due to the
inabilityof hospitalsto recovercosts from the health card fund
contribution, and the problems with the strict referral sys-
tem, which tended to disregard geographical proximity
Tantiserani(1988) found that the Health Card Program was
active in about 33 percent of the villagesand 70 percent of
thesubdistricts in 72provinces,
withabout2.7millionpeo-
pleholdingabout550,000familycards.A 1992surveybythe
Ministry ofPublicHealthfoundthatcoverage hadextended
to 36percentofthevillages, but thepopulation coveredwas
reducedto 30 percentof the subdistrictsin 68provinces,
with1.3milionpersonsholding260,000familycards.
Thisreductionin coveragehas beenattributedto the
ministry'slack of policydirectionduring this period
(Kiranandana 1990).Theprogramwasimplemented in an
ad hoc mannerand thrivedonlyin provincesthat actively
encouragedvillagerparticipation. Moreover, the program
wasneverlucrativeforlargehospitals,andhad to be sub-
sidizedby othersourcesof hospitalrevenue.
In 1990the health card schemewas modifiedand
renamedthe NewHealthCardApproach,whichempha-
sizedtheconceptof risksharing.Pricerestructuring (rang-
ingfrom200bahtfortheindividual card,andfrom300-500
bahtforthefamilycard,coveringup to threegenerations)
occurredin fiveprovinces(Manopimoke1995).In addi-
tion, the referralsystemwasrelaxedand coveragewas
expanded,withunlimited numbersoffreevisitsandno cov-
erageceilingpervisit.About20percentofthepopulation
in theselectedareasparticipatedin theproject.
In 1991the differenttypesof cardswerediscontinued,
withonlyfamilycards(pricedat 500baht)offered.Since
1994the Ministryof PublicHealthhasprovidedan equal
contributionof 500baht per card.In addition,no limits
areimposedon thenumberof episodesor the costcover-
ageper visit.Moreover, moreflexibilitywasbuiltintothe
referralsystem,and eachprovincecouldimposewhatever
conditionsit deemedappropriate.Someprovinces,such
as Rachaburi, allowcardpurchasers to payin installments
(Mongkolsmai and others 1994).Administrative changes
werealsoimplemented,withthe healthcardfundman-
agedby a committeeat the districtlevelin coordination
withvillage-level bodies.This approachwasintendedto
expandthe enrollmentbasebeyondthevillageto the dis-
trictlevel.In addition,80percentofthecardpricenowgoes
to providersof medicalcare,whilethe remaining20per-
centisretainedfor marketingand salesincentives.
Thehealthcardis likelyto be usedto expandcoverage
forpeoplewhocurrentlylackinsurance.Studiesarebeing
undertakento ascertainitsfeasibilityin urbanareasaswell.
RURAL HEALTH CARE FINANCING IN THAILAND
189
Social
security
TheSocialSecurityActof 1990provideshealthinsurance
as part of an overallpackageof benefitscoveringillness
unrelatedto work,maternalbenefits,disabilityunrelated
towork,death,childbenefits,oldage,andunemployment.
Companieswithten or moreemployeesand employees
themselves areeachrequiredto contribute1.5percentof
the employee's wagesto theSocialSecurityFund,withan
equal1.5percentprovidedby thegovernment. Sincethe
contribution is basedonincomeandnotthe expectedrisk
or incidenceof illness,risksare pooledand benefitsare
skewedin favorof high-riskindividuals.
Of thiscontribution,2;45percentagepointsis usedto
providemedicalcarefortheinsuredforillnessandmater-
nityand 2.05percentagepointsareusedfor disabilityand
deathbenefits.Expandingbenefitsto includechildand
the elderlywouldrequirean increasein the contribution,
and so hasbeenpostponed.Unemployment benefitswill
be implementedat a muchlater stagefor fearof moral
hazardbehavior-thatis, providingunemployment bene-
fitscouldinduceunemployment.
Asnoted,socialsecurityis mandatedforthemostman-
ageablegroup,namelyemployeesin formalsectorestab-
lishments withten or moreemployees. Theschemecovers
about4.5millionworkers,or 7.6percentofthepopulation.
TheSocialSecurityAct alsoprovidesfor expansionof
theschemeona voluntary basistoincludetheself-employed,
suchasfarmers,own-account workers,andotheruninsured
groups.However,thisprovision wasprobablyincludedfor
politicalexpediencyto protect againstchargesthat the
programonlycoversthewelloff.Evenif theself-employed
wereallowedto participate,the problemremainsof how
to dealwiththeircontribution. Thustheprogramhasbeen
proposedas a practicalform of coveragefor the self-
employed, butThailandisstillexperimentingwiththisidea.
Amoreimmediateconcernisthecurrentsystemofpro-
vidingmedicalcare.Insuredworkersarerequiredto regis-
ter at a specifiedhospital,calledthe "maincontractor,"
wheretheyreceivefree medicalcare (exceptfor certain
typesof treatment,suchas cosmeticsurgeryandoptome-
aresimilartothoseundertheHealth
try).Theserestrictions
CardProgram.Themaincontractorreceives an annualpre-
paymentor capitationfeefromthe SocialSecurityFund,
initiallyequal to 700 baht per insured person registered,
regardlessof actualutilization.Capitationwas chosenover
a fee-for-service
paymentscheme(toproviders)forits admin-
istrativesimplicityand to preventthe cost escalationthat
invariably occurs with fee-for-service payments. The
Workmen'sCompensationFund, for example,payshospi-
tals on a fee-for-service
basis, and has incurred substantial
costescalation,aswellas the administrativeburden of deal-
ing with massiveclaimsdocuments.Capitation,however,
maynot be attractiveenough to induce medicalproviders
to enrollin the scheme.The relativemerits of alternative
paymentmethods are sunmmarized in table 4.
The main contractor is able to contract with subcon-
tractors,whichprovide lowerlevelsof care,as wellas with
supracontractors,whichprovidehigherlevelsof care.Both
publicand privatehospitalsparticipatein the scheme.One
of the main problemswith.the scheme is confusionwith
respect to insured persons receivingcare in hospitals in
whichthey are not registered.Accidentvictims,for exam-
ple, madeheadlineswhentheywere allegedlydeniedtreat-
ment by noncontractinghospitals.
At firstemployerschosethe maincontractor,whichmay
explainlowutilizationin the firstyearof the scheme'simple-
mentation.In 1991the utilizationrate for outpatientcare
was0.32visitsper insuredpersonper year,and in 1992the
rate dropped to 0.22 visits.In calculatingthe capitation
fee,itwasestimatedthat therewouldbe 3.0visitsper insured
person per year.Inpatient utilizationwas also lowerthan
expected;in 1991and 1992the utilizationrate was 0.016
and 0.012visitsper insuredpersonper year.The estimated
utilizationrate was 0.05 visits.
Apart fromthe inconvenienceofworkersreceivingcare
at a hospitaltheywere not able to select,and the limited
number of participatinghospitalsto choose from, several
other problemsemerged.Many workerswere ignorant of
their rights,somewere not awareof the contributionsthey
made to the Social Security Fund because of automatic
deductions from their wages, and many did not under-
standthe proceduresthat had to be followedto obtainmed-
icalcare (Patichon1995).Further,providersof medicalcare
werenot preparedto managethe schemein terms of health
care delivery.As a resulteffortshavebeen madeto improve
the deliverysystemand the quality of care so that insured
persons have better accessto the care to which they are
entitled.
Specifically,the Social SecurityOffice issued a policy
directive in 1992 to grant insured persons the right to
choose their own hospital. This policywas implemented

TABLE
4
Strengthsandweaknesses
of alternativemethodsof payinghealthproviders
Paymentmethod Strengths Weaknesses
Feefor service Provider'srewardcloselylinkedto levelof effort Tendsto causecost inflation
andoutput Requiresprocessing andverificationof massivenumbers
Allowsfor easyanalysis
of provider'spractice of claimsdocuments
Createsincentivefor excessiveandunnecessary treatment

Percase(for example,using Provider'srewardfairlywell tiedto output Technical

difficultyof forcingallcasesinto standardlist
diagnostic-related
groups) Givesproviderincentiveto minimizeresource canleadto mismatchbetweenoutputand reward
Providersmaymisrepresent diagnosis in order to receive
useper individualtreated higherpayment

Capitation(per patient Administratively

simple;no needto breakdown Givesproviderincentiveto selectpatientbasedon risk
undercontinuouscare) physician'swork into proceduresor cases andrejecthigh-costpatients
Facilitates
prospectivebudgeting Maycreateincentives for providerto underservice
Givesproviderincentiveto minimizecostof treatment acceptedpatients
Allowsfor consumercloutif patientcanselectown provider Difficultto analyzeprovider'spractice

Salary(straightpayment Administratively
simplest Lossof patientinfluenceover providerbehaviorunless
per periodof work) patientchoicelinksprovidersalaryto patientsabisfaction
Facilitates
prospectivebudgeting Can easilycreateincentives
for providerto underservice
patientandto reduceproductivity
Source:
Wodd Bank1993.

INNOVATIONS
IN HEALTHCAREFINANCING
190
in one province in 1992 and expanded to nineteen and There is also evidence of abuse. That is, the medical
forty-fourprovincesin the followingtwo years.The pol- providers that some networks have recruited are so far
icyis being implementedin sixtyprovincesat present. As apart-in somecases,providersare in provincesseparated
a result the portion of insured persons choosing their by severalhundredkilometers-that insured personshave
ownhospitalincreased from 2 percent in 1992to 37 per- fewopportunitiesto use them.Providerscan claimthe addi-
cent in 1995. tional paymentthat is calculatedon the basis of services
The SocialSecurityOfficeis alsoencouragingthe for- provided,but these servicesare providedto differentpeo-
mationof providernetworksto increaseefficiencyin health ple,andthe insuredpersondoesnot benefitfromthe appar-
care delivery,improveaccessibilityof services(particularly ently wide range of providers available.These problems
ifnetworkmembersare geographically dispersed),andpool are being studied and further adjustments made to the
risks. Moreover,additionalpaymentis givento the main system.
contractorsin proportionto the inpatient and outpatient
care that they provide. Thus medical providers have an Conclusion
incentiveto increasetheirmarket shareand engagein mar-
keting for contractswith workers.As a result the number Thebasicissuewithrespectto financinghealthcareinvolves
of main contractorsincreased from 137 in 1991to 1,879 determininghowmuch of a country'sresourcesshouldbe
in 1995.Amongprivate providers,the number of network devoted to health services,how much shouldbe spent by
membersincreasedfrom 69 in 1992to 620 in 1995.In the the govemment,and howmuch relianceshould be placed
public sector the increasehas been more moderate:from on private sources of finance. The main questions are:
671 in 1991to 1,257providersin 1995.The switchfrom Who should payfor the cost of providingservices(recipi-
publicto privatehospitalshas alsobeenincreasing.In 1991 ents throughuser charges,governmentthrough subsidies,
only 16percent of insured personschoseprivatehospitals, or other funding sourcessuch as private business,collec-
but by 1994that share was59 percent. tive bodies, and charitable organizations)?How should
Twopattems of networkformationare used. The firstis health care be organized?And what role should financial
a directcontractnetwork,in whichthe maincontractorcon- intermediariesplay?
tractsdirectlywith networkmembersand assumesrespon- In Thailandvariousschemesare used to addressthese
sibilityfor managingthe capitationfee receivedfrom the issues.Someoverlap,coveringthe same population,while
SocialSecurityOfficeand acting as a "secondarymedical other populationgroupsareleft unprotected.There arealso
careprovider."The secondis an indirectcontractnetwork, differencesbetweenthe benefitsthat canbe obtainedunder
in whichthe maincontractortransfersthe capitationfee to each scheme. Some schemes, such as the fringe benefit
anetworkofficeresponsibleformanagingthe fundsreceived schemethat coversgovernmentofficialsandthe Workmen's
from severalmain contractors.The networkoffice is also CompensationFund, pay for health care on a fee-for-ser-
responsibleformarketingandrecruitingnetworkmembers. vicebasis.Others,likethe socialsecurityscheme,use a cap-
In the second case the main contractor is responsiblefor itationmethodofpayment.Someschemes,suchasthe social
medicalservicesonly. securityschemeand the Workmen'sCompensationFund,
In practice,however,there is considerableconfusionin are compulsory.The Health Card Program, on the other
the system,with great variation between networks with hand,is voluntary,and raisequestionsof adverseselection.
respect to network coverageof the three levels of care Someschemesrequirecopaymentsfrombeneficiaries, while
(subcontractor,main contractor,supracontractor),ability othersrestrict the types of servicescovered.
to managefunds and the paymentmechanismto encour- Someschemes,suchasthe LowIncomeSupportProgram
age costcontainmentand fosterfinancialfeasibilityfor the and coveragefor veteransand the elderly,requireno direct
network, quality and standard of medical servicespro- contributionfrombeneficiaries.Others, such as the health
videdbythe networks,andcoordinationbetweennetworks card and social security,require explicit contributions.
and networkmanagementat the nationallevel. Governmentsubsidiesexistformostschemes,somein terms

RURAL HEALTH CARE FINANCING IN THAILAND

191
 of explicitbudgetaryallocation(such as for veterans, the conduciveto risk pooling,which requiresa largenumber
elderly,low-incomegroups,socialsecurity,and the health of members.Thus the second preconditionis government
card).Implicitgovernmentsubsidiesalsopervadethe pub- initiativeand partial subsidy.This is the course that the
lic healthsystemthrough the pricingof servicesat govern- Health Card Programhas taken, and even though it has
ment hospitals,givingdifferentbenefitsto differenttypes beenbeset withproblems,it is neverthelessa viablescheme
of users (Khoman1995). to start off with in a rural community
Althoughsome schemeswere designedspecificallyfor
the ruralpopulation,the distinctionbetweenthe formaland References
informalsectorsis evenclearer.In fact,many groupsin the
informalsector are excludedfrom existingschemes. Hsiao,WilliamC. 1994. "HealthCareFinancingin Thailand:
Thereislittlecoordination betweenschemes,andThailand Challenges forthe Future."Paperpresentedat a facultysem-
wiT
haveto figureouthoration
esetween cscrepanies,indfilanc inarattheCollegeofPublicHealth,Chulalongkom University,
willhaveto figureout howto easethe discrepancies in financ- Bangkok.
ing.Differentschemescontainvaryingelementsof subsidy, Khoman,Sirilaksana. 1992.'HouseholdChoiceof Health-Care
andcouldcausethe allocationofresourcestoworsenbetween Providerin Thailand."ResearchReport92-02.International
urban and ruralareas.Allthese aspectsmustbe considered HealthPolicyProgram, Washington, D.C.
as Thailandcontinuesto make adjustmentsto increaseeffi- . 1995."ThePricingof Govemment HealthServicesin
ciencyciencyand
equityin health care financing.
lessons m beadrawnfromg.
Thailand:EquityVersusEfficiency?"
Regional Conference
Paperpreparedforthe
on HealthSectorReformin Asia,held
Some lessons can be drawn from Thailand's experi- at theAsianDevelopment Bank,May22-25,Manila.
ence,however,particularlyin terms of the typesof financ- Khoman, Sirilaksana,
andDowMongkolsmai. 1993."Public-Sector
ingschemesthat areviablein ruralor informalsectorsettings Health Financingin Thailand:A Synthesisof Findings."
and the preconditionsthat are required for such schemes ResearchReport93-04.International HealthPolicyProgram,
to succeed. Among other things, given the problems of Washington, D.C.
semi-subsistence
assesinidby Kiranandana,'Thienchay1990.EvaluationoftheHealthCardProject.
assessingincomesin areas dormnatedby seml-subslsten(In Thai.)MinistryofPublicHealth,Bangkok.
or informalsector employmentand the need to avoidcre- Krongkaew, Medhi.1995. "ThaiSocietyandthe Distribution of
atingexcessivefiscalburden, a workablesystemcould be Opportunity andIncome."Paperpresentedat theannualsym-
based on some form of communityfinancing,such as the posiumorganized bythefacultyof economics,March30-3 1,
Health Card Program. But certainpreconditionsmust be Thammasat University,
Bangkok.
in place for sucha schemeto get off the ground. Kultap,Praneet.1983."AStudyofIndustrial Injury:ACaseStudy
The most importantof these preconditionsis sufficient of the TextileIndustry"M.A.thesis.ThammasatUniversity,
facultyof economics, Bangkok.
s^,ocial
capital.The conceptof socialcapitalstemsfrom the Manopimoke, Supachit.1995."Voluntary HealthInsurancein
recognitionthat socialactionsandthe developmentof social Thailand:Development andAchievement." Paperpresented
organizations-such as a community health scheme- at the RegionalConference onHealthSectorReformin Asia,
depend on expected mutual benefitsthat are affectedby heldat the AsianDevelopment Bank,May22-25, Manila.
social norms and obligations within a community Social Mongkoismai,
Dow.1993. "The Social Welfare for Health Care."
capitalis accumulatedthrough the coordinationand coop- PaperpresentedattheNational
Workshop onHealthFinancing
erapitaionsof
aumuter wthinrsocialstheucturessuhandfp- in Thailand,November12-13,Phetburi.
eration of members withinsocial structuressuch as fami- Mongkolsmai,
Dow,and Sirilaksana Khoman.1993. "Studyof
lies, associations,clubs, and communities.The closer a Beneficiaries
ofPublicHospitals
in Thailand."
ResearchReport
communityis, the strongerare the norms that require com- 93-02.International
HealthPolicyProgram.[city?]
pliance;and the more stable a communityis, the stronger Mongkolsmai,Dow,PlearnpitSatsanguan,AnongRojvanit,
is the likelihoodof socialcapitalaccumulation.Moreover, Sirilaksana
Khoman, PraphatsornLeopairote,andThanakorn
socialstructuresthat are well-endowedwith socialcapital Ungswad. 1994."Features
ofVoluntaryHealth
Insurance:
Case
are moretamenable to coarective
well-en .owet
thesmaal num Studyof the Health Card ProjectforSelectedPopulation
are more amenableto collectiveaction.Yetthe smallnum- GroupsinRachaburi Province."
Researchreportsubmittedto
ber of membersrequired for communitytrust and cohe- the ThailandMinistryof PublicHealth,Officeof Health
siveness-the building blocks of social capital-is less In.urance,Bangkok.

INNOVATIONS IN HEALTEI CARE FINANCING

192
Patichon, Preeya. 1995."Choiceof Hospital and Utilizationof
Medical ServicesAmongInsured PersonsUnder the Social
SecuritySchemein NonthaburiProvince."(In Thai.)Master
of Economicsthesis. ThammasatUniversity,facultyof eco-
nornics,Bangkok.
PopulationStudiesInstitute. 1988. "ThailandDemographicand
Health Survey' ChulalongkomUniversity,Bangkok.
Rojvanit,Anong.1995."TheEpidemiologyof Inequity:
Preliminary
Findings."Paperpresentedat the IntemationalHealth Policy
ProgramAnnualMeeting,October 2-7, Beijing.
Satsanguan,Pleampit, and PraphatsomLeopairote.1992."AStudy
of UnitCostsinPublicHospitalsinThailand."ResearchReport
92-02.IntemationalHealthPolicyProgram,Washington,D.C.
Tantiserani,Pichai. 1988."Health Card in the RuralAreas."(In
Thai.)Paperpresented at the Workshopon Health Insurance
Systemfor Thailand,August18-19,Pattaya.
RURAL HEALTH CARE FINANCING IN THAILAND
193
ThailandMinistryof Public Health. 1980.RuralHealthDivision
SurveyReport.(In Thai.)Bangkok.
. 1992.Evaluative
ReportonSocialSecurity.(In Thai.)Office
of the Permnanent Secretary,Bangkok.
- . 1995.PublicHealthStatistics1993.Bangkok.
ThailandMinistryofPublicHealthand FacultyofHealth Sciences,
MahidolUniversity. 1988.AStudyoftheCoverage ofFreeMedical
CareServicesfortheLow-IncomePeoplein Thailand.(In Thai.)
Bangkok.
ThailandNational StatisticalOffice. 1990.Surveyof Population
Change1987-88.Bangkok.
ThailandOffice of the PrimeMinister.1995.GovernmentBudget
1995.Bangkok.
WorldBank. 1993. WorldDevelopmentReport1993:Investingin
Health.NewYork:Oxford UniversityPress.
Market-Based Reform of U.S. Health
Care Financing and Delivery: Managed
Care and Managed Competition
Alain C. Enthoven
T he U.S. health care financingand deliverysystem
is undergoingaprofoundand rapid transformation
from a model characterizedby fee-for-servicepay-
ments, indemnity insurance coverageof completelyfree
choice among providers, and a fragmented and nonac-
countable deliverysystem,to a model made up of com-
peting,integrated,accountablecomprehensive caresystems
genericallyknown as managedcare.This transformationis
not being driven by publicpolicy,whichfor the most part
has been protectiveof the status quo. Rather,it is being
drivenbyconsumerandemployerchoiceand marketforces
in the privatesector.
Indeed, the large governmenthealth care programs-
Medicare(thefederalgovernment'sprogramforthe elderly
and the disabled),Medicaid(the federaland state govern-
ments' programsfor some of the poor), and the Veterans
AdministrationHealth System-have barely started the
transformationand willbe the last to completeit. (In July
1994just7 percentof Medicarebeneficiariesand 8 percent
of Medicaidbeneficiarieswere in healthmaintenanceorga-
nizations;InterStudy 1995;HealthCareFinancingReview
1996)The health maintenanceorganization(HMO) rep-
resentsthe more advancedtype of managedcare. (HMOs
are definedbelow.)In 1980about 9 millionAmericanswere
enrolled in HMOs. At the end of 1996 about 65 milion
were,and the numberis growingby about 12percenta year.
Preferredproviderinsurance,a minimalform of managed
care,almostdid not existin 1980.Today,it coversabout 90
miflionAmericans.Thetraditionalmodelofindemnityinsur-
ance,characterizedby alackof anycontractualrelationship
AlainC. Enthovenis the MarrinerS. EcclesProfessorof PublicandPrivateManagementin the GraduateSchoolof Businessof
StanfordUniversity.
195
betweeninsurerand provider,is quicklydisappearing.This
paper describesthe traditionalsystemas a point of depar-
ture and as a point of contrast to the new system.It then
describesthe emergingparadigmofmanagedcareand man-
agedcompetition.
The Traditional System and Its
Consequences
The traditionalU.S. health care systemwasthe product of
interactions among different groups pursuing their own
interests without any overarchingpublic policy to guide
them or generalagreementon the goalsof the system.As
a result the systemultimatelybecame an exampleof pro-
found and multifacetedmarket failure.
Naturalmarketfailure
The market failurebegan with what mightbe called nat-
ural or inherent market failure. As Arrow (1963) noted,
the incidenceof illnessand the efficacyof treatment are
veryuncertain,creatinga natural desireamongrisk-averse
people to insure against largeand unpredictablemedical
expenses.But with insurancecomes moral hazard-that
is, insuredpeople becomeunconsciousof the costs of the
care that they receive.Over time the prices and standards
of U.S.healthcareadaptedto a marketof cost-unconscious
patients.Informationon medicalconditionsandtheir treat-
mentis costly,and doctorsand their patientshaveverydif-
ferent amountsof it. Patients depend on their doctors for
adviceand for treatments, and doctors,havinga financial
interestin treatment decisions,are imperfectagents.
Complicatingmatters, different people have very dif-
ferentmedicalneeds.In avoluntarymarketforhealthinsur-
ance, the healthydo not want to be pooledwith the sick.
And pooling arrangements can be exploitedby insurers
selectivelyofferingproductsthat are more attractiveto the
healthiermembersof the pool, effectivelyisolatingthe sick
from affordablecoverage.Finally,in the U.S. mixed but
partlyfreemarket for healthinsurance,there is a safetynet
in the form of countyhospitals,lawsrequiringhospitalsto
evaluateand stabilizeuninsuredpatientswhoappearattheir
doors, and a great deal of free care givenby doctors and
hospitals.Moreover,healthcarecoveragehas becomecostly
relativeto the incomes of many low-incomepeople. In a
voluntarysystemof healthinsurancethese problemsencour-
age "free riders."As a result about 40 millionAmericans
nowhaveno healthinsurance(EmployeeBenefitResearch
Institute 1996).
"Guild
freechoice":Doctor-created
morketfailure
These market failureswere compounded by a system of
health carefinanceand organizationcreatedand enforced
bythe medicalprofession,a systemcharacterizedbyWeller
(1984) as "guildfree choice."The principlesof the med-
icalguild,in the UnitedStatesand othercountries,had sev-
eral definingfeatures:
Freechoiceofprovider-meaningthat everyhealthinsur-
F ance plan. It was an uphillbattle for HMOs to persuade
ance planlet everypatient freelychoose amongdoctors
and hospitalsfor covered serviceswithout discrimina-
tion. This approach destroys the bargainingpower of
insurers: they cannot tell doctors that their enrollees
willnot be coveredfor the doctors' servicesif the doc-
tors do not agreeto price and qualitycontrols.
* Freechoiceof treatment-meaning no practice guide-
linesor qualitymanagement.
* Fee-for-servicepayments-meaningthat no outsideentity
can controldoctors'incomes.Doctorscouldalwaysearn
more by doing more, regardlessof whether more was
necessaryor beneficialto the patient.
* Directnegotiationoffees betweendoctorand patient-
meaningthat ifpatientsare in painor worriedabout their
health, depend on the good will of their doctor, and
INNOVATIONS IN HEALTH CARE FINANCING
196
lackinformationon the quality,price,andtreatmentpat-
terns of other doctors, they are unlikelyto be able to
negotiatewith their doctor as an equal.
* Solopractice-meaning that multispecialtygroup prac-
ticewas resisted becausesuch a group couldbreak the
seamlesswebof mutualcoercion(throughreferrals)that
the profession used to maintain the guild model.
Moreover,grouppracticeintroducedan importantele-
mentof qualitymanagementthroughpeerreview.Doctors
who deviatedfrom the solopracticemodelwere denied
hospitalstaff privileges,medicalsocietymembership,
and referrals.From the 1940sthroughthe 1960s,what
we now callHMOs were few in number and compara-
tivelysmall, and experiencedintense opposition from
organizedmedicine(Weller1984).
The enforcementoftheseprinciplespreventedthe devel-
opmentof an ordinaryeconomicmarket in whichalterna-
tiveapproachesto cost-effective
carecoulddevelop(Stephan
1978).
Otherhumon-mode
morketfailures
These market failureswere reinforcedby the behaviorof
employersand of organized labor. Most Americanswho
are notretired,disabled,or poorgethealthinsurancethrough
theiremployers(EmployeeBenefitResearchInstitute1996).
Fromthe 1940sthroughthe 1970s(and evenbeyond),most
employersofferedtheiremployeesno choiceof healthinsur-
employersto offer them. Employerssawno advantageto
offeringa choiceamongplans and preferred to stickwith
the fee-for-service model.Employersthat did offerchoices
typicallypaid in full the cost of the fee-for-serviceplan,
depriving employees who chose HMOs the economic
rewardsof choosinga more economicalhealth plan.
EmployersavoidedHMOs becausetheirpremiumswere
often higher,sincein most casesHMO coveragewasmore
comprehensive coverage(includingdoc-
than fee-for-service
tor officevisits,preventiveservices,andthe like).Thefinanc-
ing system created open-ended demand and a perpetual
shortageof doctors,soHMOs didnot havemuchbargaining
power when it came to doctor incomes. Organizedlabor
saw comprehensivehealth insurance coverageas a bar-
gainingprizeanddemandedthat employerspaythe fullcost
of fee-for-service coverage. Trade union leaders acted as if
they believed that health insurance premiums came at the
expense of employers rather than wages. Employers and
labor failed to see health services as a purchasing problem
in which they had a common interest in getting value for
their money. Moreover, most employers were too small to
spread risks and serve as effective purchasers of health
care coverage. Market failure is particularly severe in the
market for small employment groups and individuals.
Insurance carriers compounded the market failure, using
marketing strategies and "product design' to try to select
the best risks or to avoid the worst risks. They differenti-
ated coverage contracts in an attempt to create inelastic
demand by segmenting markets and making it very costly
for consumers to compare value for money and to switch
insurance plans.
Finally, government was a major contributor to market
failure. First, federal and state governments excluded (with-
out limit) employercontributions to employee health insur-
ance from the taxableincomes of employees.Thus employers
and employees had a strong incentive to agree on more,
rather than less, costlyinsurance plans (Enthoven 1984 and
1985). Medicare and Medicaid were frozen in the "guild
free choice" model from which they are only now begin-
ning to emerge. Until 1983 Medicare paid hospitals for inpa-
tient services on a retroactive basis. Some U.S. states and
the federal government passed laws requiring that allinsur-
ance contracts cover the services of providers who sought
protection from the legislatures (for example, insurers in
Rhode Island must cover pastoral counseling), even though
competent buyers and sellers would have preferred less
costly coverage that did not include those services. And
finally, the federal government generously subsidized the
expansion of medical schools and hospitals and the train-
ing of specialists far beyond what was needed, and acade-
mic health centers became important protectorates for some
members of Congress.
Consequences
of marketfailure
The result of these market failures was that, although health
care is the joint product of many components (doctors,
hospitals, laboratories, pharmacies), the U.S. health care
systemwas fragmented. Patients depended on doctors and
MARKET-BASED REFORM OF U.S.
hospitals for care that might appropriately occur in vari-
ous settings, but the different providers were paid out of
separate accounts, and there was no management control
or responsibility to see that the pieces were put together
effectively or economically. The system was nonaccount-
able. Providers could waste resources or make decisions
that were bad for the quality of care without negative
consequences. There was no contractual link between
providers and third-party payers. (The patient was the first
party, the provider the second, and the payer the third.)
Except for a few HMOs, no one was responsible for link-
ing resources to the needs of the population served. The
government created health systems agencies that were sup-
posed to do comprehensive health planning at the com-
munity level, but no one knew how to do health planning;
the agencies could not overcome the cost-increasing incen-
tives described above, and they were ineffective (Enthoven
1980). Hospitals were largely independent nonprofit orga-
nizations that competed for doctors by offering more
advanced technology and convenience. (Such competi-
tion was referred to as the "medical arms race.") The
links between doctors and hospitals were weak. Many doc-
tors practiced in two or three hospitals and played them
off against each other.
Information on treatments and costs was decentralized
to the local care site. Doctors had records of what went on
in their offices, hospitals had records of what went on in
the hospital, pharmacies had records of prescriptions, and
so on. But outside of HMOs, nobody had access to the com-
plete picture. Outcomes were rarely systematicallyfollowed
outside the context of a few studies. In the absence of
comprehensive longitudinal records that stored informa-
tion on all the treatments a patient had received, there was
no way to link outcomes to treatments.
One consequence of this system was cost-unconscious
(or inelastic) demand. Providers had incentives to resolve
all doubts about care by doing more. A phenomenon devel-
oped known as "physician-induceddemand"-that is,where
there were more surgeons per capita there was more surgery
per capita, unmediated by a reduction in price. Because of
their superiorinformation,physicianscould shiftthe demand
curve for their services (Fuchs and Kramer 1972, Fuchs
1978). Providers had little responsibility for quality or cost.
There were, and remain, wide variations in medical prac-
HEALTH CARE FINANCING AND DELIVERY
197
tice. Wennbergand Gittelsohn(1982) found tenfold vari-
ations in the incidenceof medicalpracticeswith no mea-
surabledifferencein need or healthoutcomes.Thisfinding
suggeststhat manypeople are being overtreatedor under-
treated.
Anotherconsequencewastoo manyhospitalsandbeds,
too manyspecialists,and too manyspecializedfacilities.For
example,Californiahas 120 hospitalsthat perform open-
heart surgery,half of them with annual volumesof fewer
than 200 cases.There is also a great deal of inappropriate
care. Winslow(1988) found that about one-third of the
carotid endarterectomiesperformedin the United States
were inappropriate-that is, the patientswouldhavebeen
better off without them. There was little effectivequality
management.And by 1994U.S. spendingon health care
accountedfor 13.7percentof GDP,farmore than anyother
country(Levitand others1996).Thislevelof spendingseri-
ouslystrainspublicfinances,andputs healthcare coverage
out of reachfor manyfamiliesof moderatemeans.
ManagedCare and Managed Competition:
CorrectingMarket Failure and Getting the
Incentives Right
Managedcare andmanagedcompetitionarestrategiesused
by the purchasersof health care servicesto:
* Createhealthservicesdeliveryorganizationscapableof
acquiringappropriatehealth care resources,obtaining
value for money,deployingthe resources to care for
enrolled populations, designing and executing care
processesthat produce good outcomes and value for
money,andmeasuringand monitoringperformance(out-
comes,satisfaction,and cost) and continuouslyimprov-
ing it (that is, managingcare).
* Develop a frameworkof incentivesfor such organiza-
tions to improvequalityand lowercosts.
* Use market forcesto transformthe healthcare delivery
system from its fragmented,nonaccountablemode to
efficient,integrated,comprehensivecare organizations
constantly striving to improve (Enthoven 1988 and
1993a).
In brief, managedcare organizationsare the playerson
this field,and managedcompetitionrefers to the rules of
the gamein whichtheyplay.
INNOVATIONS IN HEALTH
198
Manogedcare:Organizing
the supplyside
The mainoriginsof managedcare in the United Statescan
be traced to severalsources.First, there was the prepaid
group practicemovement,whosefoundationswere laid in
the yearsafterWorldWarII with the creationof the Kaiser
Permanente Medical Care Program and Group Health
Cooperative of Puget Sound (Somers 1971). Kaiser
Permanentewas the direct descendantof the medicalcare
programsorganizedin the 1930sand 1940sto careforwork-
ers in Henry J. Kaiser'sindustrialenterprises.These non-
profitorganizationscombinedmultispecialty grouppractice,
per capitaprepayment,voluntaryenrollment,and physician
responsibilityfor the managementof care. These organi-
zationsare describedin greaterdetail later in this paper.
Second,in some communitiesin whichprepaid group
practicesweresuccessfulandgrowing,doctorsin fee-for-ser-
vice solopracticewere feelingcompetitivepressure.They
formedindividualpracticeassociationsthroughwhichthey
could offerpatients the financialequivalentof the prepaid
group practiceswhile continuingto practicein their own
offices and also see patients with traditional insurance.
(Theseassociationsbecame humorouslyknownas "defen-
sivealliancesagainstKaiser.")
The term healthmaintenanceorganization was coinedin
1970by Dr. Paul Ellwoodas part of a national strategyto
solveAmerica'sproblemsof uncontrolledhealth expendi-
ture growth,fragmentation,and lack of accountabilityby
fosteringcompetition amongnongovernmentalcompre-
hensivecareorganizations (Ellwood,Anderson,andMcClure
1971). In 1973 Congress passed the HMO Act, which
definedHMOs asbeingof eitherthe grouppracticeor indi-
vidualpracticevariety;providedgrants and loans to help
startnonprofitHMOs;requiredthat employerswith twenty-
fiveor moreemployeesthat wereofferingtraditionalinsur-
anceofferto theiremployeesthe choiceofonegrouppractice
and one individualpracticeHMO as alternativesto tradi-
tional health insurance (if such HMOs served the areas
wheretheir employeeslivedand asked to be offered); and
overruledstate lawsthat inhibitedHMO growth.This act
had an importanteffectin openingthe market to compe-
tition.
Seekingto bring soaringhealth care costs under con-
trol,someemployerswantedto be ableto offertheiremploy-
CARE FINANCING
ees healthinsurancebased on selectiveprovidercontract- concurrentreview).A more advancedform of utilization
ing-that is,insurancethat resembledthe traditionalmodel, managementisbasedonthe recognitionthat medicaluncer-
exceptthat employeeswouldbe offeredpreferentialterms taintyis often greatandpracticevariationsare wide.Teams
of coverageif they used contractingproviders.Such con- of physiciansstudy particularmedicalconditions,review
tractingenabledemployersandinsurersto negotiateprices the medicalliterature,analyzetheir owndata, and recom-
and utilizationcontrolswith providers.But until 1982,in mend practice guidelinesbased on consensuswithin the
compliancewith the principlesof "guildfree choice,"this team.These guidelinestypicallyreflectthe leastcostlyway
kind of insurancewas illegalunder the insurancecodes of of achievingthe best possibleoutcomes.
most states.In 1982,in a majorlegislativebattleinCalifornia, The thirdprincipleis negotiatedpayment.Thebasicidea
employers,insurers,andlaborunionsjoinedforcesto defeat is to trade patient volume for better prices. Compared
the CaliforniaMedicalAssociationand securethe enact- with the usual feesin the fee-for-servicesystem,managed
ment of new legislationallowinginsurersto contractselec- careorganizationstypicallyobtain discountsof 20to 40per-
tivelyand passthe savingson to the insureds.Most other cent. These negotiated payments often include some
states followed.This move authorizedpreferred provider bundlingof services-for example,all-inclusivepayments
insurance,the other form of managedcare. perinpatientday(fordifferenttypesof patients)orper inpa-
tient hospitalcase.
Essentialprinciplesof managedcare.Managedcare has The fourthprincipleis qualitymanagement.For exam-
four essentialprinciples.The firstis selectiveprovidercon- ple, a managedcareorganizationis likelyto surveypatient
tracting.Insurerscan selectprovidersfor qualityand econ- satisfaction.Theymayrewardproviderswhoscorewellwith
omy.Qualityisimportantbecauseemployerscare aboutthe bonuses,and maynot renewcontractswith providerswho
healthandsatisfactionof theiremployees(or,iftheydo not, scorepoorly.Sophisticatedmanagedcareorganizationswill
trade unions are likelyto), insurerscare about their repu- attempt to measure outcomes of care or performance of
tations,andmistakescostmoneyMoreoftenthannot,qual- processesof care and report them to consumersand pur-
ity and economygo hand in hand. Providersare alsochosen chasers.
for theirwillingnessto cooperatewith a managedcareorga-
nization's quality and utilizationmanagementprograms, Preferredprovider insurance.Minimal managedcare.
and its reportingrequirements. Preferredproviderinsuranceis the form of managedcare
The second principleis utilization management.This most like the traditionalmodel. (Preferred providerorgani-
varies from the crude to the sophisticated.For example, zationis sometimesused to parallelthe better-established
some managedcare organizationshave retained actuarial healthmaintenanceorganization.) In most casessuch insur-
consultingfirmsto developguidelineson howlongvarious ance is not providedby medicalcare organizations;rather
inpatient cases should remain in the hospital, and these it is provided by insurance companiesthat contract with
guidelinesare translatedinto limitson what the insurance large numbersof providersthat are not otherwiserelated
willpay.Manymanagedcareorganizationsemploy"primary (Boland1985).
care gatekeepers"-primary care physicianswho control The typicalpreferred providerinsurer contractswith a
referralsto specialists.Many managedcare organizations largenumberof doctors,hospitals,laboratories,homehealth
dealingwith doctors from the fee-for-servicesector (who agencies, and the like. It creates incentives for insured
are thought to be overutilizers)require preauthorization patientsto choosecontractingproviders.For example,the
beforea nonemergencypatientcan be hospitalized.A man- insurancecontractmightpayin full for the servicesof con-
agedcare insurancecontractmayincludea deductiblefor tractingproviders,but payonly80 percent ofwhat it would
hospitalizationsthat is waivedif the patient obtainsautho- have paid contractingproviders for the servicesof non-
rization.Somemanagedcare organizationsemployutiliza- contractingproviders;the patient must paythe rest. The
tionmanagementnursesto checkonthe hospitalinpatient's insurernegotiatesdiscountedfees, and the provideragrees
conditionand plan promptdischarges(a processknownas to accept those fees as paymentin full from contracting

MARKET-BASED REFORM OF U.S. HEALTH CARE FINANCING AND DELIVERY

199
patients.Finally,the insureradoptsutilizationmanagement
tools such as preauthorization for hospital admissions,
length-of-stayguidelines,reviewof provider credentials,
and so on.
Somepreferredproviderinsurerscovercomprehensive
health care services.Others specialize,focusingon a sub-
set of comprehensiveservicessuch as mentalhealth, phar-
macy,cardiology,or radiology.These specializedinsurers
serveas subcontractorsto insurersthat covercomprehen-
siveservices.Theycan offergreater detailedknowledgeof
their specialty.And theymaycontractwith severalinsurers
that cover comprehensivecare and subcontractthe com-
ponents.
Preferred providerinsurance often servesas an impor-
tant part of the transition from the traditional unman-
aged fee-for-servicesystem to the HMO that uses per
capita prepaymentinstead of fee-for-service.A group of
doctors maybegin with a discounted fee-for-servicecon-
tract, acquire experience on which they can base a per
capita payment,and eventuallyconvertto per capita pre-
payment (see below).
Healthmaintenanceorganizations. A more fundamental
changefromthe traditionalsystemis representedbyhealth
maintenanceorganizations(HMOs).Thetermbealthmain-
tenanceorganization wasoriginallyused to describeprepaid
group practice, the main example of which was Kaiser
Permanente.It subsequentlywasappliedto individualprac-
tice associations.
Now the termis usedto describea remark-
able variety of organizations. Some are based on
multispecialtygroup practices,someon doctorsin individ-
ual practice,and someon both. Someare merelyinsurance
carriersthat complywith the regulatorydefinitionof HMO.
In general, an HMO is a health insurancecarrierthat
coversa comprehensivelist of health care services:physi-
cian and hospitalcare,laboratorytesting,diagnosticimag-
ing, and usuallyprescriptiondrugs.The coverageprovides
for nominalcopaymentsat the point of service-for exam-
ple, StanfordUniversityemployeespay$10per doctoroffice
visit,but thereis no deductibleand no limiton the amount
the HMO willpay for necessaryacute care. Copayments
are not supposedto be solargeas to constitutea barrier to
care. The HMO is supposedto controlcosts by managing
care, not byimposingdeterrent fees.
INNOVATIONS IN HEALTHI CARE FINANCING
200
The HMO contractswithemployersand individualsub-
scriberson the basisof per capita prepayment-that is, all
the medicalcare the patientneeds for a periodicper capita
paymentfixedin advanceand independentof the person's
actual use of services.Thus the HMO bears the full risk
for the cost of medicalcare.The amount and type of risk
sharingwith providersvary widely.But in most casesthe
HMO sharesrisk withproviders,explicitlyor implicitly. An
explicitrisk-sharingarrangementmightbe a contract with
a medicalgroup to provide all necessaryprofessionalser-
vices fora fixed per capita payment.An implicitrisk-shar-
ing arrangement might pay individual doctors on a
discountedfee-for-service basis,but the HMO keepstrack
of the per patientcostsof eachdoctor,adjustedforage,sex,
and possiblydiagnosis.Doctors whose costs consistently
exceed norms mightfind themselvesreceivingextra coun-
seling on practice patterns, or not have their contracts
renewed.
There are nowmanyHMO models, as the competitive
marketplacehas motivateda great deal of innovationin
the searchfor better waysto organizeand financemedical
care.Historically,the firstmaincategoryof HMO included
prepaid group practice and "staff model" HMOs. Under
the prepaid group practicemodel HMOs are based on a
medicalgroup that contractswith the HMO. The medical
group acceptsthe risksof costsof care and usuallyrewards
the partners if the group is successfulin managingcosts.
Under the staff model the doctorsare salariedemployees.
Thoughthe two modelsare usuallygroupedtogether,there
is a difference.The doctors in the group model are more
likelyto see themselvesas part ownersof the enterprise
and feel moreresponsiblefor its success.Theseare HMOs
"from the ground up." They attempt to organizecompre-
hensive care systems.Their doctors care exclusivelyfor
patients enrolled in their affiliatedhealth insuranceplan.
And they generallycare for patients in HMO-owned or -
leasedfacilities.The prepaidgrouppracticemodelhas been
much more efficientand effectivethan the fee-for-service
model, and in some states (California, Massachusetts,
Minnesota,Oregon) its role was to force fee-for-service
providersto join network and individualpracticeassocia-
tion modelsto compete.
The secondcategoryof HMO includesthe networkand
individualpracticeassociationmodels,whichcontractwith
former fee-for-serviceproviders and enable them to compete
with the group and staff models. In these models indepen-
dent medical group practices and individual doctors contract
with insurance carrier HMOs to see the patients enrolled
with those carriers, while also continuing to see patients
enrolled in traditional insurance, Medicare, Medicaid, or
other or no coverage. The doctors continue to practice in
their own offices.The medical groups are paid on a per capita
basis for professional servicesunder contracts that include
incentives for efficient hospital use. A typicalcontract might
provide a fixed monthly amount per enrolled person for
professional servicesplus a risk-sharingarrangement for hos-
pital costs. Individual doctors are paid negotiated fees for
services, given incentives for economical behavior, and are
usually monitored for the economy of their practice pat-
terns. A typical contract might pay primary care doctors 80
percent of their fees soon after deliveryof services,with the
other 20 percent withheld to ensure that enough money is
left in the pool. At the end of the year the doctors are paid
in proportion to their billings if there is money left over. In
addition, the pool of primary care doctors may share in the
savingsfrom efficient specialist referrals and hospital use.
The trend in these models is toward more integrated sys-
tems of care (see below). The original individual practice
associations were sponsored by county medical societies
and included every fee-for-service doctor who wanted to
participate. Today's individual practice associations select
doctors and drive hard bargains with them. For example,
U.S. Healthcare, founded in Pennsylvania, is based on
doctors in individual practice. It selects a panel of partici-
pating primary care doctors, evaluates them continuously
(through surveys of patient satisfaction, reviews of sample
records and referral patterns, office inspections, and so on),
and pays them an age- and sex-adjusted capitation fee for
primary care services. Good performers win cash bonuses;
poor performers are dropped from the program. U.S.
Healthcare contractswith selected specialistson a discounted
fee-for-service basis, and gives primary care doctors as a
group a financial incentive to control specialist referrals.
HMOs translate the broad incentive of capitated pay-
ment into payment to doctors in a great variety of ways.
Some pay salaries. Others pay salarieswith bonuses for pro-
ductivity,patient satisfaction, and overall economic success.
Others pay various forms of fee-for-service with manage-
MARKET-BASED REFORM OF U.S. HEALTH
201
ment controls. HMOs are empirically tuning their meth-
ods to find what works in their marketplaces.
A third category of HMO that is only just beginning to
emerge is called thephysician-hospitalorganization,in which
one or a group of hospitals team up with their medical staffs
to offer subscribers comprehensive health services on a per
capita prepayment basis. Physicians and hospitals are moti-
vated to do so by what appear to them to be the large profit
margins of for-profit HMOs, and by their perceptions of
the way that they are treated by them. Such organizations
reflect a desire by providers to control their own destinies.
Thus, in principle, managed care could be introduced into
a country without a health insurance industry.
Of course, physician-hospitalorganizationswill find that
they need to develop functions and capabilitiesthat are usu-
allyassociated with insurance companies, includingthe abil-
ity to enroll members, collect premiums, set prices on their
services,make actuarial estimates of the costs of caring for
different groups, make arrangements for covered services
that are beyond their capabilities (for example, to contract
with national or regionalcenters for advanced care), provide
coverage for enrolled members when they are outside the
organization's direct service area, and reinsure exceptional
medical costs. And they will need financialreserves to cover
unplanned losses. Thus the physician-hospitalorganization
ends up creating (or partnering with) an insurance company.
Managed care, especially the HMO, requires a change
in patterns of access to doctors, from complete free choice
to choice from among the managed care plan's contracting
providers. People who are used to the traditional system
often do not understand this change and the reasons for it.
However, experience in California and other states with
high penetrations of managed care shows that people even-
tuallyget used to the new pattern. Surveysby large employer
coalitions in California report that employees are highly sat-
isfied with some HMOs.
When they are seriously ill, some patients want to be
able to go to a famous regional or national referral center
and take their insurance with them. And they often do not
understand why they cannot. The comparatively free mar-
ket for managed care in the United States has generated an
innovation to address such concerns: a new kind of insur-
ance plan called the "point-of-service HMO" that func-
tions as an HMO for people whowant to staywith the medical
CARE FINANCING AND DELIVERY
group that they have chosen within their HMO, but also icalcareorganization.Thepremiumsthat are paidreflect
includesa preferredproviderinsuranceplan,whichthe mem- the overallefficiencyof the providerorganizationaswell
ber can accessby payinga deductible(typicallythe firstsev- as the healthrisksof the enrolledpopulation.Per capita
eralhundreddollarsofexpense)anda portionofeachmedical prepaymentimpliesa reversalof the economicincen-
bill. Suchpatients can alsoopt for a traditionalinsurance tives in the fee-for-servicesystem.Doctors prosperby
plan with somewhatstrongerfinancialdisincentives.These keepingpatientshealthyand bydiagnosingand treating
add-onsallowthe coveredperson the full range of choice theirmedicalproblemspromptlyand effectively. Tertiary
of provider,but grant more favorablefinancialterms for care(open-heartsurgery,organtransplants,servicesthat
usingthe HMO's primarycare group. Such arrangements are usuallyperformedin regionalreferralcenters, and
are provingto be verypopular.The point-of-service HMO soon), considereda majorprofit centerin the traditional
was introduced in the mid-1980s.In March 1987 eleven system,has become a "cost center." Under the tradi-
HMOs reported point-of-serviceenrollment of nearly tional systemfilled hospitalbeds were an indicatorof
400,000.ByJuly1995,318 HMOs servedmore than 5 mil- success;under the new paradigmempty beds are good.
lionenrolleesin point-of-service
plans.Most peoplein these Per capita prepaymentholds providersaccountable
arrangementsend up stayingwith their primarycare group for costs,and for the costs of poor quality.If an opera-
for more than 90 percent of the servicesthat theyuse. tion is done poorlyand leads to complicationsand the
In 1978the HMO industrywasmadeup almostentirely need formoretreatment,providerspaythe extracosts-
of local nonprofitHMOs and KaiserPermanente(then a not the insurersor patients.In that senseper capitapre-
large nationalorganizationserving3.5 millionenrolleesin paymentcancreateincentivesforhigh-qualitytreatment.
sixstates).By 1995thirty-fivenationalHMOs served42.7 Patientswithunsolvedor poorlymanagedproblemscon-
millionof the industry's53.8 millionenrollees.Somedata tinueto imposecostsonthe healthcaresystem.Per capita
on the HMO industry'sgrowthare shownin tables 1 and 2. prepaymentfacilitatesthe alignmentof doctors' incen-
tiveswith patients' interestsin high-quality,economical
Integrated deliverysystems:
Thesevenintegrations.
Theterm care.It paysforand rewardscost-effective preventiveser-
integration,as appliedto healthcare financingand delivery, vices,suchasincreasedoutreachof prenatalcareto reduce
refersto the (at least)sevenwaysin whichthe system'scom- thecostsofneonatologyandbettermanagementof chronic
ponents are being linked more closelyto realizeimportant diseasesto minimizeacute episodes.Prepaymentalso
gains in qualityand economy.Integrationneed not mean providesa frameworkfor cost-benefitanalysis,helping
common ownership.Indeed, the trend is toward integra-
tion by contractualrelationships. TABLE2
1. Betweenfinancialresponsibilityand delivery,throughper U.S. HMO industrycomposition:modeltypesand
capitaprepaymentsbythe purchaserto the chosenmed- tax status,1978,1985,and 1995
(millions
of members)
TABLEI
U.S. HMO growth: localand national firms, 1978, 1978 1985 1995
1985, and 1995 Modeltype
Staff a 3.0 0.8
1978 1985 1995 Individual
practce
associations0.6 6.4 22.1
Network n.a. 5.0 3.3
ofHMOs
Number Group 6.7 6.6 9.1
Local
HMOs 183 298 192 Mixed n.a. n.a. 18.1
Branches HMOs
ofnational 12 187 385 Total 7.3 21.0 53.4
Total 195 485 577
Taxstatus
Numberofmembers
(millions) Nonprofit n.a. 13.6 22.1
Local
HMOs 3.7 10.1 11.0 For-profit n.a. 7.4 31.2
ofnational
Branches HMOs 3.6 10.9 42.7 Total n.a. 21.0 53.3
Total 7.3 21.0 53.8
Total_7_3_21_0_53-8 a. Includedin groupmodel.
InterStudy
Source: data. Source: InterStudydata.

IN HEALTFICAREFINANCING
INNOVATIONS
202
 to determinethe most effectivewaysto spend limited
resources.Andit rewardscost-reducinginnovation,such
as the manyinnovationsthat havedramaticallyreduced
the lengthofhospitalstaysfortotalhipreplacementoper-
ations(Kestonand Enthovenforthcoming).
2. Betweenprovidersandpopulations. This integrationfacil-
itatesandencouragespopulation-basedmedicine,which
adds an epidemiological perspectiveto encounter-based
medicine. Providers examine their encounters with
patients to understand the underlying causes of the
patients' complaintsand to determineif there are effec-
tive methods of prevention.One of the enduring leg-
ends of KaiserPermanente is that when its founding
doctor,SidneyGarfield,wastreatingconstructionwork-
ers whohad nail-puncturewoundsin their feet, he went
to the constructionsitewithahammerandpoundeddown
nails.SomeHMOsnowgivechildrenbicyclehelmetsand
videotapesexplainingwhythey should alwaysbe used.
ThusHMOscan allocateresourcesto maximizethe well-
nessof theirenrolledpopulationsthroughpreventiveand
patient educationservices.Moreover,the defined pop-
ulationbase enablesHMOs to matchthe numbersand
typesofdoctorsin theirgroups,aswellasother resources,
to the needsof the enrolledpopulation.
3. Amongthefull spectrumof healthcareservices,including
inpatientcare,outpatientcare,doctors'offices,andhome
nursing,as wellas with drugs and other services.Thus
the resourcesthatHMOshavebeendevotingto improved
preventiveservicesand outpatient care are more than
being paid for by reduced inpatient costs. As noted,
HMOs seek to delivercare in the least costlyappropri-
ate setting.Theyare motivatedto organizeseamlesscom-
prehensivecareso that patientsare not left to their own
deviceswhen they leavethe hospital.In the best-man-
aged HMOs, committeesof doctors and pharmacists
choose drugsthat produce the best outcomesand min-
imizetotal costsof care,not merelythe leastcostlydrugs.
4. Among doctorsand with otherhealthprofessionals. This
meansthe rightnumbersand typesof professionalsand
the rightmnixof specialiststo ensurethat patients have
good accessand to ensurethat specialistsare proficient
in caring for the patients they were trained to see. It
means rational referral patterns and an efficient
specialist-generalist
divisionof labor.For example,spe-
MARKET-BASED REFORM OF U.S. HEALTH CARE FINANCING AND DELIVERY
203
cialistsmayserveas consultantsto generalistswhoactu-
allydeliverthe care.And it meansefficientuse of para-
medicals,for example, nurse practitioners and social
workersto work in teams with primarycare physicians.
5. Betweendoctorsand hospitals,givingdoctors a serious
interestin reducinghospitalcosts.Underthe fee-for-ser-
vicesystemdoctorshadincentivesto behavein waysthat
increasedhospitalcosts.In awell-integratedsystemdoc-
torsdeveloppracticepatternsthat facilitateefficienthos-
pital operations. They work with hospitalsto reduce
unnecessaryrecordkeeping.Andtheysupport"valuefor
money"investments.
6. Amonghospitals.Groupsof hospitalsin a regioncombine
to shareadministrativesupportfunctions,indudingman-
agementpersonnel,to consolidatevolume-sensitive clin-
icalservicessuchas open-heartsurgery,neonatology,and
laboratories.
7. Withpatientinformation.In the traditionalmodelinfor-
mationondiagnoses,treatments,andcostswasscattered
throughoutthe system;integratedsystemsdevelopcom-
prehensivelongitudinalrecordsfor eachpatient so that
everyproviderwho contactsa patient can have a com-
plete picture of the patient's medical history. This
approachhelpsto avoidduplicatetests and unfavorable
druginteractions.Moreover,the informationcan serve
as a basisfor researchon the relationshipbetweendiag-
noses, treatments, and outcomesthat is impossibleto
performwithout such information.
Controversy overmanaged care.Managedcarehas become
extremelycontroversialin the United Statesevenwhile(or
perhaps because) it is growingwith remarkable speed.
Concernsare expressedalmostdailyin the press. Perhaps
the greatest concern, often expressedby doctors, is that
the incentivesof per capitaprepaymentwillmotivatedoc-
tors to do less than they otherwisemight do to improve
patients'health("underservice"or "skimping").Thispoten-
tial shortcomingis often seen as the mirror image of the
incentivesfor excesstreatment under the fee-for-service
system.
For allthe sound and fury,there is remarkablylittle evi-
dence of reductionsin servicesat the expenseof patients'
health,at leastin establishedHMOsthat servethe employed
middle class. (Managedcare has been more problematic
whenstates havecontractedwith or createdorganizations
that specializein caringfof the poor.) Forexample,a recent
surveyof the literaturefound that "the HMO and [fee-for-
service]plansprovidedenrolleeswith roughlycomparable
qualityof care,accordingto processor outcomesmeasures.
Fourteenof 17observationsfrom 16studiesshowedeither
better or equivalent(sameor a mixtureofbetter andworse)
quality-of-careresultsfor HMO enrolleescomparedwith
[fee-for-service]enrolleesfor a wide range of conditions,
diseases,orinterventions"(MillerandLuft 1994,p. 1,516).
The frameworkin whichmanagedcareorganizationscom-
pete (managed competition, discussed below) needs to
includemeasurementandoversightof quality,aswellas dis-
incentives to skimp on or underprovide services,while
encouraging cost-reducing innovation-a delicate balance.
Related to the underservice issue is a question of trust
in doctors. Some managed care arrangements (for exam-
ple, a typical prepaid group practice) leave salaried doctors
in a neutral financialposition between doingmore and doing
less. But some arrangements give doctors strong financial
incentives to provide less care, raising questions of whether
people can trust their doctors. Compounding this prob-
lem, a few HMOs indude wording in their contracts with
doctors that are intended to limit what doctors can say to
their patients-so-called gag rules. There is nothing intrin-
sic to HMOs that requires gag rules, and the HMO indus-
try association opposes them. Apparently, this practice is
fast disappearing in the face of public protest and, in some
cases, legislation.
A related controversy inappropriately targets managed
care for something that is part of every health insurance
contract and public insurance program-that is, exclu-
sions from coverage. The federal government's Medicare
program and private health insurers, whether managed or
not, have typicallyexcluded coverage of unproven or exper-
imental therapies, and individual tragic cases have inspired
intense controversy.The U.S. media has often confused this
situation with managed care.
Traditional insurance contracts were remarkably open in
their coverage. Insurers were usually not really at risk for
the costs of care-they passed them back to employers or
to the government. So they did almost nothing to control
costs. Under pressure to cut costs, insurers are tightening
their definitions of covered services, often leading to dis-
INNOVATIONS IN HEALTII CARE FINANCING
204
appointmentsand unhappysurprises.For example,in an
effortto reducethe inappropriateuse of hospitalemergency
departments,HMOs have sometimesrefused to pay non-
contractingprovidersforservicesthat theydidnot consider
to be relatedto genuineemergencies.Whatis needed (and
are nowemerging)areindustrystandardsthat defineclearly
what is and is not coveredin waysthat people can under-
stand and accept, and proceduresfor the prompt resolu-
tion of disputes.
The arrivalof large-scalemanaged care in the United
States coincideswith the recognitionthat the resources
devoted to health care must be limited. With limited
resources, some kind of management is inevitable.The
choice is whether management willbe done by govern-
ment as the single payer, with all its rigidity,insensitivity to
local conditions, and susceptibility to "pork barrel" politics
and corruption, or whether it will be done by competing
private entities from which people can choose and that can
be held accountable by subscribers and government.
Managed care isa complexbundle of innovativesolutions
to the problems that characterized the traditional fee-for-ser-
vice system. Innovation means trial and error; in hindsight,
some effortsturn out to be poor ideas that generatemuch-pub-
licized complaints. Doing managed care well is a complex
business. Not everyone understands it, especiallywhen it is
growing so rapidly."Good" managed care seeks to substi-
tute better, more satisfying,less costlycare processes; "bad"
managedcare seeksto limitcostsbyiimposingacross-the-board
numerical limits that apply poorly to individual cases. Over
time, in an appropriatelystructured market servinginformed
people, the good will drive out the bad. But in the short run
the poor performers may be confused with the good and try
the public's patiencewith the entire endeavor.Countries con-
templating managed care should recognize that a great deal
oflearningwillbe needed on the part of patients and providers.
Of course, few developing countries will have to deal with
the expectations of a population that has had the freedom
and lack of personal responsibility for health care costs that
most U.S. citizens have had.
Finally,followers of the U.S. debate over managed care
should remember that the growth of managed care in the
United States is very much to the economic disadvantage
of doctors, nurses, hospital workers, and other providers
of care, who now find themselves in a more normal eco-
nomic market than the one theypreviouslyenjoyed.This
is bound to influencetheir judgmentsabout managedcare.
For-profitand nonprofitarrangements. Until about 1980
the HMO industrywas dominatedby nonprofitorganiza-
tions.Sincethen for-profitcarrierHMOs havegrownmuch
faster than nonprofits(see table 2). In 1995for-profitcar-
riers covered58.5 percent of HMO members.The situa-
tion is complex,however.
Forexample,StanfordUniversityemployeesare covered
by four HMOs. One, KaiserPermanente,is madeup of a
nonprofitinsuranceplan, a nonprofithospitalcorporation,
and the PermanenteMedicalGroups,whichare organized
asfor-profitprofessionalcorporations.Butthe corporations'
shares are held only by the doctors, and are not publicly
traded. AnotherHMO, Blue Shield,is a nonprofitcarrier
that marketsthe servicesof the nonprofitPaloAltoMedical
Foundationand StanfordFacultyPracticePlan. (In another
community,nonprofitBlue Shieldmightcontractwith the
for-profitColumbiaHospital Corporationof Americafor
hospitalservicesand with a medicalgroup organizedas a
professional corporation.) The two other HMOs are
for-profitcarriersthat marketthe servicesof thesetwo non-
profit providergroups.Their sharesare publiclytraded.
InmyviewStanfordemployeesarewell-served bythismix.
Thenonprofitsofferimportantbenefits,suchasmoreresearch
andcommunityservice. In 1980therewaslargeunmetdemand
forHMOsandcost-contained medicalcare.Withafewexcep-
tions,such as KaiserPermanente,the nonprofitHMO sec-
tor did not expandfast enoughto supplythe demand.The
for-profits,with theirsuperioraccessto capitaland stronger
incentivesto grow,expandedto meet the demand,bringing
the livelycompetitionandlowercostswe nowenjoy.
The setup that works willdepend on a country'sbusi-
nessculture,regulatoryinstitutions,and so on. Developing
countriesthat are interested in creatinga rapidlygrowing
HMO industrywill likelyneed to include for-profitcom-
panies in the mix because of their accessto capital and
strongerincentivesto growand innovate.Concemsabout
these companies'performancein meetingsocialobjectives
can be addressed by regulation (communityrating, rules
governingaccessto doctors,and so on), contractualprovi-
sions with purchasers,and measurementand oversightby
employers,employercoalitions,and other payers.
MARKET-BASED REFORM OF U.S. HEALTH CARE FINANCING AND DELIVERY
205
Managedcompetition:
Organizing
the demandside
Managedcompetition is a market-makingfunction per-
formedby largegroup purchasersof healthinsurancecov-
eragesuchaslargeemployers,coalitionsof largeemployers
(suchas the California-basedPacificBusinessGroup on
Health), coalitionsof small and medium-sizeemployers
(suchas the Health InsurancePlan of California,a pooled
purchasingcooperativecreatedby the state for employers
with three to fiftyemployees),the federalgovernmentfor
Medicarebeneficiariesaswellas forits ownemployees,and
state governmentsfor state-sponsoredMedicaidbenefi-
ciariesas wellas their own employees.These entities are
sometimesreferred to as sponsors.In 1986I introduced
these conceptswith the followingparagraphs:
Many proponentsand criticsof the competitionidea
sharethe misconceptionthat competitionmeansa mar-
ket madeup of healthcare financinganddeliveryplans
on the supply side and individualconsumerson the
demand side, without a carefullydrawn set of rules
designedto mitigatethe effectsof the market failures
endemicto healthcarefinancinganddelivery,and with-
out mediationby some form of collectiveaction on
the demand side. Such a market does not work. It
cannot produceefficiencyand equity.
Managedcompetitionmustinvolveintelligent,active
agentson the demandside,contractingwithhealthcare
plans and continuouslystructuringand adjustingthe
market to overcomeattemptsto avoidprice competi-
tion. I callthese agentssponsors... . A sponsoris an
agencythat assuresthe membersof a definedpopula-
tion group of the opportunityto buy healthcare cov-
erage. The sponsor contracts with health plans
concerningbenefitscovered,prices,enrollmentproce-
dures,andotherconditionsofparticipation....Sponsors
maybe publicor private.(Enthoven1986,p.106)
Sponsorsperform severalfunctionsto managecompe-
tition,outlinedbelow.
Selectingthe competitors.Sponsors must first decide
whichmanagedcare organizationsand carriersshould be
allowedto compete to serve the sponsored population.
Criteriafor selectionshouldincludequalityof care,finan-
cial solvencyand stability,competence of management,
and willingnessto operate in support of the goals of the
program.U.S. employershave had an important advan-
tage in this role. They can contract at will and use their
judgmentto decide subtle issuesof qualityand employee
satisfaction.
In general,nohealthplanhas anentitlementtobe offered,
and an employercan decide not to renewa contract with-
out havingto provepoor performancein court. However,
employershave exhibited deficienciesin selectingcom-
petitors, one of which is to prefer one or two carriersto
more choicesbecause doing so creates less work for the
employeebenefitsoffice.Governmentpurchasersare usu-
allyconstrainedby laws and rules that precludethe exer-
cise of judgment. (However,the statute establishingthe
FederalEmployees'Health BenefitsProgramfreedthispro-
gram from normal procurementlawsand regulationsand
allowsits managersto exercisediscretionaryjudgment, which
they havedone with some distinction.)Public purchasers
usuallyhave to specifythe criteriaand then contract with
anyorganizationor carrierthat meets them- evenif they
are poorlyqualifiedin the judgment of the officialsman-
agingthe program.Moreover,public officialsusuallymust
tryto enforcethe letter ofthe contractin a situationin which
the 'product" is difficultto define. The buyer-sellerrela-
tionshipgivesthe sponsora tool for enforcinggood-faith
compliancewith the spirit of the contract evenwhen its
terms are imprecise.
Creatingprice-elasticdemand. To create price-elastic
demand,sponsorsbeginby organizingcoordinatedannual
open-enrollmentperiods, which give all participants an
opportunity to consider alternativehealth plans, review
informationabouttheirpricesandperformance,andchange
plansif desired.(Experiencein the UnitedStateshas shown
that annual enrollmentswork well;in principle,a longer
period could alsowork.) To a point, a longerlist of com-
petitors creates more substitutes, and closer substitutes
help make demandmore elastic.Beyondthat point, more
choices,cspecially
if theymerelyreplicatethe choicesalready
being offered, may create a bewilderingarray of options
that raises switchingcosts and discouragespeople from
makingchanges.
INNOVATIONS IN HEALTH CARE FINANCING
206
Next, subscribersneed to be fullyresponsiblefor pre-
miumdifferences.That is, ifsubscriberschooseplanswith
higherpremiums,they shouldbe required to paythe full
differenceso that they havean incentiveto seek valuefor
money.This principleis nowobservedby about one-quar-
ter of Americanemployers.In the United States a compli-
cating factor in creating subscriber responsibility for
premiumsis the fact that employer-basedhealthinsurance
contributionsare tax-freeto the employee,without limit.
As a result, at the margin,the governmentis subsidizing
people who choose more costlyhealth plans. This short-
comingcouldbe correctedbya limiton the tax-freeamount,
set at the premiumof the low-pricedplan.
Next, sponsorsshould standardizecoveragecontracts.
Ideally,this would mean one standard contract offeredby
all the health plans competing in a sponsored group.
However,HMOs relyon providerincentivesand careman-
agementto control costs,whilepreferred provider insur-
ance relies more on patient cost sharing (coinsurance,
copayments,and deductibles)to deter excessuse. So in a
group in which HMOs and preferred provider organiza-
tions compete,it may be necessaryto offer two coverage
contracts; one for each type.And if some of the HMOs
offer a point-of-servicefeature, there may be a need for
three. Standardizationsimplifiesthe choice,shiftingatten-
tion from detailsof coverageto overallquality and price.
Standardizationalsolowersthe informationcostsof switch-
ingplans,and isan importantwayof combatingmarketseg-
mentation and making different health plans better
substitutes.
Next, sponsorsmust provide informationon the qual-
ity of care and service.(Sponsormeasurementof qualityis
discussedbelow.)However,the informationrequirement
should not be overstated.HMO competitionhas worked
quitewellfordecadesin somelargeU.S. employmentgroups
withno formalquantitativemeasuresof qualityor perfor-
mance.Peoplegot the informationtheyneeded by asking
theirfriends.The needforperformanceinformationapplies
equally to a competitive market model and a govern-
ment-managedmonopolyif the monopolyis to do a good
job of managing.
Finally,for demand elasticityand other reasons,there
shouldbe no waitingperiods or exclusionof coveragefor
preexisting conditions.When people switch plans, they
must be fullycoveredas of the day the coveragebecomes
effective.
Managing riskselection.
Theincentivesofhealthcareorga-
nizationsneed to be directed toward givingbetter care at
lowercost,not towardselectingrisks.Withoutcarefuldesign
to offsetrisk selection,avoidingthe coverageof poorhealth
riskscanbe the shortestroute to profitability,
andit is impor-
tant that managedcompetitiondesignagainstthis.
The firstcomponentof this designis to havethe enroll-
ment processrun by an independentclearinghouserather
than requiringwould-be subscribersto deal directlywith
competing health plans,because in the latter case there
would be opportunitiesfor the plans to be hospitableto
apparentgoodrisksandinhospitableto poor risks.An essen-
tial rule of enrollmentis that anymemberof the sponsored
group can join or re-enrollin anyparticipatingplanthat he
or she chooses.The health plans cannot pick and choose
enrollees.Anyattemptto do soshouldbe viewednegatively
and punishedby the sponsor.
The second componentis to standardizethe coverage
contract, because nonstandard contracts can be used to
selectrisks.
The third componentis risk-adjustedpremiums.That
is, the health care costsof the populationsenrolledin the
differentplansneedto be estimated,basedon demographic
and diagnosticfactors,and compensatorypaymentsmade
from surchargeson the premiums of the plans enrolling
the good risksto the plans enrollingthe bad risks so that
risk selectionis removedfrom the competitivemarket.For
example,health plansthat enroll a disproportionateshare
of diabeticsshould be compensatedso that they are not
punished in the marketplace,givingthem no incentiveto
cut back their endocrinologydepartmentand makethem-
selvesless attractiveto diabetics.A great deal of sophisti-
cated researchhas been done on this topic,but so far there
has beenlittlepracticalapplication(Weinerandothers1996;
Ellis and others 1996).
In fact, the competitivemarket in Californiahas been
workingquiteeffectivelyin recentyearswithoutrisk adjust-
ment,and there islittleevidenceof the negativeeffectsthat
economictheorywouldpredict. Butreal pricecompetition
is in its earlydaysin California,and these problemscould
becomemore seriousas competitionheats up. The Health
MARKET-BASED REFORM OF U.S. HEALTH CARE FINANCING AND DELIVERY
207
InsurancePlan of California,a purchasingpool for small
employers,has installeda risk-adjustmentsystembased on
age,sex, geographicarea,and medicaldiagnosticinforma-
tionobtainedfromhospitalrecords(HealthInsurancePlan
of California1995).However,this programis far too small
to havea significanteffect on the incentivesof the partic-
ipating health plans and their behavior. For risk adjust-
ment to reallywork, a majorityof the market would have
to be usingit. Still,the HealthInsurancePlan of California's
experience shows that the practicalissues of doing risk
adjustmentare manageable.
Establishingequitablerulesfor enrollmentand pricing.
Certainrules of enrollmentand pricingneed to be applied
within each sponsoredgroup (for example,employeesof
the U.S. government).Theserules are usuallyembodiedin
the contractsbetweenhealthplansandsponsors.First,every
coveredpersonis guaranteedthe rightto enrollin the plan
of his or her choice,and to renewcoverageat each annual
enrollment.
Second, the enrolledpopulationis dividedinto rating
categories,and the health plans quote uniformprices for
everybodyin a category.For example,at StanfordUniversity
wehaveseparatepremiumsfor singleadultemployees,for
singleparentswith one or more children,forcoupleswith-
out children,and for familieswith children.All the single
adults enrolled in a particular health plan pay the same
premium, regardless of their health status. The federal
governmentsimplyuses"individuals"and "families"for its
employees.The Health InsurancePlan of Californiaalso
uses age categories,becausein their market environment
they couldnot succeedif theytried to makeyounggroups
subsidizeolder groups.
Measuring andreporting
quality.Both to improvequality
and to make the market work better, major sponsorsare
investingin effortsto measureperformanceand qualityand
to report it to consumers.First, they have developedsur-
veysof consumerexperienceand satisfaction.These sur-
veysrangefrom pointedquestionsabout qualityto general
questionsaboutsatisfactionwithdoctors.(One ofmyfavorite
examplesis, "If you were hospitalizedin the past year,did
you acquire any illnessor injury in the hospital?") The
CaliforniaPublic EmployeesRetirementSystemrecently
started analyzingthe repliesof people who had been hos-
pitalizedor were high users of care to see if they were as
satisfiedas the healthyconsumers.The methods for such
surveysneed not be very complex.
Second,the NationalCommitteeforQualityAssurance,
a private nonprofitorganizationsponsored by employers
and HMOs, has developeda data set that all HMOs sup-
ply.Data include measuresof preventiveservices,such as
the percentageof two-year-olds whohavehad theirrequired
immunizations,the percentage of women who have had
periodicmammograms, and the percentageofdiabeticswho
havehad an annualretinaexamination.In Californiathese
data are audited. The California Public Employees
RetirementSystempresents these data to consumersin a
"reportcard" format.
Some states have systemsthat measure risk-adjusted
mortalityrates (thatis,mortalityratesfrom coronarybypass
graftsurgery,childbirth,andthe like,adjustedforthe patient
mix).This willgraduallybecome a widespreadpracticein
the United States.
Finally,major purchasers (government and private
employers)havecreatedthe FoundationforAccountability
to explore and endorse more advanced quality measures
relatedto specificmedicalconditionsand outcomes.
Poolingpurchasingandsponsoringtheunsponsored.
Managed
competitionis wellsuitedto largeemploymentgroupsthat
can mobilizethe resourcesto manage,contractwithhealth
plans on more or less equal terms, enforce demands for
information,and achieve economiesof scalein adminis-
tration. In largeemploymentgroupsthe employercontri-
bution to employeehealthinsuranceis the gluethat holds
the group together as a purchaser,preventingpeople with
the lowesthealthrisksfrom splittingoffand buyingcheaper
insuranceon theirown.Health insurerscan onlygainaccess
to the employer contribution by contracting with the
employer.
The most successfulmanagersof competitionare large
coalitionsof employerssuchas the PacificBusinessGroup
on Health (a California-basedcoalition of large private
employers),a nationalmultimarketconsortiumorganized
by AmericanExpress, and the Health Benefits Program
of the CaliforniaPublic Employees'RetirementSystem(a
pooledpurchasingarrangementfor 1,000public employ-
INNOVATIONS IN HEALTH
208
ers and about 1 million employees, retirees, and
dependents).
The situationis much more challengingfor peoplewho
do not belong to large groups: small employers, the
self-employed,and the unemployed.Tohelp smallemploy-
ers, the statesof Californiaand Floridahavecreatedpooled
purchasingarrangementsthat smallemployerscanjoin.The
Health InsurancePlan of Californiais open to employers
with threeto fiftyemployeesthat agreeto contributeat least
half of the premiumof the least costlymanagedcare plan
on offer.About twenty-fivemanagedcare plansparticipate,
givingmost employeesa choiceof sixto eight plans serv-
ing their area. The plan has workedwell, although it has
not grownrapidly.Its mainproblem is a lack of a powerful
financialincentiveto hold the pool together and prevent
the lowest-riskgroupsfrom splittingoff and seekinglower
premiumson their own.
In a systemof completelyvoluntaryinsurance,adverse
selectionmakessucha schemeunfeasibleforunsponsored
individualssuch as the self-employedand unemployed.If
participationis voluntaryand is not motivated by access
to the employercontribution(or the tax subsidythat sup-
ports it in the United States),people who do not antici-
pate medicalcostswill choosenot to enroll, whilepeople
who expect medicalcosts will.Costs and premiums will
become prohibitivein a phenomenonknown as a death
spiral.
Such a scheme could work for small employers, the
self-employed, and the unemployedif there were a univer-
salvoucherthat couldbe used onlyto buy healthinsurance
through a group plan like the Health Insurance Plan of
California.In the United States,however,we are nowhere
near that stage.
Simplemodelscan work.The precedingdiscussionmay
givethe impressionthat managedcompetitionis necessar-
ily complexand must be supported by advanceddata sys-
tems. This is not the case. Simple models of managed
competition have done quite well. For example, a basic
surveyof patient experienceand satisfactioncan be very
informative.And premiumrisk adjustmentmightbe based
on easilymeasurabledemographicfactorsplusthe presence
of a fewcostlychronicconditions.In fact, a general advan-
tage of decentralizedprivate markets applies here: such
CARE FINANCING
marketscan functionwith lessinformationthan is needed
by a centralizedsystem.
Consequences of managedcompetition.Large-scaleman-
aged competitionis a recent developmentin the United
States.It has beenintroducedmore extensivelyin California
than elsewhere,led mainlyby a few large employersand
purchasingcoalitions.Yeteven though most employersin
Californiahave not appliedall or even most of the princi-
ples outlinedhere, competitionhas become quite active.
In real terms the 1997premiumsfor competitiveHMOs
are about 15 percent less than in 1994.HMO premiums
have for the most part stopped rising across the United
States, although recent newspaperreports have forecast
renewedincreases(Freudenheim1997;Wmslow1997).
Medicalpracticeis changingrapidly,shorteninghospi-
tal staysandfindingwaysto avoidthem altogether.Thesur-
plusofhospitalbedshas beenexposed,andeffortsare under
way in California to find ways to overcomethe political
obstaclesto closinghospitals.A few havebeen closed,and
it is likelythat many more will be in the next few years.
(Toomanyhospitalscanbe badforthe qualityofcarebecause
a minimumvolume of patients is needed to ensure the
proficiencyof healthprofessionals.)The surplusof specialist
physiciansis alsobeingexposed,and specialistsare leaving
California.Servicelevelsare improving.There ismuchinno-
vation to improve quality and serviceand to lower costs
(Enthovenand Singer1996). Nationally,HMO member-
shiphasbeengrowingbyabout 12percenta yearsince1990.
The best managedcare organizationsare doing a great
dealto take advantageof the opportunitiescreatedbyman-
agedcareto improvemedicalcare.They are creatinginfor-
mation systems to study practice variations, measure
outcomes,identifyand adopt best practices,create clinical
practiceguidelines,and monitorprogress.Theydo (orcol-
laborate with organizationsthat do) ongoingtechnology
assessment.Theyinformpatients about healthybehaviors,
chronicdiseasemanagement,and the risksandbenefitsof
alternative therapies. Some apply continuous quality
improvementthat employsinterdisciplinaryperspectives
(Berwick,Godfrey,and Roessner1990).
Facingcompetitivepressuresin Califomia,HMOs are
workinghard to improve customer service.For example,
they offer convenientaccessto "advicenurses" who can
MARKET-BASED
REFORMOF U.S. HEALTHCAREFINANCINGANDDELIVERY
209
help patientsmake gooddecisionsabout the care they are
seeking.They are investingin callcenters to shorten tele-
phonewaitingtimesandexpediteappointments.Theemerg-
ingstandardof accessto doctorsis same-dayappointments
for patientswho think theyneed them (withsome doctor,
if not one's owndoctor).
Downsidesof managedcompetition.Effectivemanaged
competitionrequires a goodunderstandingof how health
insuranceandmanagedcaremarketswork.One of the main
problemshas come from incompleteimplementationof
managedcompetitionconcepts-for example,if employ-
ment groupsdo not maketheir employeessensitiveto pre-
miumcosts,theycan be disappointedifthey offera choice
among HMOs and do not see their costs come down
(Enthoven1993b).
Over the longerrun the incentive(createdby a lack of
good risk adjustment)to avoidor underservepeoplewho
have costlychronicconditionsis a matter of seriouscon-
cern.When provider organizationsenter this competitive
framework,unlessrisk adjustmentis done well,they have
powerfulincentives to make themselvesunattractive to
patientswith costlychronicconditions.This can happenin
subtleways:generouspayandbudgetsforpediatricsdepart-
ments,whichattracthealthyyoungfamilies,and tightbud-
gets for endocrinology(diabetes) and infectiousdiseases
(AIDS).Thisisunfortunatebecausethe patientswith costly
chronicconditionsare the ones that havethe most to gain
from well-coordinatedcare. A wise public policy would
encouragegoodrisk adjustment.
A related concern stems from the fact that Americans
move about and changejobs a great deal. Thus an HMO
consideringa morecostlybut effectivewayoftreatinga dia-
betic today in order to forestallseriouscomplicationsten
yearsfrom now has reasonto believethat the benefitswill
occurwhenthe patientis a memberof anotherHMO. (This
may mean that the qualitymeasurementprogram should
monitorsuch processesof care in detail.)
Managedcompetitionwithoutmanagedcare.I have con-
sistentlylinked managedcare and managed competition
becauseI seemanagedcompetitionin the United Statesas
a frameworkof incentivesintended to drive insurersand
providersawayfromthe traditionalmodel andtowardmore
effectivelyorganizedandmanagedsystemsofcare.Yetsome
analystshaveaskedwhether governmentsand other spon-
sors could use the concepts of managed competition to
improvethe functioningof traditionalinsurancemarkets
evenin the absenceof managedcare. (Bytraditionalinsur-
ers I mean insurersthat do not have selectivecontracts
withprovidersand thereforelack effectivetoolsto modify
their behavior.)I believethat the answeris yes,but in that
eventinsurerswouldbe competingon their efficiencyand
profitmargins,whichaccountforabout 10percentofhealth
insurancepremiums,and on customerservice.
Ibelievethat competitionon "retentions"(thatis,admin-
istrative expensesand profits as a percentage of premi-
ums) contributed to undermanagementand inflation in
the traditionalhealthcare market. Certainly,it is appropri-
ateto subjectsuch servicesto competition,and someman-
agedcompetitionconceptsareneededto makethat happen.
For example,insurancecontractscan be exceedinglycom-
plex to understand and compare.Insurers often use this
complexityto raise switchingcosts,to deter people from
makingthe effort to comparethe alternativesand to con-
sider switching.One dangeris that the best wayfor health
insurersto competeunderthose circumstances, withoutthe
abilityto managecosts,is byselectingrisks.
Apoorlydesigned
incentiveframeworkmightdrive insurersto great lengths
to select risks. Another danger is that purchasers might
assume that the lowestadministrativeexpensesare best,
leadingto undermanagementof care. The absenceof the
possibilityof managedcaregreatlyreduces the advantages
of the privatesectorin healthinsurance.If onlytraditional
fee-for-serviceis allowed,the governmentas a singlepayer
has severaladvantages-economies of scale,simplicityin
administration,efficienciesof poolingriskswidely,monop-
sonypower-that could makethat modelmore effective.
Summaryof managedcompetition.For managedcompe-
tition to work, severalelementsmust be in place:
* The existenceof managedcare organizationsin suffi-
cientnumbersto makemarketscompetitive.Awidevari-
ety of entitiesmightparticipatein this competition.
* Sponsors acting as active, intelligent agents that can
create markets, contractwith managed care organiza-
tions, set rules,monitor complianceand progress,and
offer subscribersperiodicchoice.
INNOVATIONS IN HEALTII CARE FINANCING
210
* Measuresto makedemandprice-elastic,includingperi-
odic choice,ease of switching,subscriberresponsibility
for premiumdifferences,and informationabout com-
petitors.
* Managementofrisk selection,includingan independent
clearinghousefor enrollment,rules ensuringthat con-
sumerscan choose their plan, standard contracts,and
eventualandprogressiveimplementationof risk-adjusted
premiums.
* Equitablerulesregardingpricing,access,andenrollment.
* Informationon healthplanperformancefor consumers.
* Pooled purchasingarrangements(whereneeded).
Relevance for Developing Countries
The relevanceof this experienceto developingcountriesis
neither simplenor obvious.I certainlydo not want to sug-
gestthat other countriesshouldtransplantU.S.-styleman-
agedcareand managedcompetitionto theirhealth systems
without a great deal of carefulthought about which ideas
would or wouldnot work in their societies.
I am uncertainhowthese conceptswouldapplyin devel-
oping countries.I fear the consequencesof a naivebelief
that terms like competition,marketforces,and HMOs are
magicincantationswhoseinvocationwillmake thingsbet-
ter withoutthe long,hardstrugglethat isrequiredto improve
health care systems;such naivebeliefs haveplaguedpoli-
cymakingin the United States.
In 1994WilliamHsiao wrote a soberingeditorial that
began:
Propelled by a decliningfaith in government,many
developingnationshavesearchedfor a "magicpill"to
cure the ills of their underfunded and inefficientpub-
lic sector-dominated health systems.Allured by the
successof free market mechanismsin promotingeco-
nomicgrowth,conservativepoliticiansandeconomists,
starting in the early 1980s,pushed many developing
countries to turn to the free market to finance and
providetheir health services.... The magicof marke-
tization often seducesgovernmentsinto action with-
out a criticalunderstandingof the conditionsrequired
for efficientmarketsand with no referenceto empiri-
cal evidence.(p. 351)
 Hsiao cites four countries in which elements of what
might be called a market strategywere attempted-with
poor results.He concludesthat "to contain cost inflation
and improve allocativeefficiency,the governmenthas to
regulatethe use of expensivetechnology"(p. 356).
One problemwithHsiao'sconclusionis thatin theUnited
States our experiencehas been that federal and state gov-
ernmentsare incapableof regulatingto improveallocative
efficiency.Indeed,many of our worstinflationaryexcesses
can be traced to public policy.Perhaps other countries
havedevelopeda superiorformofgovernmentthat is capa-
ble of regulatingto improveallocativeefficiency.
ButI doubt
it.
Indeed, Hsiao goeson to observethat:
publicsectorprovisionof healthservicesoften suffers
from bureaucraticinefficiency,longwaitingtime, and
unresponsivepublicsectorworkersprotectedby their
unions.Patientsmay not get value for moneyfrom a
rigidgovernment-runsystemwithinefficientanddeper-
sonalizedgovernmentclinicsandhospitals.Often a cen-
trallyplannedhealthdeliverysystemalsolacks.... the
motivationto innovate.(p. 356)
I would add pervasivecorruptionto Hsiao's list of public
sector problems.
So, implementationof managed competitionwill not
be easy,and goodresultswillnot come quickly-a serious
problem in democracieswith politicianswhose horizon
extends onlyto the next election.Thesemarketsare com-
plex and interrelated. Policymakersmust think through
the entiresystemcarefully,and be prepared to makemid-
coursecorrectionsas problemsariseand progressunfolds.
Implementingmanagedcompetitionwillnot occurwith-
out disagreementand struggle.In the United Statesman-
aged competition has been and remains intensely
controversial,mainlybecause it is underminingthe eco-
nomicinterests of providers.The changeshave come at a
time whenhealth spendingis growingrapidly.Employers
and governmentwere forcedto respond.
Managedcompetitionis not a magicpill.It willnot raise
money from workersin the informaleconomyto finance
their health care. It will not insure the uninsured unless
someoneis willingto payfor it and can raisethe moneyto
MARKET-BASED REFORM OF U.S.
do so. (But it can bringlowercosts,makingit easierto pay
for.) In the absenceof carefullydesignedrules and strong
institutions,it willnot preventhealthplansfromseekingto
serveonly the most profitablecustomers.
Eachcountry'shealthcaresystemreflectsits history,cul-
ture, andpolitical,social,and economicsystems.One coun-
try cannot simplyadopt another's health care systemand
expectsuccess.The managedcare and managedcompeti-
tion model is a good fit for the U.S. culture,whichvalues
pluralism,diversity,multiplecompetingapproaches,and
individualchoiceand responsibility,is suspiciousof con-
centrated power and dislikesgovernment,and draws on
constitutionalarrangementsthat favor free markets. The
considerableculturaldiversitywithinthe U.S. health care
systemrequiresa pluralisticapproach.
There are, however,importantfeaturesof the U.S. sys-
tem that other countriesshouldtry to avoid.These include
a lack of broadlybased socialinsurance,an overemphasis
on complextechnologiesthat yieldsmallgainsin healthsta-
tus at the expenseof primarycareandprevention(wespend
billionson neonatalintensivecarebut seemunableto ensure
everypregnantwomangood prenatalcare), and an inabil-
ity to organizecollectiveaction to make the health insur-
ance market work well for small employersand people
who want to buy their own coverage.However,managed
care and managedcompetitioncouldbe built into a system
of universalcoveragewith equitablefinancing(Enthoven
1978;Enthovenand Kronick1989).Still,I believethat the
U.S. experiencesupports severalinsightsthat are of gen-
eralrelevancefor developingcountries.
Change
shouldbe incremental
By incrementalchange, I mean continuous evolutionary
changein whicheach stepbuildson the successof the pre-
vious steps. Incrementaldoes not necessarilymean slow.
For example,HMO membershipin the United States is
growingby 12 percent a year.This is a high rate of change,
but it is a continuousprocess.
Incrementalismis one of the first laws of democracy.
Wemake changesin smallstepsso that people can under-
stand and adjustto them.Weavoidpublicactionsthat cre-
atelargewindfallgainsandlosses.Democraticgovernment
cannotbe seen as inflictingdirectharm to people(Schultze
HEALTH CARE FINANCING AND DELIVERY
211
1977).Institutionstake time to develop.Under managed
care and managedcompetition,peoplerequire a greatdeal
of time to learn how to deliverand how to receiveman-
aged care.As noted above,managedcare requires a great
dealof learning,andmanagedcompetitionrequireschanges
in expectations.Peoplein developingcountriessometimes
speak of health care reform as though it were one single
sweepingchangeafterwhichno furtherchangeis needed.
"Big bang" reform that is sudden, wrenching change is
unlikely to work well. It is better to think in terms of
"mid-coursecorrections"in aprocessof continuouschange.
Pilotprojectsandmultiplecompeting
approaches
are
useful
Theories about health systemorganizationand manage-
ment are not strong enough to predict what willwork in a
completelynew context. What looks and sounds good in
theorymayturn out to be bad in practice.In the 1992-94
U.S. healthcare reformdebatemost academicsandhealth
policy expertswere skepticalthat managed competition
could moderatethe growthof healthexpenditures-even
afterit had startedto work.Socialscienceresearchis always
backwardlooking.When prepaidgroup practicestarted in
the United States,it wasdenouncedbyorganizedmedicine
andgreetedwithsuspicionbymostAmericans.Today,Kaiser
Permanentereceivessome of the highestconsumersatis-
factionratingsamongCaliforniaHMOs, andthe restof the
health care system has had to change fundamentally to
respondto the competitivethreat it poses.Throughits prac-
tical,demonstratedexperienceit has proveditselfin a com-
petitivemarketplaceof health care and ideas.
Individualpracticeassociationsstartedwithbuilt-infatal
defects-namely, theywere dominatedby countymedical
societiesthat used them in an attempt to preservethe sta-
tus quo. They were not selectivein their physicianmem-
bership.Onlyout of failurecamethe fundamentalchanges
that enabledthem to competeeffectively. They changedto
become selective in numbers and types of doctors and
changedtheirmanagementstylesfrom"handsoff" to "hands
on."
Finally,one sizedoes not fit all.Differentmodelswork
for differentpeople in differentcircumstancesin different
partsof the UnitedStates.Forexample,KaiserPermanente,
IN HEALTHCAREFINANCING
INNOVATIONS
212
whichhas been so successfulin California,is makingsome
fundamentalchangesto adaptto marketconditionson the
East coast.Developingcountriesshouldadopt healthcare
frameworksthat allowforthe developmentofmultiplecom-
petingapproaches,sothat theycan trythemalland seewhat
worksbest.
Therearemanyroutesto managedcare
Managedcare organizationscan havea widevarietyof ori-
gins.In additionto the U.S. examplesthat I havepresented,
considerthese:
* Primarycare or multispecialtyclinicsthat are branches
of the ministry of health might be transformed into
free-standingnonprofit entities that accept capitation
paymentand competewith similarentities.
* As has been occurringin the United Kingdom,groups
of primarycarephysiciansmightbe formedthat accept
a capitationpaymentfor a longlist of services(backed
up by reinsurancefor high-costcases).
* Traditionalinsurersmightbe allowedto contract selec-
tivelywith providersand to offer patientsincentivesto
use contractingproviders.
* Traditionalinsurers might studythe performance and
prices of providers,select good performers, and offer
preferentialinsurancecoverageto patientswhouse them.
This would be a form of preferred provider insurance
with implicitcontracts.(Until the legislativechangesof
the 1980s the laws governingtraditional insurance in
the United States prohibitedinsurersfrom discriminat-
ing amongproviders.)
* Public hospitalsmight be transformedinto free-stand-
ing nonprofitentities that accept capitation contracts,
in a manner reminiscentof the U.K. National Health
ServiceTrustHospitals.
* Large employersmight create clinicsfor their employ-
ees and dependents, then open them up to the public
on a prepaid capitationbasis,in a manner reminiscent
of the industrialoriginsof KaiserPermanente.
In additionto nationalmanagedcare firms,for-profits,
nonprofits,andmutualinsurers,HMOs in the UnitedStates
havebeen sponsoredby businesses,communitycoopera-
tives,countygovernments,hospitals,multi-hospitalcorpo-
rations,physician-hospitaljointventures,medicalsocieties,
physiciangroup practices,unions,universities,and others per capita prepaymentbasis. HMOs make major changes
(InterStudy1995). to increaseeconomicefficiencywhenthey are subjectedto
competition.Ultimately,individualsandinstitutionswilldo
A developmental approachcanfosterinnovation
and what is rewardedby society(especiallywhatis rewardedby
increase popularsupport more resources)and moveawayfrombehaviorthat is not
rewarded.
Health care reform does not haveto happen all at once. There is no perfectincentivescheme.Managedcompe-
Indeed, it should not. Managedcare and managedcom- tition is an attempt to create a balance of incentives in
petition in the United States should be seen as historical favorof providersproducingvalue for money Incentives
processes. They started with some doctors, employers, can be used to correct seriousproblems,as happened in
consumers, and trade unions wanting to try a different the United States when growth in health care spending
approach,prepaidgrouppractice.Tosurvive,prepaidgroup was excessive.New incentivesmay create new problems
practicehad to sell employmentgroupson the idea of the that require adjustmentsin the incentivescheme.
individualemployeehavinga choiceof a plan.Americans
believein choice.As increasingnumbers of people began Integration of deliverysystems andcare processesis
choosingprepaid group insurance,fee-for-servicedoctors important
began to suffer and so organizedcompetingalternatives,
individualpracticeassociations. As explainedabove,the processof careis the orchestration
Thenthe governmentmadesomestrategicinterventions. ofmanycomponentsforthe purposeofimprovingthe health
First,the federal governmentofferedits employeesa wide of apatient.Effectiveness canbe enhancedand costsreduced
range of choice, includingprepaid group plans and indi- considerablyif the componentsare integratedwell-that
vidualpracticeassociations.Then, in 1973,a lawwaspassed is,iftheyworktogethertowarda commongoal.Developing
authorizinggrants and loans to nonprofitHMOs, requir- countriesshoulddevelopintegratedsystemsin their over-
ing employersto offerthem as a choiceand overridingpro- all strategiesto see if they can realizethese advantages.
visionsof state lawsthat inhibitedHMO growth.Only in Managedcompetitionis not a simple,staticmodel. It is
1991did myemployer,StanfordUniversity,adopt standard a journey guided by microeconomicprinciples,empirical
benefitsand makeemployeessensitiveto premiumprices. research,and carefulobservationof what works.
Twoyears later, having seen that Stanford survivedthe
change,the Universityof Californiasystemfollowedsuit. References
And so the storyunfolds.The keyis to create a climatein
whichdesirableinnovationscan occur. Arrow,K J. 1963."Uncertainty andthe WelfareEconomics of
Medical Care."American Economic Review53:941-73.
Incentives
matter Berwick,
D.M.,A.B.Godfrey,andJ.Roessner.1990.Curing
Health
Care.SanFrancisco,
Calif.:Jossey-Bass.
Boland,Peter.1985."TheNewHealthcareMarket:A Guideto
Perhapsbecausethe subjecthas been dominatedby health PPOsforPurchasers,
PayersandProviders."DowjonesIrwin,
professionals,
muchifnot mostof theworld'sthinkingabout Homewood, 11l.
healthsystemorganizationis uncontaminatedbyeconomic Ellis,R.P, G. C. Pope,L. I. lezzoni,J. Z.Ayanian,D. W Bates,
insight.Doctorsandnursesevenconsiderit insultingtosug- H. Burstin,and A.S. Ash. 1996."Diagnosis-Based Risk
gestthat economicincentiveshaveanythingto dowith their AdjustmentsforMedicareCapitationPayments."
HealthCare
behavior.Butthe fact is that economicincentivesdo make Review17(3):101-28.
Financing
behavior. ButthatstateooEllwood,P M., N. Anderson,andW McClure.1971."Health
a difference to the behavior of individuals and organiza- MaintenanceStrategy."MedicalCare291 (May):250-56.
tions.The U.S. experienceclearlyshowsthat. Doctorsfor- EmployeeBenefitResearchInstitute.1996."Sourcesof Health
merlypracticingin the fee-for-servicemodel changetheir InsuranceandCharacteristics
of the Uninsured."
IssueBrief
behaviormarkedlywhen their servicesare purchasedon a 70.Washington,D.C.

REFORMOF U.S. HEALTHCAREFINANCINGAND DELIVERY

MARKET-BASED
213
Enthoven,A. C. 1978. "Consumer Choice Health Plan." TheNew
England Journal of Medicine 298 (12 and 13): 650-58 and
709-20.
. 1980. Health Plan. Reading, Mass.: Addison-Wesley.
. 1984. "ANew Proposal to Reform the Tax Treatment of
Health Insurance." Health Affairs 3(1): 21-39.
.1985. "Health Tax PolicyMismatch." Health Affairs 4(4):
5-14.
- 1986. "Managed Competition in Health Care and the
Unfinished Agenda." Health Care FinancingReview (annual
supplement): 105-19.
- 1988. "Theory and Practice of Managed Competition."
In Health CareFinance.Amsterdam: North-Holland. DeVries
Lectures Vol. 9.
1993a. "History and Principlesof Managed Competition."
Health Affairs 12(supplement): 24-48.
-. 1993b."WhyManaged Care Has Failedto Contain Health
Costs." Health Affairs 12(3): 27-43. - Schultze, C. L. 1977. The Public UseofPrivateInterest.Washington,
Enthoven, A. C., and R. Kronick. 1989. "A Consumer-Choice
Health Plan for the 1990s: Universal Health Insurance in a
System Designed to Promote Quality and Economy." New
England Journalof Medicine 320(2): 29-37 and 94-101
Enthoven, A. C., and S. J. Singer. 1996. "Managed Competition
and California's Health Care Economy." Health Affairs 15(1):
39-57.
Freudenheim, M. 1997. "Health Care Costs Edging Up and a
Bigger Surge is Feared." New York Times, 21 January, p. 1.
Fuchs, V R. 1978."The Supply of Surgeons and the Demand for
Operations." Journal of Human Resources 12(supplement):
35-55.
Fuchs, V R., and M. J. Kramer. 1972. "Determinants of
Expenditures for Physicians' Services in the United States,
1948-68." NBER Occasional Paper 117. National Bureau of
Economic Research, Cambridge, Mass.
Health Care Financing Review. 1996. "Medicare and Medicaid
Statistical Supplement."
Health Insurance Plan of California. 1995. "Methods for
Calculatingand ApplyingRiskAssessmentand RiskAdjustment
INNOVATIONS
IN HEALTHCAREFINANCING
214
Measures: Results of Simulation No.2, Revised." Sacramento,
Calif.
Hsiao, W. C. 1994. "Marketization-The Illusory Magic Pill."
Heath Economics 3: 351-57.
InterStudy. 1995. The InterStudyCompetitiveEdge,PartII: Industry
Report. Excelsior, Minn.
Keston, VJ., and A. C. Enthoven. Forthcoming. "Re-Engineering
Total Hip Replacement: A Case History of Innovations to
Improve Quality While Reducing Costs." Health Care
ManagementReview.
Levit, K. R., H. C. Lazenby, L. Sivarajan, and others. 1996.
"National Health Expenditures, 1994." Health CareFinancing
Review 17(3): 205-42.
Lipsett, S. M. 1991. Continental Divide. New York- Rutledge.
Miller, R. H., and H. S. Luft. 1994. "Managed Care Plan
Performance since 1980." Journal of the American Medical
Association271(19): 1512-19.
D.C.: Brookings Institution.
Somers, A. R. 1971. "The Kaiser-Permanente Medical Care
Program." Commonwealth Fund, New York.
Stephan,J-C. 1978.EconomieetPouvoirMedical.Paris: Econornica.
Weiner,J. P, A. Dobson, S. L. Maxwell, K. Coleman, B. Starfield,
and G.F. Anderson. 1996. "Risk-Adjusted Capitation Rates
Using Ambulatory and Inpatient Diagnoses." Health Care
FinancingReview 17(3): 77-100.
Weller,C.D. 1984. "Free ChoiceAsaRestraint of'TradeinAmerican
Health Care Delivery and Insurance." Iowa Law Review 69(5):
1351-91.
Wennberg, J. 1996. The DartmouthAtlas of Health Care. Chicago,
Ill.: American Hospital Publishing, Inc.
Wennberg, J., and A. Gittlelsohn. 1982. "Variations in Medical
Care among Small Areas." Scientific American 26(4): 120-34.
Winslow,C. M., D. H. Solomon, M. R. Chassin, R. H. Brook, and
others. 1988. "The Appropriateness of Carotid Endarectomy"
New EnglandJournalof Medicine 318(12): 721-28.
Winslow,1R. 1997. "HealthCareCostsMayBeHeadingUpAgain."
Wall StreetJournal, 21 January, p. B 1.
Managed Care and Managed
Competition in Latin America
and the Caribbean
Andre CezarMedici,Juan Luis Londono, OswaldoCoelho,
and Helen Saxenian
D issatisfactionwiththe performanceof healthcare
systemshasled to an intensesearchformoreeffec-
tive meansof usingpublic and privateresources
to improvepeople'swell-being. Policyinnovationsin thisarea
depend to a largeextent on the historyand startingpoints
of differenthealthsystems-particularlywithrespectto the
role of the state in the provisionof healthcare services.In
Europe,where the state traditionallyhas playedan impor-
tant role, countrieshave started to questionthe extent of
the state's involvementin healthcarefinancing,andto look
formore effectiveand equitablewaysto separatethe financ-
ingof healthcarefromthe provisionof services.In countries
likethe UnitedStates,wherethe privatesectorplaysan active
role in the fundingand provisionof health services,efforts
havebeen madeto restructureprivatemarkets.Enthoven(in
thisvolume),for example,proposesvariousmodelsof man-
agedcompetitionfor differenttypesof systems.
Overthe pasttwodecadesmanagedhealthcareprograms
in the United Stateshave grownrapidlyand receivedcon-
siderableattention.Althoughmanyother countriesalsohave
managedcare programs,theyhave receivedless attention.
Suchprogramsare becomingincreasingly importantin Latin
Americaand the Caribbean.Althoughthese programsdif-
fer, they share severalcharacteristics.They typicallyeither
provideservicesdirectlyor use selectivecontractingto pro-
videservicesto coveredindividuals. In addition,coveredindi-
vidualsaremoreconstrainedintheirchoiceof serviceproviders
than undertraditionalindemnityhealthinsuranceplans.
Managedcare programsintegratefinancialresponsibil-
ity for and deliveryof healthcare services.Instead of using
solelyfee-for-servicepayments,healthcare purchasersuse
techniquessuch as per capita prepaymentsto providers,
whichput providersat riskfor the costof servicesprovided.
Well-developedmanagedcareprogramsalsouse utilization
reviewand qualitycontrolmanagementto containcostsand
ensurequalitycare.
Advocatesof managedcare arguethat it offers consid-
erablecost savingsovertraditionalinsuranceand improves
the qualityof health care. The realizationof these poten-
tialbenefits,however,dependson the contextin whichman-
agedcaredevelops-particularlythe regulatoryframework
andits enforcement.Managedcareprogramscan growwith
or withouta regulatoryframeworkbased on managedcom-
petition.Managedcompetitionrefersto a regulatoryframe-
work (rulesof the game)that usescompetitionand choice
to enhancethe functioningofhealthcaremarkets.1Managed
competitionoften refers to governmentregulation,but it
canalsoreferto the regulatoryframeworkthat a largegroup
purchaserof healthinsuranceputs into place.California's
PacificBusinessGroup on Health is an exampleof the lat-
ter (Enthovenin this volume).The performanceof man-
agedcare-in termsofits contributionto efficiency,coverage,
and equity-depends on the nature of the market and on
the regulatoryframeworkand its enforcement.Assuming
a varietyof forms and approaches,the conceptsof man-
agedcare and managedcompetitionare highlyrelevantfor

AndreCezarMediciis healtheconomist in SocialProgramsDivision1 andJuan-LuisLondonio is leadeconomistin the Officeof the

ChiefEconomist at the Inter-American
Development Bank.OswaldoCoelhois healthcareconsultantandHelenSaxenian is principal
economist
in the HumanDevelopment Department at theWrld Bank.Theauthorsaregratefulto DeborahChollet,JillianCohen,and
GeorgeSchieberforhelpfulcomments andto Madelyn Rossforeditorial
assistance.
Thispaperisdedicatedto thememoryofJose-Luis
Bobadilla.

215
LatinAmericaand the Caribbean.This paperreviewsexpe-
rienceswith managedcare and the regulatoryframeworks
in whichit operates in selectedcountriesin the region.
The paper draws two basic conclusions.First, there is
amplescope forthe developmentof managedcare in both
publicandprivatehealthcare systemsin LatinAmericaand
the Caribbean. Traditionalsocial insurance institutions
andprivateprepaidmedicinehavemuchto learnfromman-
agedcare modesof financingand delivery.Second,several
varietiesof managedcompetitionarerelevantforthe region's
healthcaresystems.Underan appropriateregulatoryframe-
work greater competition amongintegrated deliverysys-
tems, togetherwith greater consumerchoice,can improve
the efficiencyand qualityof services.The specificmethods
used to managecompetitionwilldependas muchon coun-
tries' size and level of development as on the methods
used to mobilizefinancing.Giventhe unequaldistribution
of incomeand wealthin the region,the state needs to play
an importantrole in mobilizingresourcesto ensureequity
and efficiency.The weaknessof marketsalsorequiresthat
the state play an activerole in enhancingmarket develop-
ment, to increasethe chancesof improvingthe coverage,
efficiency,equity,and qualityof services.
The Region's Health Care Systems
LatinAmericaand the Caribbeanis a heterogeneousregion:
incomelevelsvaryenormously,as do health status,health
care spending,and health care financingand delivery.In
1995,for example,Haiti's per capita income was $250-
lessthan one-tenthCostaRica's$2,610.In 1990Haiti'sper
capita health spending, at $62 a year,was 13 percent of
Costa Rica's $460.In 1995 life expectancyfor women in
Haiti was 57 years,comparedwith 79 yearsin Costa Rica.
In 1994 the region spent an estimated 6.3 percent of
GDP on health, about half of which came from public
sources.Totalhealth spendingrangesfrom 2.7 percent of
GDP in Guatemalato more than 8.0percent in Argentina,
Belize,CostaRica,andUruguay(WorldBankforthcoming).
Althoughthe region'shealth marketsand health insur-
ance institutionsdiffer significantly,there are four basic
systems:healthcare providersfinanceddirectlybyprivate
out-of-pocketpayments,privatehealthinsurancemarkets
financedby prepaid contributions,socialinsurance mar-
kets financedby mandated employerand employeecon-
tributions, and publicly delivered health care services
financedby generaltaxes and operated by ministriesand
decentralizedregionalhealth offices (table 1). Managed
careappearsin both privateinsuranceandsocialinsurance
systems.
Thesefoursystemsservedifferentbut overlapping groups.
The poor tend to use public hospitalsand clinics,supple-
mented by out-of-pocketexpenditureson private practi-
tionersanddrugs.Formalsectorworkersare typicallycovered
by socialinsurancesystems,althoughmanysupplementthis
coveragewith out-of-pocket expenditures and, in some
cases,privateinsurance.The richtend to buy privateinsur-
ance, although they also may be coveredby social insur-

TABLEI
The four main health servicedeliverysystemsin LatinAmerica and the Caribbean

Sourceof Cost Providerpayment Consumer Criteria

for
System Regulation financing controls mechanism options Ownershipcoverage
Privateout-of-pocket Minimal Out of pocket None Feefor service Many Private to pay
Ability
spending
Privateinsurancea Somefinancial Employers
and Rangeof Rangeof Many Private Ability
to pay
regulation households costcontrols paymentmechanisms
Socialinsurancea Littleor no extemal Mandated Rangeof Rangeof Limited Collective Mandated
regulation payrolltaxes costcontrols paymentmechanisms contributions
Public
servicesfinanced No extemal Generaltax Public Installed
capacity Restricted Public Universal,but
bygeneraltaxrevenue regulation revenue budgets especially
as a
lastresortfor
the poor
a.Caninclude
managedcare.
Source: fromIDB1996.
Adapted

INNOVATIONS
IN HEALTHCAREFINANCING
216
ance andmayrelyon the public sectorfor high-technology
services.This broadclassificationleavesoutimportantvari-
ations, however.In Brazil, for example,the government
contractsto the privatesector manyservicesthat are pub-
liclyfinanced.
Universalaccess,quality,and efficiencyare majorissues
in the region'shealthcaresystems.The poorhavethe worst
access.And despiterapidlyrisinghealthcare costsin many
countries,some analystsbelievethat the qualityof public
serviceshas declined.Manycountries'health care systems
are fragmented.Populationsare servedbyoverlappingand
uncoordinatedsystems.Fewproviders-pubic orprivate-
havethe incentivesto adopt an integratedapproachto the
healthproblemsof the populationstheyserve.Governments
generallyhavefailedto coordinateand regulatethe diverse
segmentsof the health system.
Analystsin the regionhavetraditionallyemphasizedthe
public sector's rolein financingand providinghealth care.
Onlyrecentlyhavetheystarted to examinethe importance
of privateparticipationin healthservices.In the early1990s
total spendingon healthwas believedto total 4.0 percent
of regionalGDP,including1.6percent that wasnot part of
publicspending(WorldBank1993).Researchershavesince
estimatedthat publicspendingon healthserviceswasmuch
higher-more than 3.1percent of GDP-and that private
spendingwasmore than twiceashigh-3.5-4.0 percent of
GDP Moreover,recentresearchhas illustratedthe tremen-
dous progressmade in the organizationof health services
in the Americas.Almost 100 millionLatin Americansare
nowservedby integratedhealthcare deliverysystems,and
more than 60 million are registered with prepaid, non-
governmental,integrateddeliveryorganizations-almostas
many as are servedby health maintenanceorganizations
(HMOs) in the United States.
Thus public financing and provisionof health care is
just one part of the health sector,albeit an importantone.
Althoughthe organizationand structureofthe privatehealth
sector are weak,the conceptsof managedcompetitionare
highlyrelevantto the developmentof the region'shealth
care systems.The specificmethodsused to developman-
agedcare in LatinAmericaand the Caribbeanwilldepend
on countries' size and level of developmentas well as on
the modalitiesof health care systems.In this regard,there
are three types of countriesin the region.
MANAGED CARE AND MANAGED COMPETITION IN LATIN AMERICA AND THE CARIBBEAN
217
In countrieswhere the public sectorhas playeda dom-
inant role in the fundingand provisionof healthcare, the
governmentwillhaveto reorientits role,becominga con-
sumer advocateand workingto ensure greater autonomy
forhealthcareproviders,greaterfreedomof choicefor con-
sumers,and more efficientmodes of resource allocation.
In countries with better-developed markets for private
finance and provision, as well as more experience with
integrateddeliverysystems,itwillbe possibleto experiment
withmore competitivesystemsusingthe financialresources
provided under socialinsurance schemes.And finally,in
countries wherepopulationgroups are highlysegmented
within the health care system, and that typicallyexclude
the poor from health care, competitionmay best be pro-
moted within a publiclyfinanced and regulated system
that expandscoverageand improvesthe qualityof service.
EmergingManagedCare Models
Thedevelopmentof integratedhealthcare deliverysystems
isnot newin LatinAmericaandthe Caribbean.In Argentina
and Uruguaythe developmentof organizationsof health
serviceusers(mutuales)datesfromthe end of the nineteenth
century.SinceWorldWarII many countrieshave experi-
enced growth in social insurancesystemsthat have inte-
grated finance and deliveryof health servicesfor formal
sectorworkers.Theseinstitutionswereconstitutedas pub-
lic monopoliesin most countries,though not in Argentina
and Uruguay,were pluralismprevailed.What is new in the
regionis the recent rapid growthin privateinsurance.
Anumberof managedcare modelshavebeen developed
in the region,includingChile'sProvisional
HealthInstitutions,
Uruguay'sCollectiveInstitutions of MedicalAssistance,
Colombia's Health Promotion Entities, the Dominican
Republic's Igualas Medicas, Brazil's Medical Group
Organizations, andArgentina's ObrasSociales.
Characteristics
of thesemodelsaresummarizedin table2. Morethan 75mil-
lionpeopleare enrolledin these organizations.
These managedcare programsare similarto preferred
providerorganizationsin the UnitedStatesin that, in almost
allcases,coveredindividualsmustuse providerswithinthe
program's network. Providers in the network are often
paid on a capitatedbasis,whichputs them at risk for the
cost of servicesprovided-and can create incentives to
TABLE
2
Characteristicsof managedcare organizationsin selectedLatin American and Caribbeancountries

Integrated Selective
Country/ delivery provider Utilization Negotiated Quality Number
organization systems? contracting? management? payments? management? enrolled
ChileASAPREs Yes Some Yes Yes Heterogeneous 27% of population(3.8 million)
UruguayAAMCs Yes,in mostcases Some Some Fixedpricesset by Heterogeneous 65% of population(2 million)
govemment(in
somecases)
Colombia/EPSs Yes Some Some Yes,risk-adjusted Heterogeneous 50-60% of population(I18-22
capitationset by million)
govemment
DominicanRepublic/ Yes,in mostcases Some Some Yes No 7% of population(0.5million)
IgualasMedicas
Brazil/MGOs Yes Yes Some Yesa Weak 25% of population(40 million)
Argentina/ Some Yes Some Yes,butnot uniform Weak 29% of population(10 million)
ObrasSociales
a. Pricetablesareused,with generalagreements
definedeachyear.
Source:IDBand World Bankstaffestimates.

underprovide services unless countervailing forces check
this tendency. Some of the managed care programs listed
in table 2 provide services directly, and in this way resem-
ble health maintenance organizations(HMOs) in the United
States. Utilization review and quality control management
are the two standard elements of managed care that are
least developed in Latin America and the Caribbean.
How Can Countries Achieve Managed
Competition?
Health care systems in Latin America and the Caribbean
reflect each country's history, culure, and political, social,
and economic development. Since many countries have
highly pluralistic systems, one of the main challenges that
governments face is creating regulatory mechanisms that
encourage competition to induce equity and efficiency and
improve the quality of services.
Unregulated health care markets generate many mar-
ket failures (Arrow 1963; Musgrove 1996). Unregulated
insurance markets, for example, result in adverse selection
by individuals and risk selection by insurers-both major
obstacles for efficient and equitable risk pooling. The goal
of regulation should be to establish rules of the game that
mitigate the market failures endemic to competitive health
care financing and delivery-an objective known as man-
aged competition.2 At the same time, regulatory mecha-
nisms must be adapted to a country's public sector man-
agement capacityand abilityto enforce regulations.Another
challenge for governments is improving access to and qual-
ity of servicesfor the poor in a competitive framework. Some
govemments, for example, are exploring demand-side sub-
sidies so that the poor can choose from competing health
plans (including, in some cases, public providers).
Enthoven(in thisvolume)arguesthat govemmentsand
largepurchasers of health insurance ("sponsors")must deter-
mine, based on factors such as quality of care and financial
solvency and stability, which health insurers or managed
care organizations should be allowed to compete to provide
health care services for the sponsored population. In Latin
America and the Caribbean the sponsor might be the min-
istryof health, social security institute, large private employ-
ers, or other institutions. The sponsor's main roles include:
* Defining the basic health care service package provided
by insurers or managed care programs. A standardized
benefits package simplifiesconsumer choice and focuses
attention on quality and cost.
* Providing information on quality and cost to consumers
so that they can make informed choices.
* Establishing rules of the game so that consumers can
choose the insurer or managed care program that best
suits their needs. These rules typically require coordi-

INNOVATIONS IN HEALTH CARE FINANCING

218
 nated annual open enrollmentand no waitingperiods an averageof 7.4 percent a year (in real terms).The aver-
or exclusionof coveragefor preexistingconditions. ageannualdomesticsavingsrate is 25 percent.And unem-
Managingriskselectionsothat healthcare organizations ployment, which averaged28 percent during 1983-89,
canfocusonprovidinghealthcare efficientlyratherthan droppedto 5.6 percent in the firsthalf of the 1990s.
on tryingto enroll populationswith the lowestrisk of As a regionalleader in terms of economicadjustment,
healthexpenditures.Standardizedbenefitpackageslimit Chile'seconomicandsocialreforms,includingthoseinvolv-
health care organizations'opportunitiesto selectrisks. ing socialsecurityand health care, are consideredmodels
In addition,the sponsor can use risk-adjustedpremi- for the rest of the continent.Between 1960and 1995,for
ums-that is,makecompensatorypaymentsto plansthat example,the under-fivemortalityrate dropped from 155
enrollhigh-riskpopulationsand cut paymentsto plans per 1,000 to 15 per 1,000 (WorldBank forthcoming).
that enrolllow-riskpopulations. Moreover,the maincausesof deathare nowsimilarto those
Sponsorsalso have a role to playin increasingequity. in industrialnations.
They can, for example,channel governmentsubsidiesso In 1981Chilelaunchedtwo parallelreforms:creatinga
that the poor can participatein the systemandenrollin the systemof privateintegrateddeliverysystems-ISAPREs-
same plansas the rest of the population.They alsocan set and partly decentralizingpublic health care to twenty-six
contributionrates in socialinsurancesystemsso that con- regionalhealth centers.Public health care now covers73
tributions from high-incomeworkersand their employers percent of the population;the remaining27 percent is cov-
subsidizecontributionsfromlow-incomeworkersand their ered by the ISAPREs. ll formal sector employeesmust
employers. contribute7 percent of their salariesto the nationalhealth
No country has yet achievedthis idealizedversion of system.(Thereis no employercontribution.)Workerswhose
managedcompetition.In LatinAmericaandthe Caribbean salariesreach a certainlevelcan purchasean ISAPRE(pri-
twodifferentmodelsofmanagedcompetitionhaveemerged. vate)healthinsuranceplan.If workerschooseto buy a more
In some countries(Colombia,Uruguay)the government, expensivehealthplanthantheir 7 percentcontributionper-
as sponsor,is playinga biggerrole in managingcompeti- mits, theymust pay the difference.
tion, interveningin financeand regulation.In other coun- Althoughthe governmentestablisheda basic package
tries (Brazil, the Dominican Republic) large private of servicesto be providedby the ISAPREs,the plans are
employersare playingthe roleof sponsor,withoutanypub- increasinglysellingmore comprehensivepackagesto work-
lic financeor regulation.These effortsare analyzedin the ers. The unemployedand workerswhose salarycontribu-
case studiesthat follow. tionis not highenoughto buy an ISAPREplanare covered
by the National Health Fund, which is generallyconsid-
Case Studiesof Health SystemReforms ered to provide lower-qualitycare. This is not surprising
and ManagedCompetition given that in 1995 per capita spendingby the ISAPREs
was $646, compared with $121 by the National Health
Thissectionanalyzesreformsthat are introducingelements Fund. Copaymentsare significantin the system,and with
of managedcompetitionin LatinAmericanand Caribbean the growthin the system'scoststheyhavebecomeanimpor-
health systems-in Chile, Uruguay, Colombia, the tant sourceof financingfor the servicesprovidedby both
DominicanRepublic,Brazil,and Argentina. the ISAPREsand the NationalHealth Fund.
In recentyearsthe governmenthas implementedother
Chile'ssystemhasincreased competition,but equityisa public sector reforms, such as provider paymentreforms
problem (paymentby diagnosis).Budgetsfor localhealth posts are
nowcapitatedandare linkedto the posts'capacityto recruit
The economicand socialdevelopmentmodel followedby and maintain clientele.Public hospitalshave been given
Chile sincethe early1980sis one of the most advancedin more autonomy,and are increasinglysellingservicesto the
Latin America.During 1991-95 Chile's economygrewby ISAPREs.

MANAGED CARE AND MANAGED COMPETITION IN LATIN AMERICA AND THE CARIBBEAN

219
 Althoughthese reformshave introducedsome compe- forthcoming).The health systemis pluralistic,with many
tition,in that the ISAPREscompeteto providehealthcare privateand publicserviceproviders.Althoughin theoryall
servicesto one segment of the population,the minimum Uruguayansare coveredby a health plan, data indicate
conditionsformanagedcompetitionhavenot beenachieved. that about 6 percent of the populationis not enrolledin
Most ISAPREsenroll healthy,young,high-incomework- anyhealthplan.In 1994healthspendingwasalmost $300
ers. Moreover,the ISAPREshave powerfulincentivesto per capita-about 8.5percent of GDP (WorldBankforth-
makethemselvesunattractiveto patientswith costlychronic coming).
conditions. When individuals' risks for health expenses During the 1970s and 1980sUruguay'shealth system
increase-whether becauseof agingor healthproblems- underwenta seriesof reforms.Formalsectoremployeescan
the ISAPREsraise their premiumsor excludeconditions now choose among health plans, which are provided by
from coverage.The populationis dividedinto rating cate- private Collective Institutions of Medical Assistance
goriesbyindividualrisk. Asa resultmanyhigh-costpatients (IAMCs).Thesystem'sfrequentdeficitsarecoveredbygen-
return to the public system. eral tax revenue.
A regulatoryofficewas recentlyestablishedto regulate The Bank of SocialProvisionfinances,on a capitation
the behaviorof the ISAPREsand to maintaina minimum basis, each worker's basic health plan. The IAMCs now
number of providersin the system.But this officedoes not cover about 65 percent of the population.Worker contri-
havethe legal authorityto ensurethat the ISAPREsoper- butions are set by the government,whichalsohas defined
ate under equitable rules regarding pricing, access,and the basicpackageof servicesto be providedby the lAMCs.
enrollment.Moreover,risk selectionis not managed. Given fixed contributions,IAMCs control their costs by
In summary,Chile'shealth systemstill has manyshort- adjustingthe volumeand qualityof servicesand by exclud-
comings.Equity is a major problem giventhat the public ing high-riskindividuals.The degree of consumerchoice
systemispoorlyfundedrelativeto the ISAPREsandreceives amongthe IAMCsdependsheavilyonlocation.In Uruguay's
no subsidies from the ISAPREs.In essence the country capital,Montevideo,manyIAMCscompetein the market.
maintainsa two-tiersystemwithdifferingservicesandqual- In rural areas,however,there are usuallyonly two IAMCs
ity of care.Althoughqualityof care and consumerchoice per district,and in severaldistrictsthere are none.
haveimprovedfor many high-incomeworkers,the regu- Employeeand employercontributionsto the IAMCs
latory frameworkneeds to be reformed so that competi- only cover servicesprovidedto employees.Thus employ-
tion is better managedto minimizemarket failuresand to ees tend to selectthe IAMCthat offers the best terms for
improveequity,quality,and costcontrol.A managedcom- familycoverage,or they enroll their childrenand spouses
petition regulatory framework is needed so that the in partialinsuranceinstitutionsthat are not requiredto offer
ISAPREshave incentivesto improve servicequality and the basic package.
reduce costs,not to maximizeprofits by attracting low- The public sector coverspeople who are not servedby
risk populations.Public sector networks could be orga- the IAMCs,includingpoor householdswhosehead does
nized to competewith the ISAPREsfor consumersunder not eam enoughto purchasecoverageby an IAMCor a par-
comparablehealth care packages.Direct subsidiescould tialinsuranceinstitution.AswithChile'sISAPREs,the pub-
be givento low-incomegroupsto enhancethe demandside licsectoralsocoverspeople-usuallythe elderly-who have
of the system. been forced to leave the IAMCsbecause of increases in
their health risk.
Uruguay provides extensive
coverage,
but needsto Both the public and privatesectorsof Uruguay'shealth
strengthen itsregulatoryframework systemuse copaymentsto recovercosts.Thus the IAMCs
are financedby a combinationof prepayments(in the form
Uruguay,with 3.2 millioninhabitants,has relativelygood of monthlypremiums) and copaymentsforservicesreceived.
socioeconomicindicatorsand incomeequality.In 1995the The public systemhas manylevelsof copaymentthat vary
under-fivemortalityrate was 21 per 1,000 (WorldBank with familyincome.

INNOVATIONS IN HEALTH CARE FINANCING

220
 The Ministryof Economyand Finance sets prices for
copaymentsand contributions,andhas avoidedsignificant
priceincreasesoverthe past ten years.As a result the oper-
ationaldeficitof the lAMCshas increasedsubstantially, and
somehave closed.Contributionsdo not vary by age, and
the IAMCsdo not receiveany risk adjustmentsbased on
their enrolledpopulations.As a resultmany youngpeople
voluntarilyleave the system (because they can purchase
cheaperplanselsewhere),whilethe elderlytryto retaincov-
erage.Thisimbalancehas been a majorfactorin the finan-
cialcrisisfacingthe IAMCs.
Youngpeoplehavetwo alternativesto the IAMCs.They
can enroll in a partial health insuranceplan offered by a
private institution (which generallycosts less than the
IAMCs),or they can seek servicesfrom the public sector.
Sincemanychoosethe latter,the number of people served
by the publicsectorhas beengrowing.Publicspendinghas
increasedand servicequality has deteriorated.Moreover,
an increasingnumber of people are covered simultane-
ouslyby two or three modalitiesof assistance.
Publicresourcesare also strainedby policiesregarding
the financingof high-technology services.Manyhigh-tech-
nologyservicesare coveredbythe publiclyfinancedNational
ResourceFund and are suppliedbythe Institutesof Highly
SpecializedMedicine.The cost of some of these services
(renaltransplants,forexample)is almostfivetimesthe price
chargedin neighboringcountries (such as Argentinaand
Brazil).In 1995the NationalResourceFund spent half of
its budget on high-technologyservices for about 8,000
people.
As in Chile,Uruguay'spluralistichealh systemhas ele-
ments of competition,but the regulatoryframeworkis not
sufficientlydevelopedto enhancecompetitionand choice.
There are many market failuresand problemsin the sys-
tem that put pressureson healthcare costs.Problemsarise
fromthe waycontributionsare controlled,the lack of mea-
sures to control adverse selectionby individualsand risk
selectionby insurers,the fact that dependentsare not cov-
eredunder a commonframework,andpoliciesonthe reim-
bursementof high-technologyprocedures.
The main suggestionsfor developinga more consistent
managedcompetitionframeworkare related to the trans-
formationof publicservicesand budgets.Publicproviders
could be organizedas autonomousnetworks,financedby
MANAGED CARE AND MANAGED COMPETITION
221
risk-adjustedpaymentformulas,that competeto coverlow-
income groups. Risk-adjustedcapitation formulas could
be usedto avoidthe cream-skimmingbehaviorofthe IAMCs.
The effortswould, of course,require some adjustmentof
Uruguay'srigidregulatoryframework.
Colombiahosmodeimpressive progress
in structuring
competition
in recent years
Colombiais in the middle tier of LatinAmerica'sdevelop-
ing countries,with a populationof 37 millionpeople, per
capitaincomeof $1,800,and the region'smost stableecon-
omy.Havingbeen sparedthe externalshocksexperienced
by most other LatinAmericannationsduringthe debt cri-
sis, and with no deep recessionsor inflationaryspirals,it
graduallymanagedto narrowpronouncedincomeinequal-
ity duringthe 1980s.Still,nearly30 percent of its people
survivedon lessthan $2.aday,and the country'ssocialsec-
tor institutionswere undeveloped.
Untilthe early1990shealthcaresystemswerehighlyseg-
mented, characterizedby heavy out-of-pocketspending.
Theneediestwereleftwithoutcoverage.Anambitiouspack-
ageof institutionaland financialreformslaunchedin 1993
soughtto implementa healthinsurancesystemwith a strong
element of competitionbut guided by principlesof soli-
darity.In the processColombiaproved itselfa pioneer in
its pursuit of managedcompetition.In the firstthree years
of the reformprogram,despitedauntingtechnicalandpolit-
icalproblems,coveragehas been vastlyexpanded,equity
has been increased,and the qualityof health serviceshas
been much improved.
In yearspast Colombiahadthree parallelhealthcare sys-
tems. A socialinsurancesystemrun by the SocialSecurity
Institute providedhealth servicesto formal sector work-
ers, and complementaryagenciessuppliedhealth services
to their familiesand to governmentemployees.Afterforty-
fiveyears,this publichealthcare systemserved21 percent
of the population-mainly middle-classcitizens,whowere
dissatisfiedwith servicequality.A traditionalpublichealth
systemoverseenby the Ministryof Health combinedpub-
lichealthinterventionsandopen-accesshospitals.Repeated
effortssincethe 1970sto decentralizepersonalmedicalcare
and emphasizeprimaryhealth care for the poor had borne
littlefruit, creatinga situationin whichthe poor had scant
IN LATIN AMERICA AND THE CARIBBEAN
access at the same time that hospital utilization rates were
very low (less than 50 percent).
Operating alongside the two public systems was a pri-
vate health care system, itselffar from uniform. Privatelyrun
hospitals accounted for 25 percent of hospital discharges
and surgeries. Some 75 percent of outpatients were attended
to by private doctors or practitioners. One million wealthy
Colombians had access to high-quality and expensive pre-
paid medical systemsand privateinsurancecompanies,whose
coverageessentiallyduplicated what was offered bythe social
insurance system. This segmented system cost a great deal
and generated sizable disparities between regions and
between segments of society.Health care expenditures were
enormous, equivalent in 1992 to about 6 percent of GDP,
and more than half came out of families'pockets. The coun-
try's infant mortality rate of 25 per 1,000 live births was an
average of the 15-20 per 1,000 rate in large cities and the
more than 200 per 1,000 in remote rural areas. At least 20
percent of the population-the poorest rural dwellers, res-
idents of marginal urban districts-had no access to health
care, a situation they attributed largely to their inability to
pay.The poorest 20 percent of Colombians paid about 18
percent of their income for health care.
Between 1990 and 1993 Colombia laid the institutional
foundations for a complete overhaul of this fragmented,
segmented, and inequitable health system. A 1991 consti-
tutional reform created the framework for decentralizing
social services and developing a social security system
grounded in the principles of universality, solidarity, effi-
ciency, and private sector involvement. In 1993 Congress
approved Law 60 and Law 100, the comerstone of a new
system made up of public finance, mandatory affiliation,
decentralization and competition in service delivery, con-
sumer choice, and broad participation.
At the heart of this new system was mandatory affilia-
tion, by those able to pay,with a social security systemguided
by a uniform resource mobilization and allocation scheme
and plurality of health service organizations. Thus finance
was separated from service delivery, and the purchaser-
provider split of services was developed.
The system's funding scheme is grounded in socialinsur-
ance principles. All participants contribute 11 percent of
their earnings to enroll their families. These resources are
placed in the SolidarityFund, a compensation fund for redis-
INNOVATIONS IN HEALTIH CARE FINANCING
222
tribution among population groups with differing economic
means and health needs. The fund assigns to the organiza-
tion chosen by each family a capitation payment that is
risk-adjusted for sex, age, and geographic location, with
reinsurance for catastrophic illness.The average capitation
payment is $120 per person per year.In addition to the basic
contribution, there is a copayment system to encourage
rational use of services. Public funds do not finance the
Solidarity Fund; they are gradually being turned to directly
fund health insurance for the poorest 30 percent of
Colombians.
At the core of health care reform is a new system of
multiple organizations in charge of enrollment (by delega-
tion of the Solidarity Fund), insurance, and organization
of service delivery.Health Promotion Organizations (EPSs)
combine insurer functions with the type of service articu-
lation performed by U.S. HMOs. Without privatizing the
SocialSecurityInstitute, the reforms divested it of its monop-
olyin this area. They allow unrestricted competition among
EPSs-public or private, nonprofit or for-profit, integrated
delivery or by contract-subject to certain minimum finan-
cial conditions. Families can freely choose their EPS. Users
are guaranteed a universal package of high-quality health
services, including medicines. One aim of this arrange-
ment is to stimulate competition for service delivery and
spur the development of supplementary plans for inter-
ventions not covered under the universal package.
The new social security systemseeks to integrate the pub-
lic health care delivery system. Government budget out-
lays (bolstered by a "solidarity payment" of 1 percent of
contributions) are being strictly targeted to public health
interventions or to the poorest 30 percent of Colombians.
Public hospitals have become autonomous corporations,
akin to state industrial enterprises. Regional agencies now
purchase services on behalf of disadvantaged groups.
Growing volumes of direct transfers from the national gov-
emient to municipalities are intended to fund insurance
for the neediest. To that end, a system of community coop-
erative healh organizations has been actively promoted in
the poorest areas to purchase basic health care services or
to allow people to join an EPS.
In tandem with this separation of the system into three
streams of funding, servicepurchase, and serviceproduction,
the Ministry of Health, freed of direct responsibility for pro-
viding services,is now responsible for general oversight.The
ministryis aidedby a nationalboard representingbusiness,
labor,andhealthserviceusers.Thistripartitebodyhas broad
powersto developrules and regulationsunder the law.In
addition,the now-autonomousHealth Superintendency
has been strengthenedto better performits financialcontrol
functionsand supervisethe workingsof the system.
Thus Colombia'snew health systemcombinesthe two
preeminentprinciplesof LatinAmerica'sversionof man-
agedcompetition:coordinationof servicedeliveryin inte-
grated plans and consumerempowerment.At the core of
the systemare the EPSs, which assumethe risk of ensur-
ing a universal packageof services.The demand side is
bolstered by users' abilityto freelychoose their EPS, by
consumerorganizationthroughhealth alliancesand coop-
erative health organizations,and by government media-
tion in healthcarefunding(withrisk adjustmentsfor groups
that cannot pay, and direct governmentsubsidiesfor the
poor) and regulation.
In the three yearssince Colombiaadoptedthe new sys-
tem, enormousprogresshas beenmadein its development.
More than fifty decrees havebeen issued to regulate the
systemand fine-tuneits variouselements.The Ministryof
Healthhas madelocalagenciesresponsibleforservicedeliv-
eryin 80 percentof the country,and 85percent of hospitals
are nowrun autonomously. Withthe newfundingsystemin
place,nearly20 milion people-55 percent of the popula-
tion-now has healthinsurancecoverage.Some 14milion
peoplehavechosento joinone of thirtyEPSs,justover half
of themfallingunderthe SocialSecurityInstitute.The new
EPSshavean impressive recordof innovationin serviceorga-
nization,informationsystems,contracting,payments,and
qualityassurance.SixmiDlion of Colombia'spoorest have
receivedvouchersgivingthemaccessto the system:fourmil-
lionthroughEPSsandtwomilionthroughcooperativehealth
organizations.Accordingto a surveyconductedin early1997,
83percent of EPS memberswere satisfiedor verysatisfied
with the servicestheywere receiving,and 73percent of the
populationfelt that the healthsystemhad improvedsince
the enactmentof Law 100.
Colombia's mix of public funding and managedcom-
petition has substantiallyexpanded coverageand made
the systemmore equitable.Still,the transitionhas had its
share of technical,institutional,and politicalproblems:
MANAGEDCAREAND MANAGEDCOMPETITIONIN LATINAMERICAANDTHE CARIBBEAN
223
Growing violence, resulting in 30,000 homicides a year
(more than in the United States, which has ten times
the population),is addingto the burden of disease.The
associatedskyrocketingdemandformedicalcareis gen-
eratingadditionalcostsof more than 1 percent of GDP.
* The processof identifyingbasicparametersfor the new
systemhas been one of trial and error.The capitation
rate,forinstance,mustbalanceout systemrevenueswhile
defrayingthe cost of universal service.Difficultiesin
measuringcostsand forecastingrevenuesof newmem-
bers have prompted more changesin these areas than
wouldhavebeen desirable.
* The absenceof strong politicalresolveto curb evasion
has kept systemrevenuesdown.
* The transformationfrom a systemthat allocatedpublic
fundsto onebasedon demandfnancinghas takenlonger
than expected,drivingup costsunnecessarily.
* Not allthe adjustmetitsto the universalhealthpackage
havebeen based on sound cost-effectivenesscriteria.
* In the absenceof fullydevelopedinformationsystems,
the allocationof publicresourcesandconditionsfor con-
sumerchoiceare still not fullytransparent.
* Politicalproblemsstil arise. One result of the govern-
ment's politicalcrisiswas a successionof three health
ministersin lessthan threeyears,leavingthe systemwith-
out strongdirectionat criticaljunctures.Andthe shiftin
the balanceof powerbetweenusers,providers,organiz-
ers, and fund holders has created friction, sometimes
open and sometimesnot, that has hampereddecision-
makingand tested providers'supportfor the system.
In the final analysis,Colombia'sinnovativereformsto
introducecompetitionintoitshealthcaresystemhaveyielded
resultsfar more quicklythan in other countries.The chal-
lenge in the years ahead will be to consolidatethe gains
and step up the pace of the transition,to givethe country
a more efficientand equitablesystemoverall.
Costa Rica hasachieved
someelements ofmanagedcare
The CostaRicangovernmentcreatedthe CostaRica Social
SecurityFund(CCSS)in 1941to providesicknessand mater-
nity coverageand pension funds for urban salariedwork-
ers. In 1961thismandatewasexpandedto coverthe entire
populationwithinten years,but politicaldisputesdelayed
implementationten years.In 1971the CCSSwasrequired
by lawto proceedwith the universalizationof coverage.
This goalwas achievedin 1973with the enactmentof a
lawthat transferredall Ministryof Health hospitalsto the
CCSS and entrustedthe CCSSwith providingfree health
care to the poor.That sameyear,the NationalHealth Plan
defned the scopeof work for the Ministryof Health and
the CCSS. The ministrybecame responsibleforproviding
preventivehealth servicesand the CCSS for curativeser-
vices.The socialsecuritysystem'sextensivecoveragehas
left little room for the private sector. (Both the ministry
and the CCSS serve patients who are uninsured and do
not deny services to anyone, regardless of income.) A
1982-85 survey,for example,found that the privatesector
managedjust 1.9percent of hospitalbeds.
CostaRica'seconomy,like manyothersin LatinAmerica
andthe Caribbean,sufferedconsiderably duringthe economic
crisisof the early1980s.In 1981inflationwas81 percentand
the currencywas devaluedby320percent,andduring1982-83
unemploymentreached9.4 percent.The country'scapacity
to payits externaldebt was underminedas debt payments
started to consumemore than half of exportearnings.The
government'scapacityto collecttaxeswasreduced,andpub-
licexpenditureswerecut.Asaresultthe NationalHealthPlan
wasrevisedto redefinethe legalrolesoftheMinistryofHealth
and the CCSS, to promoteadministrativedecentralization,
and to createaltemativemodelsof healthdelivery.
During the 1980sthe governmentexperimentedwith
transferringpart of the health deliverysystemto the pri-
vate sector.In particular,the CCSS developedthree pro-
grams-enterprisephysicians,mixedmedicine,andmedical
cooperatives-to incorporatethe private sector in service
delivery.Thesemeasuressoughtto reducecostsandimprove
serviceefficiencyand quality.
Under the enterprisephysicianprogram,private enter-
prises (630 to date) hire a doctor and provide him or her
with an officeand a nurse. Under the mixedmedicinepro-
gram,the insured choosesa doctor from amongmembers
of the programandpaysfor the service.Bothprogramsare
supported by the CCSS. During 1986-87 these two pro-
gramsreacheda combined7 percentof the populationcov-
eredbythe CCSSwhileconsumingonly4percentof x-rays,
3 percent of laboratorytests, and 6 percent of approved
paid sickleave.
INNOVATIONS IN HEALTH
224
Under the medical cooperative program, the CCSS
signeda contractwith twohealthcooperatives,makingthem
responsiblefortwoclinicsin the SanJosemetropolitanarea.
Under one of these agreements,the CCSSconstructedand
delivereda $1.5millionclinicto COPPESALUD,a physi-
cian'scooperative,whichbecameresponsibleforoutpatient
care in the Pavasdistrict.The CCSSpaysabout $30 a year
foreachinsuredresidentin the area.COPPESALUDhealth
servicesare deliveredby a basic health careteam madeup
of a physician,an assistant,and a communityhealthworker
foreach3,000-3,500persons.Staffturnoverislowbecause
salariesare high.
Although Costa Rica's health system does not have
enough insurersto create competitivemarket conditions,
and the privatesectoris small,the three privatehealthpro-
gramsdo contain elementsof managedcare. In the enter-
prise physicianprogram,doctors haveto compete for the
salariedpositionswithincompanies.Like the general and
familyagentsof the primarycare network,enterprisedoc-
torsserveasgatekeepersto specializedservicesand as agents
of cost control.Under the mixed medicine program,the
insuredmust selecta doctor fromthe list of physicianspro-
vided by the insuranceorganizationin whichhe or she is
enrolled.This approachcan save moneyfor the systemif
the insured signs up with the most economicalproviders.
Finally,the medicalcooperativeprogramrepresentsa type
of capitationmodelin whicha group of physiciansprovide
health care servicesin return for a fixed per capita pay-
ment providedby the government.
Given that Costa Rica'ssmallpopulationimpedes the
developmentof competitivehealth plans, future innova-
tions could come from increasingchoicefor beneficiaries
of the FamilyHealth Program,using more managedcare
techniquesin primarycarecentersandpublichospitals,and
increasing transparency and efficiencyin the system of
resourceallocation.
The Dominican
Republic
is movingtowardgreaterprivote
provision
With 7.5 millioninhabitants and a per capita income of
about $1,000,the DominicanRepublicspends about 5.3
percent of its GDP on health (WorldBank forthcoming).
Publicspendingaccountsfor38percentofthe total.Despite
CARE FINANCING
the lack of in-depth studies on servicecoverage,it is esti-
mated (basedon a studyconductedin the capital)that 56
percent of the populationis coveredby the private sector,
30 percent by the Sub-Secretariatof Public Health and
SocialAssistance(SESPAS),10 percentby the Dominican
Institute of SocialSecurity,and 4 percent by the Institute
of SocialSecurityof the Armyand PoliceForce.
SESPASplaysa contradictoryrolesince it regulatesand
inspectsallpublicand privatehealthserviceswhilealsopro-
vidingpubliclyfinancedhealth servicesthroughits 57 hos-
pitals, 79 health posts, and 587 rural clinicsand medical
dispensaries.The DominicanInstitute of SocialSecurity
managesa maternityhospital,20 polyclinics,and 13urban
and 128 rural medical offices.Its financingcomes from
voluntarycontributionsto socialsecurityaccountsandtrans-
fers from the federalgovernment.It chargesa fee for some
servicesto partiallyrecovercosts.
The private sectoris composedof both nonprofitinsti-
tutions(suchas NGOs) andfor-profitorganizations.Over
the past twentyyears the organizationof private services
has changed as coveragestrategiesand health plans simi-
lar to those of U.S. HMOs have been adopted. Although
it controlsjust one-quarterof the country'shospitalbeds,
the private sector accounts for more than half of outpa-
tient visits.
The economicand institutional crisis experiencedin
the 1980screatedproblemsfor the DominicanRepublic's
publichealthsystem.Since1985the government,with assis-
tance fromthe U.S.Agencyfor InternationalDevelopment
(USAID),hasbeendevelopingaltemativefinancingschemes
and extending private coverage to low-income families
and workersin the informalsector (La Forgia1990).
There are atleastthreekindsof privateinsurance:Igualas
Medicas(prepaid HMO-type health plans), health insur-
ers, and self-administeredinsurers.The twenty-oneIgualas
Medicasadministerthe organization,financing,and pro-
visionof healthservices.Most have their own clinics.
The healthinsurersare specializeddepartmentsof insur-
ance companiesorganizedto coverhealth riskswithin an
integratedinsurancepackage.In general,theycontractwith
medicalclinicsfor serviceprovision.The insurersprovide
two typesof services:those based on a systemof preferred
providers(where the user has limited choices)and a sys-
tem offering freedom of choice within establishedprice
MANAGED CARE AND MANAGED COMPETITION
225
parameters.The secondschemeis usuallymore expensive
and has a more sophisticatedsystemof control.
Finally,the self-administeredinsurersarenonprofitorga-
nizationsestablishedby institutions,companies,and unions
to managehealth insuranceplans for their specificpopu-
lations.Tensuchinsurersare operating,the largestof which
is the MedicalInsuranceof Teachers,with about 120,000
enrollees.
The Igualas Medicas are the fastest growingof these
three typesof insurers.With the backingof USAID,these
enterpriseshavereceivedtechnicalassistancewith admin-
istration,costcontrol,qualityassurance,andactuarialcapac-
ity.Althoughthey are now tryingto expand their services
to low-incomegroups,they are doingso withoutthe finan-
cial backing of the state, which continues to provide its
own healthservices.
Despitethe institutionalformsof prepaymentdeveloped
in the DominicanRepublic,most insurerslack basic tech-
nicaland administrativeexpertisein terms of information
systems,methodologiesfor establishingrisk,cost controls,
andcost-efficientmethodsof allocatingresources(Santana
1996). They alsolack adequate mechanismsfor respond-
ingto consumers'complaintsand comments.
Despitethese deficiencies,the healthservicesprovided
by the private sector are relativelycomprehensiveand of
higher quality than those provided by the public sector.
Privateinsurersadmit familiesand all categoriesof work-
ers, and theirbeneficiarieshaveaccessto better clinicsand
services.Their units also offer greater choiceof doctors
and more personalizedattention.
One frequent criticismof the privatesystemis that its
packageofhealthcoverageexcludesmanyservicesandhigh-
costpatients.But thiscriticismshouldnot overshadowthe
fact that the IgualasMedicasare aviableoptionfor improv-
inghealthservicesand eventuallyreformingthe Dominican
Republic'ssocialsecuritysystems.
Further developmentof managed competition in the
DominicanRepublicrequires the developmentof a regu-
latory frameworkthat increasesthe accountabilityof the
IgualasMedicaswhilefosteringcompetitionwith SESPAS
and the Institute for SocialSecurity.Consumerempower-
ment can be increasedby developingan essentialpackage
of servicesand allocatingpublicresourcesto subsidizethe
poor's accessto the competitiveintegrateddeliverysystems.
IN LATIN AMERICA AND THE CARIBBEAN
Brazil'sthree systems shouldbe betterintegrated Despite a pluralisticstructure of serviceproviders,the
Brazilianhealth model lackswell-definedrules separating
Brazil'shealthcare systemis composedof a publicsystem, the three systems.As a result there are conflictsamongthe
a privatesystemof supplementarymedicine,and an out- system'sparticipants,due both to the lackof adequateleg-
of-pocketsystem(table3).The resourcesgatheredthrough islationand to the constitutionalprinciplesguaranteeing
thesethreesystemstotaled$23.2billionin 1994, withexpen- the universalright to free and comprehensivehealth ser-
dituresequivalentto 4.5 percent of GDP vices. Low-incomecitizensaccount for 78 percent of the
The public system-known as the Unified Health public system'susers, yet must share the systemwith the
System-was establishedby the 1988Constitutionand is 22 percentof peoplewho are alsocoveredby privateinsur-
financedby federal and local governments.In 1994fed- ance.Otherwise,the poor haveto pay for medicalservices
eral healthexpendituresreached$10.4billion,whilethose out of pocket.
of states and municipalitiesreached $4.1billion. Most of Since the public sector is not organizedon a competi-
the resourceslocalgovernmentsspend on health aretrans- tive basis, servicesare essentiallyrationed-as evidenced
ferred to them by the federal government.In exchange, bylonghospitallinesand lackof materialsand medication.
localgovernmentsare responsiblefor the directprovision And becausethere are no barriersblockinguse of the pub-
of servicesor for contractingout servicesto private estab- licsystem,peoplecoveredby privatehealth insurancetend
lishments. to use the most sophisticatedservicesand examsprovided
The private system consistsof four types of organiza- by the public sector. Thus the government is indirectly
tions: medicalgroup organizations(MGOs) that operate subsidizingprivatehealth insurance.
like U.S. HMOs, health insuranceinstitutionsthat func- A fewinnovativeexperimentsare under wayto reverse
tionundersimilarcriteriabut do notprovideservicesdirectly, this situationat the local level,where states and munici-
medicalcooperativesthat use prepaymentsystems(though palitiesare assumingincreasingautonomyin the adminis-
their operationsresemblethoseof preferredproviderorga- trationofhealthservices.One suchexperimentis the Health
nizations),and medical servicesprovided by companies, Plan of Actionintroduced in 1995in Sao Paulo,Brazil's
whichmaybe withintheir facilitiesor contractedfrom any largestcity (15 millioninhabitants).
of the other three organizations.The private systemrelies The goalof the Health Plan of Actionis to organizeSao
on prepayments,coversabout 35millionBrazilians,andhas Paulo's municipalservices(includinghospitalsand basic
annual revenuesaveraging$6.6 billion. health services)into cooperatives (managed by doctors
Theout-of-pocketsystem-meaninghouseholdpayments and staff) that competefor clientele.Many such coopera-
for servicesprovidedthroughmedicalnetworksand private tiveshave been organized. Doctors and staff have been
hospitals-has annualrevenuesof $2.1biDlion. The number transferred from the human resourcesdivisionof the city
ofpeoplecoveredunderthissystemisresidual,sinceresources governmentto administrativeand managementpositions
from the publicand privatesystemsare the mainsourceof in hospitalsand otherhealthinstitutions.Under thisscheme
financingforhospitals,healthservices,and doctors. citizenscan choose the cooperativethat best meets their
TABLE3
Health care servicesin Brazil
Publicsystem
Group (UnifiedHealthSystem) Privatesupplementary
medicine Direct out-of-pocket
payments
Informalsectorworkers/low income Primarycareand None ComplementUnifiedHealthSystem
hospitalization
Formalsectorworkers/middleincome Hightechnologyand Primarycare,hightechnologyand Not usedmuch
sophisticated
procedures sophisticated
procedures
Highincome Hightechnologyand Primarycare,hightechnologyand Usedheavily
sophisticated
procedures sophisticated
procedures

INNOVATIONS
IN HEALTHCAREFINANCING
226
needs,andthe municipalgovernmentmakesan annualpay- cent for each dependent)and3-6 percent paid byemploy-
ment of about $225per capita to each cooperative. ers. Taxable income is capped at $3,750 a month. The
Experiencewith the Health Plan of Actionis limitedto nationalObra Socialsystemcovers10millionbeneficiaries
relativelyuncomplicatedservicesin Sao Paulo.Publichos- in about 300 Obras. In addition to the national system,
pitalsareresponsibleformorecomplexservices.TheHealth twenty-fourprovincialObras cover5 millionpublicemploy-
Plan of Actionhas four main weaknesses: ees and their dependents (WorldBank 1996).
* Lack of incentivesto provide preventivehealth care to About 4 millionretired, disabled, and pensioned per-
plan members. sons and their families are covered by Integral Medical
* Limitsto competition,since the cooperativesare orga- AttentionforRetirees,whichis fundedby an additionalpay-
nized into geographicallyseparatenetworks. rolltaxof 5 percentthat is sharedbyemployersand employ-
* Ablurredrelationshipbetweenthe HealthPlan ofAction ees.Altogether,this socialinsurancenetworkcoversnearly
andthe UnifiedHealthSystem,whichcreatescross-sub- 20millionArgentines,or about61percentofthe population.
sidiesbetweenthe two systemsandpreventsclearanaly- There is also a large market for private health insur-
sis of the Health Plan of Action. ance.About 200 private plans cover more than 2 million
* The percapitapaymentsthat the citymakesto the Health people.Another1 millionbelongto nonprofitmutualinsur-
Plan of Action could be used to foster competition ance funds. Coverageis often duplicated,however.Many
between public cooperativesand private health struc- workerswhocontributeto the nationalObras Socialesalso
tures. buy voluntaryhealth insurancebecauseof dissatisfaction
Despite these problems,80 percent of users are satis- with the servicesprovidedby the Obras.In addition,some
fied with the Health Plan of Action.The challengeis to employersprovide privatehealth insurancein additionto
givecontinuityto the programandto better defineits rela- the mandated Obra Socialcoverage.People that are not
tionshipwith the other healthsystems. coveredby either the Obras Socialesor by private insur-
Brazilcanlearn a lot from the varietiesof managedcare ance-mainly the poor-rely on the publiclyfinancedpub-
and managedcompetitionin the region.The government lichospitalsystemand on out-of-pocketpurchaseof health
could developa systemof competitivepurchasersof health services.
careand simplifythe systemof allocatingpublicresources. The health deliverysystemis mixed.About half of hos-
The public systemof providerscould be transformedinto pital beds are in privateinstitutions.Althoughsomeof the
integrateddeliverysystemsthat compete withprivate net- Obras Socialesprovidehealthservicesdirectlythroughtheir
works.A commonsystemof regulatingpublic and private ownfacilities,mostpurchaseservicesfroma networkof pri-
purchaserscould then be developed,with attention to a vate providers.Because accessto providersis restrictedto
comparablepackageof essentialservicesand a systemof those within the network,many Obras Socialesresemble
quality assurance.Within this frameworkthere is ample preferred providerorganizations.Over the past few years
spacefor the applicationof managedcare techniques,espe- the Obras Socialeshave moved awayfrom fee-for-service
dailyin ambulatoryservices.Brazil,however,hasnot reached providerpaymentstoward capitatedpayments.
consensusabout future reforms. Still,the systemisfacingfinancialdifficulties.
Thenational
and provincialObras Sociales(indudingIntegralMedical
Argentina's systemhasweaknesses-but the government AttentionforRetirees)haveincurredhuge deficitsin recent
istryingto address them years.Moreover,consumerdissatisfactionwith the Obras
Socialesis widespread.Except for white-collarworkers,
In Argentinamost formalsectorworkersand their depen- until very recentlyworkers have not been able to choose
dents are required to participatein an Obra Social(statu- their Obra Social-it is dictatedby their employment.The
tory sicknessfund) linked to their place of employment. Obras Socialesalsohavehuge differencesin both per capita
The national Obras Socialesare funded by a compulsory funding and performance.Some workers receivealmost
payrolltax of 3 percent paid by employees(plus 1.5 per- no healthcare servicesfrom their Obras. Others belongto

MANAGED CARE AND MANAGED COMPETITION IN LATIN AMERICA AND THE CARIBBEAN

227
Obras that provide comprehensiveservices.Many Obras
Socialesare probablytoo smallto be economically efficient.
Manyhaveweak managersand are overstaffed.
The provinces,whichare responsiblefor almostallpub-
lic hospitals,alsohave experiencedsizablefiscaldeficits.
Equity and efficiencyare both big problems.Yetthe gov-
ernment's approachto reformin one part of the system-
the nationalObrasSociales-suggeststhat it recognizesand
is respondingto at least some of these challenges.
The governmentis nowestablishinga regulatoryframe-
workthat supportsa competitiveenvironmentforthe Obras
Sociales.The envisionedframeworkcontainsmanyelements
of managedcompetition.Itwouldguaranteestandardhealth
benefitpackages,supportconsumerchoice,improveequity,
and minimizeriskselectionbyworkersand Obras Sociales.
Manycomponentsofthe reformshavealreadybeenimple-
mented.The governmenthas signeddecreesallowingwork-
ersto choosetheirObra Socialandto taketheircontributions
withthem.A preliminarystandardbenefitspackagehasbeen
defined.The formulasfor redistributingObra Socialcon-
tributionsfrom high-incometo low-incomeworkerswere
overhauledto make the systemmore accountable,trans-
parent, and equitable.Efforts are being made to develop
risk-adjustedpaymentsforthe Obras Socialesbasedon the
riskprofilesof thosewho enroll.Obras are restructuringto
cut staff and modernizetheir operations.Finally,steps are
beingtakento strengthenregulatoryoversightofthe Obras.
Designingand implementingthese reformsis difficult,
anddependson both politicaland technicalfactors.Design
and implementationis made more difficultby the weak
informationbase that supports the Obras Sociales.The
Obras Socialesenrollmentdatabase,for example,is incom-
pleteandinaccurate.Thenumberof dependentsis not dear.
Data to support risk-adjustedpayments(age, sex, health
risks)arelimited.Andmuchworkwillbe neededto develop
measuresofObra Socialqualityto supportconsumerchoice.
The consolidationof managedcompetitionin Argentina
willrequireimprovingthe systemofpayments(includingrisk
adjustment)to integrateddeliverysystemsandincreasingthe
scopeof choicefor consumers.Obras Socialeswillneedto
becomemoreaccountableto theirconsumersandfacemore
competitionwith privateproviders.Over the long run the
poor couldbe broughtinto the systembyredirectingpublic
resourcestowardplansthat providethemwith coverage.
INNOVATIONS IN HEALTH CARE FINANCING
228
Conclusion
Managedcare modelsof health servicedeliverynow make
up a significantshare of the health insurance market in
severalLatinAmericanand Caribbeancountries.Although
these modelsshare manyfeatureswith managedcare pro-
gramsin the United States-the countrywith the largest
share of the population(57 percent) enrolledin managed
care in the world-they have distinct featuresthat reflect
theirpoliticaland socioeconomiccontexts.Theirgrowthis
likelyto continueoverthe nextdecadegivenboth the growth
in demandfor private insuranceand the reformsthat sev-
eral governmentsare supportingto promote competition
in healthinsurancemarkets.
The concepts of managedcare and managedcompeti-
tion have found fertile soil in Latin America and the
Caribbean.Colombiaand Uruguayhave gone furthest in
introducingvarietiesof managedcompetition,with about
half of their populationscovered.In Argentina,Chile, and
southern Brazil new institutional modalities have been
extended to cover about a quarter of the population. In
CostaRicaandsomepartsofthe English-speaking Caribbean
the introductionof neworganizationalmodelsis stillin the
experimentalphase.
A country's abilityto introduce managedcompetition
depends on its characteristicsand institutional features.
Three basic determinantsof managedcompetition'ssuc-
cess includea country'ssize,its levelof development,and
the way its health system is organized. The smaller and
poorera countryis, and the less developedits institutional
capacity,the more trouble it willhavemanagingcompeti-
tion. The achievementsof Uruguay(a smallcountry)and
Colombia(a middle-incomecountrywith a highlyhetero-
geneouspopulation)do not followthese generalguidelines,
however.
As noted, the thirty-three countries of Latin America
and the Caribbeanshowconsiderablediversity.At one end
of the scaleare elevenmiddle-incomecountrieswith fewer
than 1 million inhabitants (mainly English-speaking
Caribbeancountries)and highlevelsof coveragefinanced
and operated by the public sector.These are followedby
sevencountrieswith 1-5 millioninhabitantsand highpop-
ulation densities (for example, Costa Rica, El Salvador,
and Jamaica). In these countries discussionsof internal
TABLE4
Trendsin LatinAmerican health care markets
Higher-incomecountries
Smallhomogeneous
countries Managedcarein publicsystems
Largeheterogeneous
countries Managedcompetitionwith a mix of publicand
privatefundsand institutions
markets and opportunities for managed care are proba-
bly more relevant for primary care, and must be com-
bined with incentive systems that improve the public
administration of services. At the other end of the scale
are the large Southern cone countries (including south-
em Brazil), which have achieved considerable economic
and institutional development. Given these countries' more
developed markets, the integration and plurality of their
service providers, and their health care experiments over
the past decade, full competition between integrated deliv-
ery systems, with public financing, has the greatest poten-
tial. Rather than encouraging development of separate
private and public systems, innovations should promote
convergence among existing systems in order to mini-
mize duplication of coverage, control costs, and improve
the quality of services.
In other large but poor countries (Bolivia, Guatemala,
Haiti, Honduras) the empowerment of consumers would
be more effective if the state redirected its efforts from
providing the middle class with public hospital services
toward funding and purchasing primary health care services
for the poor.To that end, communityorganizationscould
forythe por.eT org
acthat roleind strucommurinimarety ions btion
cd
play a more active role in structuring markets from the begin-
ning, drawing on some of the ideas of sponsors presented
by Enthoven (in this volume) or the community financing
schemes identified by Hsiao (1992). The experiences of
Colombia's Empresas Solidarias and Peru's community devel-
opment cooperatives are particularlyrelevant inthis regard.
Other countries have more options for developing health
care systems with greater competition, freedom of choice,
and solidarity.These indude the larger,higher-incomecoun-
tries with good institutional development (Mexico,
Venezuela)in which health reforms aimed at managing com-
petition
would be highly effective. There are also smaller
petition would be highly effective. There are also smaller
countries withmore diversifiedsocialand institutional struc-
tures (the Dominican Republic, Ecuador, Peru), where an
MANAGED CARE AND MANAGED COMPETITION IN LATIN AMERICA AND THE CARIBBEAN
229
Lower-incomecountries
Competitionfor primaryhealthcarewith publicfundsfor the poor
Structuredpluralismwith a strongerelementof publicfinance
active configuration of managed competition could help
overcome the segmentation of existing systems,expand cov-
erage, and improve the quality of services (table 4).
Latin America and the Caribbean lag far behind the rest
of the world in terms of education and health indicators.
Catching up over the next decade will require more than
increased financial and human resources. Above all, it will
require redoubling the pace and scope of institutional inno-
vations that the region has experimented with during the
1990s in order to use additional resources more efficiently
and equitably.The challenges are particularly great in coun-
tries with low levels of public sector management capacity.
The methods described here under the aegis of managed
competition hold great promise in this respect. A long road
must still be traveled to define the new models for the health
sector. Managed competition systems will play a major role
in this process.
Notes
1. Londofioand Frenk (1997) suggesta related concept of
structuredcompetitionto characterizea systemin whichcompeti-
is structured with an explicit and common set of rules that
aelglyvldfraldlvr ytm n pnos hscn
are legally valid for all delivery systermsand sponsors. This con-
cept differsfrommanagedcompetition,in whichdifferentspon-
sors can manage, withmore discretion,differentrules (see Enthoven
in this volume).
2. Of course, managed competition is not the only type of reg-
ulationpossible.Regulationis often used to restrictcompetition.
References
AmericanAssociationofHealthPlans.1995.HMOandPPO Trends
Report.Washington,D.C.
Arrow,Kenneth. 1963. "Uncertaintyand the WelfareEconomics
of MedicalCare."American Economic Review 53 (5).
Birdsall, Nancy. 1993. "Pragmatism,Robin Wood and Other
Themes: Good Government and Social Well-Being in
 Developing Countries." In L.C. Chen,A. KNeinrman, and N.C.
Ware, eds., Health and Social Change in an International
Perspective.Carnbridge,Mass.:Harvard Schoolof Public Health.
Bobadilla,Jose Luis. 1996. "SearchingforEssentialHealth Services
in Low and Middle Income Countries: A Review of Recent
Studies on Health Priorities." World Bank, Human
Development Department, Washington, D.C.
Bobadilla, Jose Luis, Peter Cowley, Philip Musgrove, and Helen
Saxenian. 1994. "Design, Content and Financingof an Essential
National Package of Health Services." Bulletin of the World
Health Organization72: 653-62.
Bobadilla, Jose Luis, Julio Frenk, Peter Cowley, Beatriz Zurita,
Juan Querol, EmestiVilarreal, and Rafael Lozano, 1994. "El
Paquete Universal de Servicios Esenciales de Salud." In Julio
Frenk, ed., Economia y Salud: DocumentosparaelAnalisisy la
Convergencia.Mexico,D.F.: Fundaci6n Mexicana para la Salud.
Enthoven, Alain. 1993. "The History and Principles of Managed
Competition." Health Affairs 12 (supplement): 24-48
- 1996. "Managed Competition: An Agenda for Action."
Health Affairs (spring).
Enthoven, Alain,and SaraJ. Singer.1996a. "Managed Competition
and California's Health Care Economy." HealthAffairs 15 (1):
39-57.
- 1996b. "Market Based Reform: What to Regulate and
Whom." Health Affairs (spring).
Fiedler, J. L. 1996. "The Privatization of Health Care in Three
Latin American Social Security Systems." Health Policy and
Planning 11 (4): 406-17.
Frenk, Julio. 1995. "Comprehensive Policy Analysis for Health
System Reform." Health Policy 32: 257-77.
Govindaraj, Rarnesh,Christopher Murray,and Gnanaraj Chellaraj.
1995. Health Expenditures in Latin America. World Bank
Technical Paper 274. Washington, D.C.
Hsiao, W C. 1992. "Cooperative Health Care: A Comprehensive
Strategy to Finance Health Care for Developing Nations." In
Proceedingsfrom the World Health Organization Leadership
Conference.Geneva: World Health Organization.
IDB (Inter-American Development Bank). 1996. Economic and
SocialProgressin LatinAmerica: 1996Report.Washington, D.C.
Janovsky,Katja, ed. 1996. Health Policyand SystemsDevelopment:
An Agendafor Research.Geneva: World Health Organization.
Katz, J., and J. Laspiur. 1993. "Reflexiones Comparativas sobre
el Modelo de Organizaci6ny Comportamientodel Sector Salud
en la Argentina, Chile y Uruguay" In OPS, ed., Estructuray
Comportamientodel SectorSaluden laArgentina,Chiley Uruguay.
Cuaderno Tecnico 36. Washington, D.C.
Katz,J.,J. Laspiur,and H. Arce. 1993."UnModelo de Organizaci6n
y Comportamiento Sectorial." In OPS, ed., Estructura y
ComportamientodelSectorSalud enlaArgentina,Chiley Uruguay.
Cuademo Tecnico 36. Washington, D.C.
Labadie, G., and D. Sanchez. 1993. "El Sector Salud en Uruguay"
In OPS, ed., Estructuray Comportamientodel SectorSalud en
INNOVATIONS
IN HEALTHCAREFINANCING
230
laAi.gentina,Chiley Uruguay.CuadernoTecnico36.Washington,
D.C.
La Forgia, G. M. 1990. Health Servicesfor Low-Income Families:
ExtendingCoverageThroughPrepaymentPlansin the Dominican
Republic. Washington, D.C.: The Urban Institute.
- .1995. "UnderstandingHealthSectorReform:AFinancial
and Service Flow Model and the Colombian Case." Inter-
American Development Bank, Social Programs Division 2,
Washington, D.C.
Larrafiaga,O., R. Miranda, and R.D.P Molina. 1996. "Organizaci6n
Industrial de los Servicios Sociales de Salud en Chile." Paper
commissionedbyInter-American Development Bank, Santiago.
Lewis, Maureen A. 1996. "Measuring Public Hospital Costs:
Empirical Evidence from the Dominican Republic." Social
Science and Medicine 43 (2): 221-34.
Lewis,Maureen A., andAndre Cezar Medici. 1995. "PrivatePayers
of Health Carein Brazil:Characteristics,Costs, and Coverage."
Health Policyand Planning 10 (4): 362-75.
. 1996. "The Health Reform in Brazil: Phasing Changes."
Inter-American Development Bank, Social Programs Division
1,Washington, D.C.
Londotio, Juan Luis. 1996a. "Estructurando Pluralismo en los
Servicios de Salud: La Experiencia Colombiana." Revista de
Andlisis Econ6mico. 11 (2). Santiago de Chile.
. 1996b. "Is There a Health Gap in Latin America? A
Graphical Exposition." World Bank, Latin American and the
Caribbean Region Technical Department, Washington, D.C.
. 1996c. "Managed Competition in the Tropics?" Paper
presented at the International Health Economics Association
Inaugural Conference, May,Vancouver.
Londofho,Juan Luis, andJulio Frenk. 1997. "Structured Pluralism:
A New Model for Health Reform in Latin America." Health
Policy (JUly).
Londofio, Juan Luis, and 0. Icoechea. 1996. "Latin American
Health Care Systems: A Simple Description of its Financing
and Organization." World Bank, Latin American and the
Caribbean Region Technical Department, Washington, D.C.
Luft, H. S. 1996. "Modifying Managed Competition to Address
Cost and Quality." Health Affairs 15 (1): 23-38.
Mazzafiero, VE. 1996. "Health Conditions in Argentina: Facing
the Twenty-First Century." Journal of Public Health Medicine
18 (2): 234-37.
Medici, A. C. 1994. "Economia e Financiamento do Secor Saude
no Brazil."AdSaide S6rieTematica,Faculdadde Sa6dePblica,
Universidade de Sao Paulo, Sao Paulo.
. 1996. "Opcoes para Reformar o Financiamento dos
Servicosde Saude." In OPS, ed., ReunionEspecialsobreReforma
del SectorSalud. Washington, D.C.
Medici, A. C, and C. Czapski. 1995. "Evolucao e Perspectivas
dos Gastos Publicos com Sauldeno Brasil." World Bank, Latin
America and the Caribbean Department 1, Human Resources
Division, Washington, D.C.
Mills, A. 1993. "Economic Aspects of Health Insurance." In K.
Lee and A. Mills, eds., The Economicsof Health in Developing
Countries.Oxford: Oxford University Press.
Miranda Radic, E., and Y Vergara. 1993. "Evoluci6n y Estructura
del Sector Salud en Chile." In OPS, ed., Estructura y
Comportamiento del SectorSalud en laArgentina,Chiley Uruguay.
Cuademo Tecnico 36. Washington, D.C.
Munioz,A., 1993. "Evoluci6nde laAtenci6n de Salud enArgentina."
In OPS, ed., Estructuray Comportamientodel Sector Salud en
la Argentina,Cbiley Uruguay.Cuademo Tecnico36. Washington,
D.C.
Murray, Christopher. 1994. "Towards an Analytical Approach to
Health Sector Reform: Investing in Health Systems." Bulletin
Ofthe World Health Organization72(3): 663-74.
Musgrove, Philip. 1995. "Cost Effectiveness and Health Sector
Reform." Human Resources Development and Operational
PolicyWorking Papers 48. World Bank, Washington, D.C.
.1996.Public
andPrivateRolesinHealth:
Theoryand
Fnncing
Patterns.World Bank Discussion Paper 339. Washington, D.C.
Panadeiros, M. 1996. "Organizaci6n del Seguro de Salud en la
Argentina: Analisis y Propuesta de Reforma." Desarrollo
MANAGED CARE AND MANAGED COMPETITION IN LATIN AMERICAAND THE CARIBBEAN
231
Economico 36 (summer).
Rossetti, A. G. 1996. "The Political Economy of Policy Change:
Concepts for the Analysis of Health System Reform." World
Bank, Latin American and the Caribbean Region Technical
Department, Washington, D.C.
Robinson,J. C., and L. P Casalino. 1996. 'Vertical Integration and
Organizational Networks in Health Care."HealthAffzain15(1):
7-22.
Santana, A. D. 1996. "Los Servicios de Salud Pre-pagos en
Republica Dominicana." Santo Domingo, Tercera Vers.
Schieber, George, Jean-Pierre Poullier,and Leslie M. Greenwald.
1994. "Health Systems Performance in OECD Countries:
1980-1992." Health Affairs 13 (4): 100-12.
World Bank. 1993. World Development Report 1993: Investingin
Health. New York:Oxford University Press.
. 1996."Argentina:Facingthe Challengeof Health Insurance
Reform." LatinAmerica and the Caribbean Region,Vashington,
D.C.
. Forthcoming. "Health, Nutrition and Population Sector
Strategy Paper, Statistical Annex." Human Development
Department, Washington, D.C.
Medical Savings Accounts
for Developing Countries
Len M. Nichols,NicholasPrescott,and Kai Hong Phua

Schieber and Maeda(in thisvolume)outlinethe fun-

damentalhealthcarefinancingchoicesfor anyecon-
omy.Mostsystemsrelyon a mixofpublicand private
funds, using variations along a continuum between two
extremes:relianceon governmentthrough public financ-
ing,public providers,and governmentprice and resource
controls;and relianceon marketsthrough privatefinanc-
ing, private providers,market prices, and private invest-
mentin healthfacilities.Likethe otherpapersin thisvolume,
this paper assessesthe advantagesand disadvantagesof
financing options along this continuum for countries at
differentstagesof development.Under certainconditions
medicalsavingsaccountscould playan importantrole in
increasingresources,efficiency,and equity in market-ori-
entedhealthsystems.Thispaper clarifiesthe preconditions,
risks,and riskmanagementstrategiesforincreasedreliance
on medicalsavingsaccounts,insurance,and market forces
generally.
Featuresof Medical SavingsAccounts
Financialstructure
Medicalsavingsaccountshave two essentialcomponents:
a fund withbalancesthat are set asidefor medicalexpenses
and that mayaccumulateat the discretionof the individ-
ual; anda backupfinancingmechanism.The backupmech-
anism is some kind of insurance,for its use is contingent
on certifiedmedicalevents.This mechanismtypicallyhas
either a largeor no explicitsizelimit, andtherebyprovides
backup financingfor medicalneeds that would exhaust
the medical savingsaccount fund. Backup financingcan
be whollyor partlyprivate or public.
Medicalsavingsaccountproposalsassumethat both com-
ponents-the fund andthe backupinsurance-are present.
Both componentsare needed because the distributionof
health care expendituresis highlyskewed.In the United
States,for example,10percent of the populationaccounts
for70 percent of healthcare spendingin a givenyear(Berk
and Monheit 1992). The health spendingof this 10 per-
cent is,on average,eighttimesthe populationwideaverage,
and that of the top 1 percent is thirty times the national
meanexpenditure.Similardistributionsare observedin all
technologicallysophisticatedhealth care deliverysystems.
This skewed distributionof health spending means that
somekind of poolingor insurancemechanismis needed to
guaranteeaccessto high-costbut necessaryhealth care.
There are twowaysto pool healthrisks:cross-sectional,
across individualsand families during a given year; and
intertemporal,over manyyearsfor an individualor family.
Themainproblemwithcomprehensive insurancein a cross-
sectionalpool is moral hazard: sufficientresourcescan be
marshaledforhealthcareneeds,but lowcopaymentrequire-
ments can lead to excessivedemand for health services

LenM.Nicholsis principalresearchassociateatthe UrbanInstituteinWashington, D.C.NicholasPrescottis senioreconomist in the

EastAsiaandthePacificRegionattheWorld Bank.KaiHongPhuaisseniorlecturerin healthpolicyandmanagement in theDepartment
of Community, Occupational, andFamilyMedicine atthe NationalUniversity
of Singapore,
andadjunctfellowattheInstituteof Policy
Studiesin Singapore.Theauthorsaregratefulto DeborahChollet,AlexPreker,andGeorgeSchieberforhelpfulcommentson earlier
drafts,thoughthe authorsretainresponsibility
foranyerrorsor ambiguities.
Theopinionsexpressed herearethoseofthe authorsand
do notnecessarilyreflectthoseof the UrbanInstitute,WorldBank,orNationalUniversityof Singapore.

233
(Chollet and Lewis in this volume). Thus some type of
demandmanagementis required.The mainproblemwith
intertemporalrisk pooling is cash flow:until an individ-
ual's annual contributionsto the medicalsavingsaccount
havebeenaccumulatedovermanyyears,resourcesare insuf-
ficientforunpredictableand catastrophichealthcareneeds.
Thus some type of cross-sectionalpoolingis required to
insureagainstthese needs.
Combined,medicalsavingsaccountsand high-deductible
backup insurance can deal effectivelywith both pooling
problems.Medical savingsaccounts help make patients
more awareof health care costs and so reduce the moral
hazardassociatedwith cross-sectionalpooling,andbackup
insurancereducesthe cash-flowproblem of intertemporal
poolingover an individual'slifetime.Byitselfa medicalsav-
ingsaccountis aweakintertemporalrisk-poolingdevicefor
individualsor families.Althoughit canaccumulateandthus
be used for a variety of future purposes, it could never
financethe costsof being in the top 1 percent or even 10
percentof peopleneedinghealthcare.Bycontrast,backup
insurance,by pooling health risksacross individualsand
familiesin a givenyear,can easilyfinancethe high health
carecostsof the few whoneed themwith relativelylowper
capitapremiumsor tax payments.
Contributionsto the medicalsavingsaccountfund can
be eithervoluntaryor compulsory(through a mandate or
tax),and theycan be madeby individualsor by employers
(includinggovernments)on individuals'behalf.Toencour-
age the spreadof medicalsavingsaccounts,a tax subsidy
maybegrantedto incomeor expendituresdevotedto these
funds. Contribution and withdrawal limits may also be
imposed to prevent high-incomeindividualsfrom using
medicalsavingsaccountsto evade tax liabilities.
Backupinsurancepoliciescan be purchased in private
marketsif feasible.The state can adopt the principlesof
managedcompetitionto facilitatethe developmentof pri-
vate insurance markets (see Enthoven in this volume).
Alternatively,the state can sell insurance,with subsidies
for the poor,or directlyprovidebackup insuranceby pro-
viding services or subsidizingproviders. To encourage
patients to be parsimoniouswhen demanding care, the
policy'sdeductibleand out-of-pocketlimitshouldbe set at
high levels,higherthanthe annualmedicalsavingsaccount
contributionlimit-perhaps asmuch as 10percentof fain-
INNOVATIONS IN HEALTH
234
ily income.This meansthat patients willhaveto pay 100
percent of medicalcosts, either from the medicalsavings
accountor otherpersonalfunds,untilahighlevelof expense
has been incurred.Deductiblescan varywith income,but
this adds considerableadministrativecomplexity.
Mechanics
Medicalsavingsaccountcontributionsare accumulatedin
a fund managedby a trustee-an employer,bank, insurer,
or the state. When workersneed healthcare, they can use
the balanceofthe medicalsavingsaccountto paytherequired
fees. If medicalneeds in a givenyear exhaustthe balance
of the medicalsavingsaccount,thenadditionalout-of-pocket
funds must be expendeduntil the backup insurancepol-
icy'sdeductibleamountis reached.Atthat point the backup
policystarts to pay for the healthcare needs of workersor
their families.Dependingon the policy,the family'soblig-
ation is eithergreatlyreducedor eliminated.
If a worker's familydoes not use much health care dur-
ing the year,then the medicalsavingsaccount fund could
have a positivebalance at year's end. This balance could
be withdrawn and spent on other goods, or it could be
retainedandaccumulatedforfuturehealthcare needs.This
balancecouldbe usedas savingsforretirementorlong-term
healthcare needs.This sourceof savingscouldbe particu-
larlyimportantforcountriesthat nowhavedecliningdepen-
dencyratios and relativelyfew elderlycitizens,but expect
to havemuchlargerretiredpopulationsin the future.Interest
could be earned on these accountbalancesand taxed or
not, dependingon policyobjectives.
Health Care Financing with Medical
Savings Accounts and Insurance
offinancialrisks
Distribution
Thissectionillustratesdifferentcombinationsof publicand
privatefinancinganddescribesthreehealthfinancingmod-
els with medicalsavingsaccounts and backup insurance
policies.Allthe modelsare specialcasesof the generalone
depicted in figure 1 through T(x) and HH(x), whichrep-
resent total and householdor privatelyfinancedexpendi-
tures, respectively'
CARE FINANCING
FIGURE
I amountto providerson behalfof all patients.Whetherit
Healthexpendituredistributions
T, 7 7
H,
Pnivate \Public < medical savingsaccounts, especiallyif mandated, the greater
\ H(x) T(x)
X, X2 $/HH
Annualhousehold
expenditures
Annual household expenditures are measured along
the horizontal axis.The verticalaxismeasures the percentage
of households with particular health expenditure levels.The
entire area under T(x) represents total national health spend-
ing. This general framework can accommodate any financ-
ingstructure, for the difference between total and household
(private) spending is public spending.2 Different financing
systems have these lines closer or farther apart, depending
on the importance of public subsidies.
Household expenditures are the sum of out-of-pocket
payments, private insurance payments for health services,
and payments for health services made out of medical sav-
ings accounts. Any or all of these components could be
zero in any particular health financing system. The differ-
ence between T(x) and HH(x) can be explained in two ways,
but each interpretation represents some form of public pay-
ment for health care. Consider expenditure level X1 . T,
households have total expenditures equal toXI. HI house-
holds might pay all of X1 out of pocket, with private insur-
ance orwith medical savingsaccounts, inwhich case T 1 -H 1
households get their entire X1 from the government.These
householdscould be said to be insured by a government
program. They may be poor or privileged.
Alternatively, all households could be primarily respon-
sible for their own health care, but a government health
service subsidy equal to (T1 -. H1 ) X1 is shared among all
T1 households. That is, the government controls health ser-
vice costs below market levels or total cost and pays some
MEDICALSAVINGSACCOUNTSFOR DEVELOPINGCOUNTRIES
235
occursthrougha publicinsuranceprogramforcertaingroups
or a publiclyfinancedprice subsidyfor all,the difference
between T(x) and HH(x) can be interpreted as a measure-
mentof the publicpaymentsforhealthcareaggregatedover
allindividuals.
No householdpaysfor healthcare with private sources
beyond the catastrophic expenditure level X 2 . Care pro-
vided beyond X2 is completely under the discretion of the
state. The more comprehensive are private insurance and
is the share of health expenditures financed outside the gov-
ernment budget, since expanding private financing instru-
ments will shift HH(x) up and to the right.
Altemativemedicalsavings
accountmodels
Keeping this expenditure distribution in mind, we now turn
to examples of alternative financing models. The first med-
ical savings account system we discuss has a purely public
backup insurance mechanism; the second has a limited
private backup insurance system; and the third has a purely
private backup mechanism. In each case we assume that
contributions on behalf of each worker are by employers as
well as by workers.
Medical savingsaccountswith public backup. In the first
model, individual medical savings accounts with a public
sector backup, contributions fund only the medical savings
account. Workers draw from this fund as health care needs
arise for them or their family, now or in the future. Under
this model the state essentiallycompels individuals and fam-
ilies to create intertemporal risk-poolingdevices using med-
ical savings accounts, and provides the cross-sectional
risk-poolingbackup mechanism to finance health care needs
if the medical savings account is exhausted. This backup
assistance can be through a public insurance program,
through direct provision of subsidized care by public
providers, or through public subsidies to private providers.
For people with positive balances in their medical sav-
ings account at the end of any year, the state has an impor-
tant policy choice. (This choice is present in all the health
system financing models discussed in this section.) If the
primary goal of the compulsory medical savings account
programwasto ensure the accumulationof sufficientpri-
vate fundsto financefuturehealthcare needs(and thereby
avoidintergenerationaltransfersas the populationages),
then the state should not allowbalancesto be withdrawn
in full for non-health care purposes.But if the main pur-
pose of the medicalsavingsaccountschemewasto reduce
current health care spending by making patients pay for
healthcare,then the stateshouldallowwithdrawalsfornon-
health purposes.Althoughboth motivationsare likelyto
be presentin manycases, analystscan inferwhichmotiva-
tion is strongerby noting whichof these alternativeseach
state chooses.
Medicalsavingsaccountswith mixedbackup.The second
modelrequireslargercontributionsbyemployersandwork-
ersbecausetheymustcoverboththemedicalsavingsaccount
anda backupinsurancepolicy.This high-deductiblebackup
insuranceprovidescross-sectionalrisk poolingfor families
but is not largeenoughto fullycoverallcontingenciesonce
medicalsavingsaccounts are exhausted.This option can
be viewedas a medicalsavingsaccountwitha privateinsur-
ance"corridor."The corridorliesbetweenthe medicalsav-
ings account and the public backup system for truly
catastrophicevents.Thus the cross-sectionalrisk pooling
here is partly private (over the insured only) and partly
public (overthe wholesociety),or mixed.
Corridorinsurancepolicieshaveexplicitlifetimeor annual
benefitlimits,and whenthey are reachedthe public sector
financesextremelyhigh-costcases.Thiscorridorof limited
liability insurance may be a good way to nurture a private
insurancemarket,but the state could also sell or compel
the purchaseof insuranceproductsthat would accomplish
the actuarialpooling required. We consider this corridor
"private"regardlessof wherethe insuranceis purchased,
sincethe basicgoalis to pool risk amongthe insured group
of individualsand familiesand not the societyas a whole,
and sincethe fundingfor this insurancecomesfrom indi-
vidual contributions made by the insured (even though they
may be compulsory).It is also private in the sense that
individualshave discretionoverhowthis additionalhealth
carepurchasingpoweris spent.Privatelydirectedpurchasing
powermayhavemanyimportantimplicationsfor the health
deliverysystem(seebelow).Corridorinsurancecould eas-
ilybestructuredto financea considerableportionof a coun-
INNOVATIONS IN HEALTII CARE FINANCING
236
try's health care expenses,sincethe distributionof health
expendituresis so skewed and highlycorrelatedwith the
need for hospitalization.
Medicalsavingsaccountswithprivatebackup.The third
financingmodel has a fullyprivatebackup insurancesys-
tem for people with medicalsavingsaccounts.The insur-
ance and cross-sectionalrisk pooling in this system are
completein the sense that they cover all the health care
needsofthe insuredabovethosefinancedby out-of-pocket
paymentsand the limitedintertemporalpoolingcreatedby
medicalsavingsaccounts.In this case the public sector's
responsibilityis reducedto guaranteeingaccessforthe poor.
Again,it can do sothroughpublic insurance,direct service
provision,or providersubsidies.
These different systemsand backup mechanismsfor
medicalsavingsaccountsare illustratedin figure2. The first
bar representsa traditionalsystem,in whichthere is no pri-
vate insuranceor medicalsavingsaccounts.Instead there
is only a mix of public financing combined with private
out-of-pocket payments. Introducing medical savings
accounts can reduce public spending because private
resourcesare mobilized.Medicalsavingsaccountscan also
replacedirectout-of-pocketspending.Privateinsurancelia-
bilityis smallerin the mixedbackupcasethan in the purely
FIGURE2
Financinghealthcarewith medicalsavings
accountsand insurance
Percentage
of healthexpenditure
100
80
insurance
60
40
20 1t
Status Public Mixed Private
quo backup backup backup
Financingsystem
private backup case because the "corridor" model preserves
public liability for the catastrophicallyill. The purely private
backup systemhas the smallest public sector role, but even
it has considerable public spending because the state is
presumed to subsidize insurance premiums and health ser-
vice copayments and to directly finance service provision
for the poor.
In all medical savingsaccount systemsthat include backup
insurance, workers and their families have strong incentives
to use health resources parsimoniously at least up to a point,
while truly catastrophic health care expenses are covered
either by the public sector or by private insurance. In both
cases more private funds are drawn into the health financ-
ing system than in the traditional system, increasing the
state's options for allocating health resources. Total national
health spending could be larger under all systems with med-
ical savings accounts and much larger under the mixed
backup or private backup options, since they mobilize con-
siderable private resources to pay for health care. A larger
share of private resources would enable budgetary resources
to be better targeted to the health care needs of the poor.
In each system basic ambulatory health services for a rea-
sonably healthy family with average income could be cov-
ered from their medical savings account plus moderate
out-of-pocket obligations. Sicker and poorer families may
require additional subsidies.
Medical SavingsAccounts and Health
Policy Objectives
Medical savings accounts are controversial in health policy
research circles. Singapore's experience has been used to
support a particular side in the contentious U.S. political
debate (Hsiao 1995), and one of this paper's goals is to clar-
ifywhat Singapore's experience means for other countries-
especially developing countries.
With two notable exceptions (Bond, Heshizer, and
Hrivnak 1996; CBO 1997), reviewed below, practically all
the literature on medical savingsaccounts outside Singapore
has been theoretical-either advocacypieces (Tanner 1995;
Goodman and Musgrave 1988;Pauly and Goodman 1995)
or hypotheticalmicrosimulation
exercisesdesignedto show
who would win or lose financiallyupon switchingfrom a
comprehensive insurance policy to a medical savings
MEDICAL SAVINGS ACCOUNTS
237
accounthigh-deductible insurance arrangement (Keeler and
others 1996; Nichols, Moon, and Wall 1996; American
Academy of Actuaries 1995; Ozanne 1996; O'Grady 1996).
This section highlights analytical issues by discussing a set
of hypotheses about the effects of medical savingsaccounts
that could be tested with proper data and research design.
The discussion of these hypotheses is organized around
three broad themes common to public finance in develop-
ing countries: resource mobilization, efficiency,and equity.
Resource mobilization
In health systems that have heretofore been whollyor mostly
public, creating medical savingsaccounts and private insur-
ance or private discretion over health care purchasing may
be the most politically palatable way to make the middle
and upper classes pay more for health services, since most
consumersplace considerable value on discretion overhealth
services and providers. Thus one hypothesis is that intro-
ducing medical savings accounts will increase total spend-
ing on health as privateresources are mobilized. Introducing
medical savings accounts as well as a compulsory backup
insurance mechanism would increase health spending even
more. Underlying these hypotheses is the assumption that
the state willnot reduce the absolute levelof its health spend-
ing, but may redirect it, as discussed below.
Another resource mobilization hypothesis is that private
demandforhealthserviceswillencourageprovidersandoth-
ers to organize and invest in more efficient health care insti-
tutions in order to serve the emerging block of purchasing
power. In essence, private purchasing power may make it
profitable to invest in health facilities. Over time this could
considerably augment the supply of health facilities.
Efficiency
There are three broad hypotheses about the efficiencyeffects
of medical savings accounts and backup insurance. It has
been argued that medical savings accounts will lower costs,
rationalize health professionaland health facilitysupply,and
improve the quality of care.
Costs.As noted, the extremely skewed nature of health
expenditures makes risk pooling essential. This pooling of
FOR DEVELOPING COUNTRIES
resourcessets up a naturaltradeoffbetween shieldingindi-
vidual patients from financialruin and the moral hazard
problem described in Schieber and Maeda (in this vol-
ume) and Cholletand Lewis(in this volume).Patientsare
inclinedto use more healthresourcesthan is optimalifthe
marginalprice they face is artificiallylower than the true
marginalcost. Withinsuranceand risk pooling,this gap is
typicallyfinancedby an externalthird party.Medicalsav-
ings accounts create incentivesto consumefewer health
resourcesbecausepatientsare responsiblefor the fullcost
of care and are spendingtheir ownmoney,at least untilthe
backupinsurancetakes over.
There are two other waysto managethe excessdemand
that resultsfromrisk poolingand insurancefinancing.The
first is for the state to control the supply of health care
providers until supply constrainseffective demand. This
approachwillkeep spendingdown,but it maynot provide
adesirableallocationofresources,sincetriagedecisionsare
essentiallymade by the state.3
The secondmechanismfor dealingwith excessdemand
for health care,increasinglycommonin the United States,
is managedcare (see Enthovenin this volume).Managed
care, in all its manyforms,essentiallyenforcestriagedeci-
sions and non-price rationing by private physicians.
Competitionin the marketplacefor healthplan enrollees
protects patients from providers'incentivesto underpro-
videcare. Marketsand qualityreportingstandardsneed to
be welldevelopedfor competitionto be effective,and it
has provento be so in some markets.
Bythemselvesmedicalsavingsaccountsattack the root
cause of excessdemand: copaymentobligationsthat are
belowmarginalcost. But the key differencebetween this
approachto the moral hazardproblem of healthinsurance
and the othersis that consumers-not healthplan profes-
sionalsor the state-get to choose.In essence,consumers
ration their ownhealth care instead of havingelitesration
it for them.To the extent that healthresourcesare free to
followthe resultingdictatesof consumerdemand,medical
savings accounts represent a tool for constructing a
consumer-directedhealth care system.This is their advo-
cates' ideal (Goodmanand Musgrave1988).
Highercost-sharingobligationswillsurelyreducehealth
serviceutilization.But cost-containmentresultsmaynot be
asdramaticassomeproponentsof medicalsavingsaccounts
INNOVATIONS IN HEALTH CARE FINANCING
238
contend. First, the medical savingsaccount may still be
looked on as someoneelse's moneyto spend, especiallyif
someof the contributioncomesfromemployers(American
Academyof Actuaries1995).Furthermore,since medical
savingsaccounts can be used to pay the first incurred
expense,some patientsmay actuallyuse more health care
than they did when they faced coveragelimits or had to
make out-of-pocketcopayments(Pauly1994).
Second, providers have the power, as in any market
with asymmetricinformation,to influenceconsumerdeci-
sions,especiallyonce the deductiblehas been reachedand
the marginalcostto the patientdropsdramatically. The bulk
of national health spendingcould actuallyoccur at levels
abovethe backupinsurancepolicy'sdeductible(Nichols,
Moon, andWall 1996;Keelerand others 1996).Providers
have incentivesto encourageuse when they are paid on a
fee-for-servicebasis.Furthermore,high spendingis gener-
ally associatedwith complexhospitalinpatient stays.The
nature of these illnessesand treatment optionsmakesfam-
idiesmuch less willingto sacrificeservicesto save money
than they are in more discretionaryambulatorycontexts
(AmericanAcademyof Actuaries1995).
Third,the combinationof medicalsavingsaccountand
backupinsuranceproducessubstitutionand incomeeffects
that counteracteach other.The rightto keepunused med-
icalsavingsaccountbalancesand the requirementto pay a
highinsurancedeductiblemean that patients initiallyhave
relativeprice incentivesto reduce health serviceuse. At
the sametime,the backupinsurancepolicyis likean income
effectthat willincreaseconsumption,especiallyif this cov-
erageisnew.Thusthe net effectofmedicalsavingsaccounts
on the level and rate of growthof per capita costsis some-
what ambiguous.
Providersupply.To control costs, state-run health care
systemshave to limit the number of health professionals
and health facilities.In the discussionof resource mobi-
lizationabove,wehypothesizedthatmedicalsavingsaccounts
would lead to more privateinvestmentin health facilities.
Here we argue that medicalsavingsaccounts and backup
willraise
insurance,by directingdemandto privatefacilities,
thepricesof someprofessionals' servicesandtherebyincrease
the long-runsupplyof those professionals(assumingthat
the state relaxesartificialadmissionand certificationlim-
its). To the extent that this privately directed demand
becomesdominant,the resultingsupplyof providersmay
be morereflectiveof consumerpreferencesthanthe provider
mix producedby the state planningapparatus.
Quality.Similarly,
the privatelydirectedpurchasingpower
createdby medicalsavingsaccountsand backupinsurance
will likelyincreasedemand for qualitycare outside state
facilities.In some (mostly)public systemsresource con-
straints and rational triage require lowerquality in some
areas of care in order to provide adequate qualityto the
larger number of patients in most areas. Medicalsavings
accountsandbackupinsurancecreatethe abilityto payfor
higherqualityin more limitedareas. This may or maynot
affectthe averagequalityof care,dependingon other pol-
icy decisionsdiscussedbelow.
Equity
Mobilizingprivate resources to finance medical savings
accountsand insurancewillcause middle-classcitizensto
paymore andlow-incomecitizensto paylessforhealthcare,
ifthe statetargetsspendingon the low-incomepopulation.
Thusmedicalsavingsaccountsandprivatehealthinsurance
make it possibleto expand accessto health care without
increasingpublicresourcerequirements.There is no guar-
antee, however,that the state willtake this approach.
Moreover,there are equityrisks,at leastduringthe tran-
sition.Providersthat suddenlyhaveexcessdemandmayraise
their pricesas governmentsdismantlethe price controlsof
regulatorypublic systems.In the long run, free entry into
healthprofessionsand facilityconstructionmayguarantee
that competitivemarketswillprevail,but educatinghealth
professionalsandbuildinghealthfacilitiestakeconsiderable
time.The resultscouldbe enrichedproviderswholeavethe
controlledpublicsector for the more lucrativeprivatesec-
tor,abetter deliverysystemforthe wealthy,and fewerhealth
resourcesavailablefor everyoneelse.This risk mayjustify
usingstate healthpersonnelpolicy,at leastduringthe tran-
sition to competitivehealth servicesmarkets,to guarantee
that enoughproviderscontinueto servepublicsectorpatients.
Suchpersonnelneed not remain state employees,but they
maybe requiredto work one or two daysa week at public
facilitiesand receivelowerremunerationfor those days.
MEDICAL SAVINGS ACCOUNTS FOR DEVELOPING COUNTRIES
239
Anotherdownsideriskis that privatedemandmayfinance
ahigh-qualityuppertier in a two-tieredhealthsystem,espe-
ciallywhere highlytrained physiciansand state-of-the-art
hospitalsare scarce.The existenceof the upper tier could
reducewillingnessto supportpublicfinancingof the lower
tier. Thusit could destabilizethe entire system.Cross-sub-
sidies may be necessary,as well as vigilant attention to
qualitydifferentialsbetweentiers.
In addition,at least with the private backup insurance
mechanism,there is some risk that out-of-pocketobliga-
tions could be largerelativeto income.Waysto minimize
this risk include limits on the out-of-pocket maximum
and policiesthat maintaincompetitiveor regulatedhealth
serviceprices. Some analysts fear that the incentives of
medicalsavingsaccountswilllead patientsto neglectpre-
ventive and basic care and thus frustrate the deliveryof
cost-effectivecare. There is some evidencethat patients
in industrial countries are not wellinformed about com-
plexhealthservices,and that whenfacedwith strongfinan-
cial incentives,they tend to reduce both necessaryand
unnecessaryhealth servicesin equal measure (Newhouse
and the InsuranceExperimentGroup 1993).At the same
time,reports fromU.S. companiesthat haveimplemented
medicalsavingsaccountsin the past fewyearsdo not indi-
cateanyproblemsof this sort, althoughadequate data are
not yet availablefor systematicand independent study
(CBO 1997).
Riskselectionconcems
By far the most contentiousissuein the U.S. debateover
medicalsavingsaccountshas been risk selection-the fear
that medicalsavingsaccountswould appealmostlyto the
healthy,andthat paymentsforhealthcarewouldthenbecome
more correlatedwith health status than they are now (see
Nichols,Moon, and Wall 1997 and the references cited
therein). This fear is relevant to all voluntarysystemsin
whichpeople havethe option of buyingmore comprehen-
siveinsurance.Althoughthis maybe an importantissuefor
developingcountriesthat eschewcompulsoryparticipation,
this paper focuseson a more commonproblem associated
with the skeweddistributionof health expenditures:set-
tingthe medicalsavingsaccountcontributionappropriately.
Publicand privatebackupsare consideredseparately,
 Toillustratethe problem,considera systemthat isfinance
entirelyby publicfunds. In this case,settinga medicalsav-
ings accountlevelis equivalentto determiningthe amount
that willbe givento each citizenin lieu of public coverage
for at least some health care needs. If the citizenuses no
health services,then the state has "lost"moneyon that cit-
izenbecauseit couldhaveusedthat moneyto payforsome-
one else'shealthcare.Thetheoryisthat byprovidingmedical
savingsaccounts and their incentivesfor parsimony,peo-
ple who do use health resourceswilluse sufficientlyfewer
of them to compensatefor the publicmoneythat wouldbe
"wasted"onpeoplewhodo not use them.Institutionalizing
mnedical savingsaccountsmayhelpteach peoplethat health
care is not free for anyone,and this maybe importantin a
long-rundevelopmentstrategy.But as a short-runtechni-
cal point it should be noted that it is possible to set the
publiclybackedmedicalsavingsaccounttoohighandthereby
"lose" money if more unspent health moneys go to the
healthythan the sicksave (CBO 1996).
With private backup insurance, the medical savings
account amountmust be coordinatedwith the deductible
of the privateinsurancepolicythat accompaniesit.The ana-
lyticalproblem is similarto the previouscase,exceptthat
in this case the insurer,not the state, is the residualloser
or beneficiaryif the medicalsavingsaccountamountis set
too high or too lowrelativeto the deductible.If a medical
savingsaccount contributionis too large,it willencourage
healthcare use that increasesthe insurer'scosts.If it is too
small,it willdiscourageuse and lead to short-runinsurer
profitsand lowerpremiumsin a competitivemarket. The
simplestway to solvethis coordination problem is to let
privateinsurersoffer high-deductiblepremiumsandmed-
ical savingsaccountpackages,for then both would be set
actuarially,at least in competitiveequilibrium.
With publicbackup insuranceand a privatelyfinanced
medicalsavingsaccount,the state faces the same contin-
gent liabilityproblem as the insurer in the privatebackup
case.If the medicalsavingsaccountis too smallrelativeto
the insuranceprogram'sdeductible,it willdiscourageuse
and couldsavethe statemoneybynot coveringmuchhealth
care use. If it is too large,it couldincreasestate expendi-
tures by increasingdemandfor the high-endcare that the
state finances.Thus the state needsto coordinateits insur-
ance programparameterswith privatelyfinancedmedical
INNOVATIONS IN HEALTH CARE FINANCING
240
savingsaccounts to achievespending and access targets.
One way to do this is to dictate the parameters-contri-
bution levels,servicescovered-of the privatelyfinanced
medicalsavingsaccountoptions. This process is straight-
forwardin compulsorysavingsor tax systems.If the state
insteadallowsvoluntarymedicalsavingsaccountcontribu-
tions,then it must respondto market-determinedmedical
savingsaccounts. But medical savingsaccounts cannot
find a market equilibriumuntil the state revealsits pro-
gramparameters.This createsa strategicsituationin which
state and insurerinterests maydiverge.The generalpoint,
however,isthat medicalsavingsaccountandinsurancepara-
meters willbe coordinatedone way or the other, for one
set is contingenton the other.
Singapore's Experience
A detailed overviewof Singapore's health care delivery
and financingsystemis providedby Phua (inthis volume).
This sectionhighlightsthe salientpolicychoicesand their
effects.
features
Institutional
Singaporeis one of the East Asian "tiger" economies-a
small,high-incomecity-statewith a populationofjust 3 mil-
lionbut a per capitaGNP of $22,500(by comparison,per
capitaGNP in the United States is $25,880).Singapore's
experiencewith medicalsavingsaccountscan be used to
drawlessonsfor implementationof these accountsin other
countries.In 1984Singaporebecame the first economyto
implementmedicalsavingsaccountson a nationwidebasis.
And to this day the Medisaveschemeremainsthe world's
onlyexample of an applied medicalsavingsaccountpro-
gramintegratedwitha country'shealthfinancingstructure.
The medicalsavingsaccountmodel that has evolvedin
Singaporeover the past decade correspondsto the med-
icalsavingsaccountwith mixedbackup described earlier.
In other words, individualMedisaveaccountsare embed-
ded in a broader financingframeworkthat backs up the
medicalsavingsaccountswitha cross-sectionalcatastrophic
risk-poolingschemecalledMedishield(introducedin 1990)
anda means-testedsafetynet for the poor calledMedifund
(introducedin 1993).This three-tierpackage-Medisave,
Medishield,and Medifund-is backed up by government
financingof supply-sidesubsidiesto publicprovidersaimed
at loweringthe net priceschargedto patients (table 1).
The mechanicsof the Singaporemodel are straightfor-
ward. Contributions to Medisaveare an integral part of
Singapore'scompulsorysocialsecuritycoverage,provided
by the CentralProvidentFund (CPF). Enrollmentin the
CPF is mandatoryfor allemployees(and, sinceJuly 1992,
for the self-employed).The CPF is funded by a mandatory
payrolltax equivalentto 40 percent of the wage bill, split
evenlybetween employersand employees.Of this 40 per-
cent contribution,between 6-8 percentagepointsare allo-
cated to the member's Medisave account. These
contributionsare incometax-deductibleandinterestbear-
ing.The Medisavebalancecan accumulateup to S$ 19,000,
beyondwhichincrementalsavingsare rolled over into the
CPF member'sordinaryaccount,fromwhichit canbe with-
drawn after age 55.
Medishieldis alsomanagedunder the CPF umbrella-
allCPF membersareautomaticallycovered,andthe annual
premiumsare deductedfromtheir Medisaveaccountunless
theychooseto opt out.Medifund,however,is not managed
bythe CPF.It is an autonomousendowmentfundintowhich
periodicbudgettransfersare made, andfromwhichinvest-
ment incomeis used to financedemand-sidesubsidiesfor
the poor.
Withdrawalscan be made from individual Medisave
accountsto paymedicalbillsincurredbythe accountholder
and immediate familymembers. However,withdrawals
are subject to two important exclusions.First, because
Medisaveis designedto pay for hospitalizationexpenses,
ambulatorycare is generallynot covered.Second,eligible
MedifundCommittee.This safetynet is targeted at house-
holdswith two adults and three childrenearningless than
S$ 1,400 a month (the lower one-third of the income
distribution).
Effectson healthpolicyobjectives
Resourcemobilization.The main policyobjectiveguid-
ing the introduction of medical savingsaccounts was to
mobilizenonbudgetaryresourcesto helppayforthe increas-
ing health costs of Singapore'sagingpopulation.Medical
savingsaccounts were expected to absorb pressuresthat
wouldotherwisedrain the budgetof the Ministryof Health,
freeingthe ministryto focus on the prioritygovernment
functionsof publichealth and protectingthe poor.
A completeassessmentof Medisave'seffectivenessin
mobilizingresourcesforhealthmust take into accountboth
staticanddynamicperspectives.Astaticviewof Medisave's
importanceis givenby its weightin the overallstructure of
health financing.In 1995Singapore'sratio of total health
care costs to GDP was estimated at just 2.7 percent. Of
this, private expenditures accounted for 60 percent and
publicbudgetaryspending,for 31 percent.
Public extrabudgetaryexpenditures financed 9.6 per-
cent of total health spending-mostly Medisave(8.5 per-
cent),followedbyMedishield(0.8 percent)and Medifund
(0.3 percent).This staticperspectiveshowsthat Medisave
TABLEI
Sourcesof health financingin Singapore, 1984-95
(percent)
1984 1986 1988 1990 1992 1994 1995

hospitalizationexpensesare capped (S$ 300 a dayfor hos- Publicsector 37.1 37.2 33.0 32.0 34.0 39.3 40.3
pital chargesplus limitsfor each surgicaloperation).Thus Budgetary 35.3 27.8 21.3 21.2 24.4 29.7 31.0
the averagehospitalbill requires a significantcopayment Capital 26 o 2.9 25 615 8.8 11I2
Current(net) 26.8 24.2 18.3 18.8 17.9 20.9 19.8
in addition to the portion paid by Medisave.Claims for Extrabudgetary 1.8 9.4 11.7 10.8 9.7 9.6 9.6
Medishieldbackupcoveragefor catastrophicexpensesare Medisave 1.8 9.4 11.7 10.7 9.4 8.8 8.5
Medishieid n.a. n.a. n.a. 0.0 0.3 0.5 0.7
subjectto a high annualdeductible(S$ 1,000)as wellas a Medishield Plus na. na. na. n.a. n.a. 0.0 0.1
20 percent copayment.In addition, Medishieldcoverage Medifund n.a. n.a. n.a. n.a. n.a. 0.3 0.3
exdudes preexistingconditionsand is subjectto claimlim- Private sector 62.9 62.8 67.0 68.0 66.0 60.7 59.7
its of S$ 20,000per policyyearand S$ 80,000per lifetime. Out-of-pocket 62.9 62.8 67.0 66.5 63.9 58.7 57.7

Finally, Medishield coverage expires at age 75. As a last Prvateinsurance n.a. n.a. n.a. n.a. 1.6 1.8 2.0

resort,patientsunableto paytheir billsat. government

.
hos-
~~~~
Tota0.
~ ~ ~~~~~Total
100.0 100.0 100.0 100.0 100.0 100.0 100.0
00 0 0 00 . 0
pitalscan applyfor a mean-testedgrant fromtheir Hospital Source:Woad Bankstaf estimates.

MEDICAL SAVINGS ACCOUNTS FOR DEVELOPING COUNTRIES

241
has succeeded in playingan important though not at all not wasteits resourcepool on financingmedicaleventsthat
dominant role in mobilizingresources to finance health aregenerallyaffordableaspartofroutineconsumptionexpen-
expenditures.Indeed, if the role of Medisaveis expressed ditures,and for whichinsurancedoes not produce anywel-
relativeto hospitalizationexpenditures-which is all it is fare gain.Sinceoutpatientcare could consumeas much as
intendedto helpfinance-it appearsmuchmoresignificant two-thirdsof healthexpenditures,this significantexclusion
(almost30 percent). frees a largeamountof contributionsto be rolled overinto
From a dynamicperspective,however,it seemslikely individuals'lifecyclesavingspool. This, in turn, allows
that Medisave wiDlplayan increasingly
importantrolein financ- Medisaveto achievesomedegreeof intertemporalriskpool-
ing as Medisavebalancescontinueto accumulateover indi- ing to cover intermediatefinancialrisks,whichcould pro-
viduallifecycles.
Today'sflowof disbursements fromMedisave duce a welfaregainto its beneficiaries(dependingon their
accountsunderstatesthe flow of contributions,and there- rate of timepreferenceandthe payoutrate).In otherwords,
fore the accumulatingstock of Medisavebalancesavailable excluding outpatient coverage has an indirect dynamic
to financefuture claims.The actualMedisavepayoutratio resourcemobilizationeffect by loweringthe payoutrate,
(expendituresas a share of contributions)has consistently whichallowsMedisaveto performan efficientintertempo-
been around 20 percent.In fact, the payoutratio has been ral insurancefunctionfor hospitalization.Thus the outpa-
faling becausethe nominaldailycap of S$ 300 has not been tientexclusionturnsoutto be awin-winpolicyonboth counts.
adjustedsinceinception.Forexample,in 1995Medisavepay- Second,the insuranceeffect of Medisavecoveragefor
mentstotaledS$ 291 miDlion, comparedwithestimatedcon- intermediate(capped)hospitalizationcostsis accompanied
tributionsof S$ 1,809million.Thatsameyear,the cumulative byefficiencyincentiveson both the demandand the supply
savingsgeneratedbythe low payoutrate had alreadymobi- sides.On the demandside,the Medisaveexpenditurecaps
lizeda total Medisavebalanceof S$ 12,700million-equiv- meanthat a significantportionof hospitalbillsare not cov-
alent to four years'worth of Singapore'saggregateannual ered by Medisave.WorldBank estimatessuggestthat 60
healthspending.Thedynamicsofthe underlyingsavingsaccu- percentof hospitalizationcostsin publichospitalsare sub-
mulationareillustratedbytherisein percapitabalances(from sidized.The residual40 percentthat is chargedto patients
S$2,595in 1985to S$5,400in1995)aswellasbytheincrease is splitabout evenlybetween Medisaveand out-of-pocket
in the shareof employeeswho attainthe requiredminimum payments.Thus patientsfeel the bite of individualrespon-
balanceby age 55 (44 percentin 1995). sibilitynot onlyin the form of an average20 percent coin-
Medisave'sexceptionalresourcemobilizationpotential surancefractionpaidindirectlyout oftheirMedisaveaccount,
results from two crucial features. First, its revenues are but of another20 percent paid directlyout of pocket. On
supportedby a sizablepayrolltax base madepossiblebythe the supplyside, Singaporerecentlyimposedrevenuecaps
extensiveformalemploymentin Singapore'srich, urban- onthe government'srestructuredhospitalsthatlimitedallow-
ized economy.Nearlyuniversalcoverage-the number of able increasesin hospitalrevenueper patient per dayto 5
individualMedisaveaccountsis equivalentto 80 percent percenta yearduring1994-96.Singapore'sremarkablylow
of the population,and dependentfamilymembersare also ratio of health care costs to GDP is consistentwith these
eligible-has been achievedbytappinginto the well-estab- efficiencyincentivesfor containingcosts.
lished socialsecurityoperationsof the Central Provident
Fund. Second, Medisave'sexpendituresare sharplycon- Equity. Medisave's effects on equity are not directly
tained by its twin policyexclusionslimitingcoveragepri- observablein terms of differentutilizationratesor afford-
marilyto hospitalization,with definedexpenditurecaps. abilitybetweenincomegroups.There are,however,two sig-
nificantly pro-poor elements built into the financing
Efficiency.
Medisave'sdesignis consistentwithprinciples frameworkin whichMedisaveis embedded.
of economicefficiencyin two importantways.First,by not First,governmentsubsidiescontinueto playanimportant
coveringlow-cost,high-probability outpatientcare-unlike backup role in financinghospitalcare,and are targeted to
comprehensivehealthinsuranceschemes-Medisavedoes the poor.Asnoted above,about 60 percentof publichospi-

INNOVATIONS IN HEALTH CAREFINANCING

242
tal costsare subsidized.Subsidiesare channeledto public
hospitalsandreflectedinpublicpricingpolicies.Explicitprice
discriminationis based on the four types of accommoda-
tionsin publichospitals,rangingin ascendingorderof com-
fortfrom classesC, throughB2 and B1, to A. Subsidyratios
are highlydifferentiated-averaging84 percentof hospital
costsin classC, 71 percent in classB2, 36 percent in class
B1,and 13percentin classA.Publichospitalsprovidefinan-
cialcounselingto prospectiveinpatientsto facilitateselec-
tion of an affordableward dass based on the expectedsize
of the hospitalbilland the individual'sMedisavebalance.
Thesedifferentialsubsidyratioshelp equalizethe afford-
abilityof dass-specificprices relativeto the incomes and
Medisavebalancesof the patients who select them. For
example,in 1995the 75th percentilehospitalbill charged
to patientsin classC was S$ 524, comparedwith S$ 854 in
classB2. Thesebillscomparewith medianmonthlyhouse-
hold incomes of S$ 787 in the bottom quintileof house-
holds and S$ 1,657in the second quintile. Government
estimates suggest that at those income levels the corre-
spondingbills were equivalentto 7.4 and 9.5 months of
Medisavecontributions,respectively.
The secondimportantpro-poorfinancinginstrumentis
Medifund.Althoughintended as a safetynet of last resort
forthe poor,availabledataindicatethat,duringits firstthree
years of operation,Medifund responded favorablyto 99
percent of requests for financialassistance,and paid the
entire medicalbill in 87 percent of those cases.Patients
receivingassistanceaccounted for 5 percent of hospital
admissionsat the classC and B2levels(3.3percentin 1993,
4.2 percent in 1994,and 5.8 percent in 1995).
Experience in Other Countries
AlthoughChinaand Malaysiaare exploringmedicalsavings
accountoptionsfor their healthsystems,the UnitedStates
is the only other country wheremedicalsavingsaccounts
are part of the health financingstructure and reports of
their effectshavebeen made. CBO (1997) surveysall the
publidyavailableevidencefromemployersthat haveswitched
from comprehensiveinsurance plans to medicalsavings
accountsandhigh-deductible insurance,andBond,Heshizer,
and Hrivnak(1996) surveyedtwenty-sevenfirms offering
medicalsavingsaccountsto their employees.It should be
MEDICAL SAVINGS ACCOUNTS FOR DEVELOPING COUNTRIES
243
kept in mind that the overridinggoal of U.S. policyis to
containcosts,not to mobilizeresourcesforthe healthsector.
Ozanne (1996)focuseson howfirmsreport their results
differently,and how the overall satisfactionwith medical
savingsaccountsdiffers.Forexample,someemployersreport
that their costshavefallensubstantially,whileothersreport
modestsavings.Somefirmshavediscontinuedmedicalsav-
ings accountsafter usingthem for a while,though no rea-
sonsare publiclystated. CBO (1997)alsooutlinesthe data
that need to be reported to properlyevaluatemedicalsav-
ingsaccounts,concludingthat "thereportsgenerallydo not
provide enoughinformationto fullyassesshow successful
the plans were at reducing medical spendingby and on
behalf of employeesand limitingadverseselection"(p. 2).
Most of the reported cost data refer onlyto firm costsor
insuranceclaimsfiled,not total costs(includingemployees'
out-of-pocketexpenditures).Furthermore,there are little
data on the healthstatusof workersbeforetheychosetheir
plan and the coststo employeesof differentplans.
Bond, Heshizer,and Hrivnak (1996) asked two basic
questionsof firmsthat had switchedfrom comprehensive
insuranceplans to medicalsavingsaccounts:Did employ-
ers' costs and employees'maximumout-of-pocketliabili-
ties go up or down?They found that employershad saved
an averageof 12percent and employees'maximumout-of-
pocket liabilityhad fallenby about a third. Althoughboth
resultsare partly due to higherthan averagebaselinecosts,
the findingsshowthat, on average,these firmsareprotecting
workers better than they were before medical savings
accounts.Whether this increased protection is commen-
suratewith loweremployercostsin long-runequilibriumis
an empiricalquestionthat willbe closelyexamined.
One reason the U.S. experiencewith medicalsavings
accountswillbe watchedeven more closelythan usual is a
newlaw,TheHealthInsurancePortabilityandAccountability
Act, signedby President Bill Clinton on 8 August 1996.
Starting1January1997,a limitednumberof employeesof
small firms (fewer than fifty employees) and the self-
employedwillbe giventax preferencesfor settingup med-
ical savings accounts along with their high-deductible
insurance plans.The U.S. General AccountingOffice is
requiredto studythisdemonstrationandreportto Congress
by 1January 1999.Atthat time the tax preferencefor med-
icalsavingsaccountsmaybe expandedor eliminated.
Lessons
for DevelopingCountries
Singapore's experience and our theoretical discussion lead
to five broad lessons for developing countries.
LessonI Medical savingsaccountscannot be usedalone
It is not actuarially feasible to have complete self-insurance
for health care. Medical savings accounts can be important
tools for mobilizing health resources, but only as part of a
comprehensive set of financinginstruments. Cross-sectional
backup insurance must also play a major role, and public
funding for catastrophic cases and for the poor will con-
tinue to be necessary.Health personnel policies and health
service price controls are also likely to be useful policy
tools to contain costs and maintain equity in the move toward
a more market-oriented health system.
Lesson2 Mobilizingresourceswill take a longtime
The resource mobilizationeffects of medical savingsaccounts
could help avoid the intergenerational transfer problems
that will be especially acute in developing countries with
rapidly aging populations and shrinking tax bases. But it
may take quite a few years to accumulate sufficient med-
ical savings account balances to transfer significant finan-
cial risk to households. Building public and provider
understanding of and support for systemwide change will
take time and is essential for a system based on private ini-
tiative and market incentives to work.
Lesson accountscanandshouldbe
3 Medicalsavings
designedto enhanceefficiency
This is particularly important in systems that are mobiliz-
ing more resources and rapidly expanding their health care
delivery systems. Increased resource mobilization and pur-
chasing power will produce a significant income effect that
could lead to substantial increases in health service utiliza-
tion. Medical savingsaccounts can be structured to increase
efficiencyin these utilization decisions byinsuring that sub-
stantial out-of-pocket payments are required before backup
insurance policies take on the bulk of the financial burden.
Medical savings accounts can create incentives or price
INNOVATIONS IN HEALTH
244
effectsthatcounterbalance
theincomeeffectofmorecov-
erage generally.The efficiency effect of these price incen-
tives may be enhanced by educating the public about prices
and comparative costs-especially through financial coun-
seling prior to utilization, as well as instant feedback on
medical savingsaccount balances upon billingand payment.
Lesson4 Medicol sovingsaccountsposeequity risks
Having households pay more out of pocket at the point of
service could raise already high barriers for low-incomefam-
ilies to fully participate in the health care delivery system.
Thus public subsidies for medical savingsaccount contribu-
tions, price schedulesthat varyby income, and public subsi-
dies for insurance premiums may be needed to ensure that
the evolvinghealth care systemis not limited to the wealthy.
Risk selection is a less serious problem in systems with
mandatory participation (and high compliance) than it is in
voluntary systems.Still,the medical savingsaccount amount
must be coordinated with the deductible of the backup
insurance policy that accompanies it. Uncoordinated lev-
els are unsustainable, especially if the system mixes public
and private insurance mechanisms. Unstable insurance sys-
tems alwaysthreaten the most vulnerable-that is, those
with recognizably great health care needs.
Lesson5 Therearemajorinstitutionalprerequisitesfor
implementingmedicalsavingsaccounts
First,per capitaincome levelsmust be highenough to finance
individual contributions to medical savingsaccounts as well
as premiumsto a backupcatastrophicrisk pool. Second,a
high degree of labor force participation in formal sector
employment is needed to provide a taxable transactions
base for resource mobilization.These conditions are unlikely
to be met in many developing countries, and may rule out
any immediate possibilityof nationwide coveragewith med-
ical savings accounts. However, such accounts could still
play a niche role for high-income urban employees, just as
existing social insurance arrangements do for civilservants
and industrial workers in many countries. Third, an effec-
tive system of payroll tax collection combined with effi-
cient fund management and claims processing is needed to
implement the financial operations associated with medical
CARE FINANCING
savings accounts. This is likely to be much easier in coun- Bond, MichaelT, BrianHeshizer,and MaryW Hrivnak. 1996.
tries that already have a social security mechanism with "ReducingEmployeeHealth ExpenseswithMedicalSavings
which medical savings accounts can be integrated. Fourth, Accounts."ClevelandState University,Ohio.
CBO (CongressionalBudget Office). 1996. "MedicalSavings
a well-developed computerized information system ii~lnk Accountsand Medicare."Washington,D.C.
ingpersonal savingsaccountswith hospital providers,backed . 1997. "AReviewof Reportsof EmployerExperiences

by strict securityand accountingcontrols, is essential. Finally, withMedicalSavingsAccounts." Washington,D.C.

in countrieswhere medical savingsaccounts and their backup Goodman,John C., and GeraldL. Musgrave.1988."Freedomof
insurance coverage are, for equity reasons, embedded in a ChoiceinHealthInsurance."NationalCenterforPolicyAnalysis
health financing structure that protects the poor through PolicyReport 134.Dalas, Texas.
Hsiao,WlliamC. 1995."MedicalSavingsAccounts:Lessonsfrom
self-targeted subsidies, the state must have the administra- Singapore."Health Affairs14(2):260-66.
tive capacity to implement a policyof price discrimination. Keeler,Emmett B.,Jesse D. Malkin,DanaP Goldman,and Joan
This requires an ability to channel supply-side subsidies to L. Buchanan.1996. "Can MedicalSavingsAccountsfor the
providers (public or private) while monitoring and regulat- NonelderlyReduceHealthCareCosts?"Journal oftheAmerican
ing the resulting hospital price structure. MedicalAssociation275(21): 1666-71.
Newhouse,JosephP, and theInsuranceExperimentGroup.1993.
Notes FreeforAI1? Lessonsfrom
theRandHealthInsurance Experiment.
Cambridge,Mass.:HarvardUniversityPress.
Nichols, Len M., MarilynMoon, and Susan Wall. 1996. "Tax
1. lT(x)and HH(x) are drawnwith the commonlyobserved, PreferredMedicalSavingsAccountsand CatastrophicHealth
approximatelylog norrnalshape for health expendituredensity N I
functions.
functions. ~~~~~~~~~~~~~Insurance
Plans:A NumericalAnalysisof Wlners andLosers."
WorkingPaper06571-002.UrbanInstitute,Washington,D.C.
2. We assumethat employersfinance their contributionsto ____ 197 "Epria Wor on Meia aig cons
employees'healthinsurancepremiumsthroughlowerwages.Thus What WeKnowNow,What WeNeed to Know,and Howue
householdsare the ultimatesourcefor employerpaymentsas well. Mht Wearn the Rest P epeed to the annualme
3. Triageis a processfor sortinginjured or sick people into ng of the Red Pal Sciente to Janual mew
groups based on their need or likelybenefit from immediate gorleans
medical treatment. Triageis used on the battlefield,at disaster OGrady,MichaelJ. 1996. "MedicalSavingsAccountsasd the
sites,in hospital emergencyrooms, and anywhereelse that lim- DynaMicsof Advrs Selectn Congssional arch
itedmedicalresourcesmustbe allocated.Today,medicalresources ServiceReportfor the US.S.
Congress.Washington,D.C.
are consideredto be limnited virtuallyeverywhere. SrieRpr o h .. Cnrs.Wsigo,DC
Ozanne,Larry.1996."HowWil MedicalSavingsAccountsAffect
MedicalSpending?"Inquiry33(3):225-36.
Pauly,Mark V 1994.Do Two WrongsMakea Right?An Analysis
References of MedicalSavingsAccounts.Washington,D.C.: American
EnterpriseInstitute.
AmericanAcademyofActuaries. 1995."MedicalSavingsAccounts: Pauly,Mark V, and John C. Goodman. 1995. "TaxCredits for
CostImplications and DesignIssues."PublicPolicyMonograph. HealthInsuranceand MedicalSavingsAccounts." HealthAffairs
Washington,D.C. 14(1):126-39.
Berk, Mark, and Alan Monheit. 1992. "The Concentrationof Tanner,Michael.1995."MedicalSavingsAccounts:Answeringthe
HealthExpenditures: AnUpdate."Health Affairs11(4):145-49. Critics."PolicyAnalysis
228. CatoInstitute,Washington,D.C.

MEDICAL SAVINGS ACCOUNTS FOR DEVELOPING COUNTRIES

245
Medical Savings Accounts and
Health Care Financing in Singapore
Kai Hong Phua
S ingaporemaintainsthe world'slongest-runningexper-
imentwithmedicalsavingsaccounts.Theseaccounts,
describedgenerallyin Nichols,Prescott, and Phua
(in this volume),help defraypublichealth care costswhile
ensuringthat citizensreceiveadequateand affordablecov-
erage. Moreover,the systemencouragesSingaporeansto
stayhealthyand minimizethe use of unnecessarymedical
services.
Health Care System
Singapore'shealthcare deliverysystemis mixed.The pub-
lic systemis run by the government;the privatesystemis
run by voluntaryand private hospitalsand practitioners.
The health care deliverysystemcomprisesprimaryhealth
care provisionat private medicalclinicsand government
outpatient polyclinics,and secondaryand tertiarycare at
privateand publichospitals.
About 80 percent of primaryhealth care is providedby
private practitioners;the governmentpolyclinicsprovide
the remaining20 percent.For hospitalcare, whichis more
costly,the situationis reversed:80 percent of hospitalcare
is providedby the publicsectorand 20 percent is provided
by the private sector.
Patientsare free to choose amongprovidersin the dual
health care deliverysystemand can walk in for a consul-
tation at any private clinic or governmentpolyclinic.For
emergenciespatients can go to the twenty-four-houracci-
dent and emergencydepartmentslocatedin government
hospitals.Singapore'sCivilDefenseForce runs an emer-
KaiHongPhuaisseniorlecturerinhealthpolicyandmanagement intheDepartment
at the NationalUniversity
of Singapore,
andadjunctfellowat the Instituteof PolicyStudiesin Singapore.
247
gencyambulanceserviceto transportaccident and trauma
cases and medical emergencies to the acute general
hospitals.
Public health servicesare providedthroughthree gov-
ermnent ministries.The Ministryof Health is responsible
for preventive,curative,and rehabilitativehealth services.
It formulatesnationalhealth policies,coordinatesprivate
andpublichealth care,and regulateshealth standardsand
legislation.The Ministryof Environmentis responsiblefor
environmentalpublichealthservicessuchasoverseeingsew-
erage and waste disposalsystems,ensuringthe safety of
food that is prepared and sold in Singapore,and control-
ling infectiousdiseases,air and water pollutionand toxic
chemicalsand poisons.The Ministryof Labor is responsi-
ble for the industrialand occupationalhealth of workers.
Singapore'shealthcarephilosophyemphasizesthe build-
ing of a healhy populationthroughpreventivehealthcare
programsand the promotionof healthyliving.Publichealth
education programs encourage Singaporeansto adopt a
healthy lifestyleand be responsiblefor their own health.
The publicis made awareof the adverseconsequencesof
habitslike smoking,alcoholconsumption,unhealthydiets,
and sedentarylifestyles.The governmentpolyclinicsoffer
every child free immunization against tuberculosis,
poliomyelitis,diphtheria,whoopingcough,tetanus,measles,
mumps,andrubella.Theseimmunizationsarealsoprovided
at private clinicsfor a small fee. Health screening pro-
gramshavebeenintroducedforthe earlydetectionof com-
mon ailmentslike cancer,heart disease,hypertension,and
diabetesmellitus(Seowand Lee 1994).
ofCommunity, Occupational,
andFamily
Medicine
 The government ensures that good and affordable basic
medical services are available to all Singaporeans by pro-
viding heavily subsidized medical services at public hospi-
tals and government clinics. All private hospitals, medical
clinics,clinical laboratories, and nursing homes are required
to maintain a good standard of medical services through
licensing by the Ministry of Health.
To promote personal responsibility, Singaporeans are
required to use the Medisave scheme to save for their hos-
pitalization expenses, especially during old age. This is to
avoid overreliance on state welfare or medical insurance.
Under the scheme every employee puts 6-8 percent of their
monthly income into a personal Medisave account. These
savings can be used to pay for hospitalization expenses
incurred by the worker or their family As noted, this sys-
tem encourages individuals to stay healthy and minimnize
the use of unnecessary medical services.
The public system requires that patients make copay-
ment for their medical services at the time of consumption
to discourage unnecessary use. For people who choose to
be accommodated in the lower classes of wards in public
hospitals (there are four classes), hospitalization expenses
are subsidized up to 80 percent by the government. Smaller
subsidies are given to people who prefer the comforts and
personalized service of the higher-class wards. The indigent
can apply to hospital management for partial or full remis-
sion of their bills at public hospitals. No Singaporean is ever
denied access to the health care system or use of the acci-
dent and emergency services, and no one is turned awayby
hospitals. Waiting time for an elective operation average,
between two and four weeks. There is no waiting for emer
gency admissions.
Primoryhealth services
Primary health services include primary medical care for
families, health screening and preventive health programs
for schoolchildren, home nursing, day care, and rehabilita-
tion for the elderly,and health education and promotion for
all.
The public sector comprises sixteen government poly-
clinics located throughout the country. Each clinic pro-
vides curative outpatient medical treatment, immunization,
health screening and education, investigative facilities,and
INNOVATIONS IN HEALTH CARE FINANCING
248
pharmacy services.About 770 private clinicsare run by 1,060
medical practitioners. The average outpatient consultation
fee (including medication) is about S$ 10-15, well within
the means of every Singaporean. At the government poly-
clinics children and the elderly (above 60 years) are given
up to a 50 percent concession in their payment.
Hospitolservices
There are about 10,500 hospital beds in twenty-four hos-
pitals in Singapore, or about 3.5 beds per 1,000 people.
About 80 percent of these beds are in the twelve public hos-
pitals, which have between 200 and 2,500 beds. The twelve
private hospitals tend to be smaller, with 60 to 500 beds.
The public hospitals set the standard of medical care and
benchmark for hospital charges.
Of the twelve public hospitals, six are acute general
hospitals. The others specialize in areas such as obstetrics
and gynecology,psychiatry,and infectious disease. The pub-
lic general hospitals provide multidisciplinary inpatient
and specialistoutpatient services and twenty-four-hour acci-
dent and emergency service. In addition, there are spe-
cialty institutes for cancer, heart, eye, and skin diseases.
Tertiaryspecialistcare for cardiology,renal medicine, hema-
tology,neurology,oncology,radiotherapy, plastic and recon-
structive surgery, pediatric surgery, neurosurgery,
cardiothoracic surgery, and transplant surgery are central-
ized in two of the larger general hospitals, Singapore General
Hospital and National UniversityHospital. The private hos-
pitals have similar specialists and comparable facilities.
The government has also introduced low-cost community
hospitals for intermediate health care for the convalescent
sick and aged who do not require the more expensive care
of the acute general hospitals.
In public hospitals, Singaporeans can choose from dif-
ferent types of wards and accommodations upon admission.
Patients pay more for a higher level of physical amenities,
althoughthe provisionof medical care is similarfor all accom-
modations. In most cases serious medical conditions are
treated in the public hospitals by senior consultants or spe-
cialists, regardless of the type of ward chosen by the patient.
The average length of stay in the general hospitals is about
five days. Hospital beds are well utilized, with an average
occupancy rate of about 80 percent.
 Since 1985the governmenthas restructured(or corpo,
ratized)fiveof its acutehospitalsand twospecialtyinstitutes
to be run as private companieswhollyowned by the gov-
ermnent.This wasdone to givethe corporatizedhospitals
the managerialautonomyandflexibilityto promptlyrespond
to the needs of the patients. In the process commercial
accountingsystemshavebeenintroduced,providinga more
accuratepictureof operatingcostsaswellasinstillinggreater
financialdisciplineand accountability.Corporatizedhospi-
tals are differentfrom other private hospitalsin that they
receivean annual governmentsubsidyto providemedical
services.Moreover,theyare expectedto be managedlikea
nonprofitorganization.The corporatizedhospitalsare sub-
jectto broad policyguidanceby the Ministryof Health.
Dentalhealthservices
Under its dental health programthe governmentprovides
freedentalcareforallschoolchildren. Both generalandspe-
cialized dental care are provided in five hospitaldental
clinics, two community dental clinics, and the central
GovemmentDental Clinic.Dental servicesare also avail-
able in the private sector,where 320 dental clinicsrun by
500 dentalsurgeonsprovidespecializedaswellasbasicden-
tal treatment.
Aspart of the governmentprogramfor preventiveden-
tal care, the nationaldrinkingwater supplyhas been fluo-
ridated since 1956.Dentalhealth educationis providedto
all patients presentingfor dental treatment and to special
groups such as schoolchildren.The governmentregularly
organizesdental health campaignsto raise awarenessof
dental conditionsand their prevention.The dental health
educationprogramis creditedwith loweringcaries among
childrento one of the lowestrates in the world.
Major Concerns and Future Challenges
AlthoughSingaporehas come a long wayin improvingits
health indicatorsand standards of medicalservice,many
concernsand challengesremain.Theseinclude:
* Increasingcosts of providinghealth care as a result of
advancesin medicalknowledgeand technology,leading
to increased specializationand subspecializationand
greateruse of medicaltechnology.
MEDICAL SAVINGS ACCOUNTS AND HEALTH CARE COVERAGE IN SINGAPORE
249
* Risingexpectationsand demand for better and more
sophisticated health services by an increasinglywell-
informedand more affluentpublic.
* Rapid agingof the population,with the portion of peo-
ple 60 years and above estimated to increase from 9
percent at present to 25 percent by 2030.
* Shortageof staff such as nurses and health therapists,
asyoungpeoplebecomemore attractedto lessdemand-
ingjobs than those in the healthcare services.
Singaporehas sought to control supply factors in the
healthsectorthroughdeliberatestaffingand facilitiesplan-
ning.On the demandside, effortshavebeen madeto mobi-
lizecompulsorysavingsthroughthe CentralProvidentFund
(CPF), thusrationingdemandimplicitlythroughconsumer
purchasingpower.The governmentuses socialplanningto
avoidthe problemsfacedby most countriesin maintaining
a balancebetween demandpressuresand supplycapacity.
The situationcould potentiallybe more acute for a small
nationlikeSingapore,whereresourcesare limitedandexpec-
tations are risingfor more and better services.
Healthexpenditure
Inthe early1960stotal healthexpendituresweremore than
4.5 percent of GDP In the mid-1960sthe level of public
spendingwas similarto private spending.Sincethen gov-
ernmenthealthspendinghas dropped from 2.5percent to
1.0percent ofGDP,whileprivatespendinghas rangedfrom
1.5to 2.5 percent. Thus total health expendituredid not
increaseas fast as GDP during 1960-80.
In the early 1980shealth spendingas a share of GDP
wasjust 2.5percent (table 1).This lowlevelwasrelatedto
rapid growthin GDP2There was alsosteadyexpansionin
privatehealth spendingduringthis period, as reflected in
increasingnumbersof doctorsin privatepracticeas wellas
privatehospitaladmissions(table2). Today,there are more
privatedoctorsthan governmentdoctors.
Asredevelopmentandrationalizationof the government
hospitalsprogressed,governmentspendingremainedhigh
throughsubsidiesto the Ministryof Health,althoughthere
wereperiodicincreasesin userchangesto reflectrisingcosts.
Until 1985andbeforesubstantialcostrecoveryeffortswere
introduced,Ministry of Healthsubsidieswereequalto about
half of privatehealthconsumption.
TABLE
I
Health care expendituresin Singapore, 1975-95
(millionsof Singaporedollars)

Private Public Total

Shareof Shareof Shareof
Year Amount GDP (%) Amount GDP (%) Amount GDP (%) GDP
1975 249 1.86 138 1.03 387 2.89 13,373
1980 396 1.58 223 0.89 635 2.53 25,091
1985 702 1.80 419 1.08 1.198 3.08 38,924
1990 1,526 2.25 457 0.67 2,032 2.99 67,879
1995 2,368 1.96 685 0.57 3,380 2.80 120,629
Source:
Ministry
of Health,Singapore.

Total health spending now accounts for about 3 percent
of Singapore's GDP1 During the 1980s public spending
was less than 1.0 percent of GDP, while private spending
increased from 1.5 percent to more than 2.0 percent. Thus
there has been a discernible shift toward private spending,
reflecting price increases and cost recovery efforts in the
public sector as well as the consumption preferences of a
affluent population for a perceived higher quality of ser-
vice. The government currentlysubsidizesabout one-quarter
of total health care costs.
TheNational Health Plan
Until recently most medical costs were incurred by gov-
ernment hospitals, as well as the growing private hospital
sector, and financed by personal payments, limited insur-
ance coverage, and employment benefits that include com-
pany plans for workers and their families (Fong and Phua
1985). The Ministry of Health's goal is to provide quality
health care that is not only available and accessible, but
that is also affordable and must be paid for.
TABLE
2
Supplyof health facilitiesand doctors in Singapore, 1960-95
These considerations form the basis for the National
Health Plan formulated by the ministry in 1983. The plan's
key proposal, the Medisave scheme, imposes compulsory
savings and restructures the system of health care financ-
ing. The main objectives of the plan are to secure a healthy
and productive population through active promotion of
healthy lifestyles, and to improve cost efficiency in the use
of health services.In additionto promoting individualrespon-
sibility for maintaining good health, the plan aims to build
up financial resources in order to create the means to pay
for medical care during illness (Ministry of Health 1983).
Restructuring
ofgovernment
hospitals
Moves to restructure health care financing were made to
avoid the problems of a welfare state system financed by
taxes (such as the United Kingdom's National Health
Service) and to shift the burden of financing health care to
individuals, families, and employers (including the private
and voluntary sectors). The strategy used was to increase
cost sharing by users and to progressively move the provi-

Admissions Admissions
Number of Government Private to govemment to private Govemment Private
Year hospitals hospitalbeds hospitalbeds hospitals hospitals doctors doctors
1960 14 6,537 650 na na 282 358
1965 16 6,817 859 120,274 na 450 469
1970 16 6,891 869 135,952 na 496 867
1975 22 8,005 1,100 164,205 na 855 847
1980 26 8,078 1,507 234,502 31,326 914 1,121
1985 22 8,329 1,671 229,988 47,164 1,214 1,307
1990 21 7,922 1,837 259,541 77,562 1,831 1,593
1995 24 8,326 2,211 266,142 91,413 2,124 2,191
Source:Singapore
Department
of Statistics, various
Yearbookof Statistics, years.

INNOVATIONSIN HEALTHCARE FINANCING

250
sion of healthcare to the privatesector.This approachwas
initiallycarriedunderthe bannerof the privatizationmove-
ment, but during its implementationthe term restructur-
ing was preferred in view of the sensitivitiesinvolved.
Although the extent of privatizationwas never explicitly
defined in the health sector,one goal was to transfer the
managementand control of major public hospitalsaway
from the government(Phua 1991).
The hospitalrestructuringprogramfirstproceededwith
the formationof a government-ownedsubsidiary,incorpo-
rated under the umbrellaof TemasekHoldings,to manage
the newlycompleted National UniversityHospital from
June 1985 onward.This arrangementwas later changed
when the hospitalwas placedunder the controlof a new
government-ownedstructure, the Health Corporationof
Singapore.This corporationwas createdin April 1987 to
acquireand manageallrestructuredgovernmenthospitals,
beginningwiththe NationalSkinCentrebyDecember1988,
SingaporeGeneralHospitalbyApril1989,KandangKerbau
Hospitaland ToaPayohHospitalbyApril1990,Singapore
NationalEyeCentre by October 1990,and TanTockSeng
Hospitalby April1991.The NationalUniversityHospital,
Singapore'smainteachinghospital,waseventuallyput under
the governanceof the NationalUniversityof Singapore.
Under the restructuringprogram public hospitalsand
specialtyinstituteshavebeen incorporatedas privatecom-
panieswhollyowned by the government.The current aim
of the restructuringprogramis to givegreater managerial
autonomyto governmenthospitals so that they can pro-
videmoreefficientand higher-qualityservice,improvepro-
ductivity,controlcosts,andhavegreaterflexibilityto rapidly
respondto changingneeds.
Arestructuredhospitalis fullyautonomousandcan recruit
its ownstaff,set its ownterms of remuneration,and decide
on the deplbymentof resources.Comparedwitha govern-
greater
menthospital,a restructuredhospitalhas significantly
autonomyover its operations. Managers of restructured
hospitalsare accountableto the board of directors for the
hospital'sperformance.
Restructuringis no longer regarded as a privatization
exercise.The hospitalsare still 100 percent owned by the
governmentand continueto pursue the socialobjectives
of the Ministryof Health (WHO 1994).Restructuredhos-
pitals continueto subsidizetheir patients and receivean
MEDICALSAVINGSACCOUNTSANDHEALTHCARECOVERAGEIN SINGAPORE
251
annual subsidyfrom the governmentto offset their oper-
ating deficits. Since the restructuring program began in
1985,the governmenthas restructuredfive(ofseven)acute
hospitalsand two specialtyinstitutions.
Medisave
In 1981the Ministryof Health held discussionswith med-
icalprofessionalsin the public and private sectors and in
the NationalUniversityof Singaporeto gatherideasonhow
to furtherdevelopSingapore'shealthcaresystem.Additional
researchand data collectionwere then conducted,includ-
inga reviewof healthcare systemselsewhere.Membersof
Parliamentwere also consulted. In 1982 the minister of
health proposedthe Medisaveschemeand its underlying
philosophy,whichwas followedby extensivepublicity.For
the next year discussionswere held with communitylead-
ers from citizens' consultativecommittees,representative
employers'federations, trade unions, and health-related
associations.
A paper on the National Health Plan was released in
1983and was widelypublicized.Feedbackfrom the pub-
lic was activelysolicited,and there was follow-upby the
mediaon issuesrelated to health care financingand deliv-
ery.Parliament then debated the National Health Plan
before approving the Medisavescheme, which compels
Singaporeansto set aside their own savingsto meet future
hospitalizationexpenses,and recommendingthat there be
periodicreviewsof the scheme'simplementationto ensure
that appropriateadjustmentswere made.
The Medisaveschemewas givenwide media coverage
to explainhowit wouldwork.Printed materialswerewidely
distributedand public talks were held in communitycen-
ters,hospitals,andcompanies.Theschemewasimplemented
in governmnent hospitalson 1 April1984.
Beforethe schemewas expanded to indude approved
privatehospitalsin 1986,it wasintroduced as a pilot pro-
ject involvingthe new National UniversityHospital (run
by an autonomousquasi-governmentcompany)in 1985.
The scaleof chargesand Medisavewithdrawalshere were
equivalentto those in govemmenthospitals.Withdrawal
rates, however, were subjected to a daily maximum for
hospitalchargesand a maximumrate foreach surgicalpro-
cedure.The samelimitsnowapplyto the privatehospitals.
 The schemehas been modifiedseveraltimesbased on
the experienceacquired.Initially,accountholderswereonly
allowedto use their Medisaveaccountsto payfor the full
costof hospitalstaysin lower-pricedwards,and for part of
the costs in more expensiverooms.This ceilingwas later
extended to cover alnost all hospitalcharges, subject to
maximumdailylimits.In addition,Medisaveaccountsthat
are used to payfor higher-pricedaccommodationsare not
allowedto be overdrawn.
Implementationof the Medisavescheme,though for-
mulatedand coordinatedbythe MinistryofHealth,required
the activeparticipationof many groups,includingmedical
and related professionals,academics,politicians,commu-
nity and grassrootsleaders,employersand employees,and
the media.It tookmore thantwo yearsto thoroughlydebate
the issues involved,to disseminatevital information,and
to gather feedbackfrom alllevels.This approachpointsto
the importanceof bottom-upplanningandcommunitypar-
ticipation in the wide acceptanceand successfuladminis-
tration of anyinnovativepublic program.
Advantages. In essence,Medisaveservesas an additional
sourceofpersonalfinancingformedicalexpendituresincurred
by families.This shift in public cost sharingfrees govern-
menttax revenueformore urgentptioritiesand contributes
to better publichealth services.It is hoped that Medisave,
actingas a personalhealthfinancingand paymentscheme,
willcontroleffectivedemandthroughthe pricemechanism.
UnderMedisavemostpaymentsforhealthcareare made
at the point of consumption.This close link betweenpay-
mentand use better reflectsthe realcostsof healthcareand
helpspreventexcessuse. Medisavealsocaters to different
consumerpreferencesfor a range of accommodationsin
public and private hospitals.Thus,withincertain limits,it
can be used to providecompletecoveragein lesscomfort-
able wards and to subsidizemore expensivecharges.
Thusthe Medisaveschemerepresentsa majordeparture
from the socialsecurityschemesof other countriesin sev-
eral areas.First,Medisaveis not a commonpool of fundsto
be used indiscrimninately by a governmentfacingpressures
from interest groups to respond to short-termproblems;
instead,it is a schemethat coversonlydependentfamilymem-
bers.This fitsin with the conceptthat the familyis the basic
socialand economicunit of anysociety,with caringfor the
INNOVATIONS IN HEALTII CARE FINANCING
252
welfareof its members(includingthe sickand the elderly)a
collectivebut essentiallypersonalresponsibility.The aim is
to preserveand enhancethe stabilityof an essentialsocial
structureamid rapid environmentalchanges.Onlyif a fam-
ilyisunableto sharein the medicalexpensesof its sickmem-
bersdoes the stateuse publictaxesto subsidizehealthcare.
The ideais to promote self-reliance, althougha safetynet is
stillavailableforthose in need (Phua 1986).
Anotherfeature of Medisaveis that, unlike tax-based
financing,it doesnot placean undulyheavyburden on the
employedandthe young,and doesnotsubjectpublicexpen-
diture to the vagariesof economiccycles.The current gen-
eration of workersis obligedto save for the future,instead
of relyingon the uncertaintaxes of the next generationfor
support.This is in linewith officialpolicyto promote finan-
cialindependenceamongthe elderly,whosemedicalneeds
are expectedto increase(Phua 1987).
Disadvantages. AlthoughMedisaveis designedas com-
pulsorysavingsfor predictablehospitalizationneeds,espe-
cially among the elderly, it seems to have encouraged
imirediate spendingamongyoungergroupsfor expensive
hospitalservicesthat are perceivedto be of higherquality.
There has been a dramaticshift in demand from the gov-
ernrneht hospitalsto the restructured and private hospi-
tals, and a discernible upgrading from the lower- to
higher-pricedbeds. SinceMedisavedoes not cover ambu-
latorycare (except for certainprocedureslike hepatitis B
vaccinationand kidneydialysis),peoplemayhaveanincen-
tive to shiftmore care and spendingto the hospitalsector.
Moreover,there is considerableignoranceon the part
of the public with regard to the limitsof Medisavecover-
age (for example,that onlyS$ 300 a dayis allowedfor pri-
vatehospitalbills),creatingthe illusionofmoremoneybeing
availablethan in reality.Many people also view (wrongly)
Medisavefundsas frozenassetsthat, ifunspent,areretained
by the government.Thus manypatients opt for more lux-
uriousservicesthan theywouldotherwiseconsume.In prac-
tice,Medisavehasactedmorelikeasupplementaryfinancing
schemeto enable the consumptionof upgraded hospital
services,as wellas a mechanismfor recoveringcostsin the
publichospitalsystem.
A criticalassumption of the Medisavescheme is that
patients are the best judge of how their savingsshould be
spent on health care.However,effectivehealth care deci-
sionmakingrequiresconsiderableknowledgeabout prices
and the quality and probable outcome of medical treat-
ments, as wellas a high standard of ethics on the part of
practitioners.Sincemanyof these preconditionsare not in
place, there are grounds to intervenefor the public good
and to protect consumers'interests.
Most of the scheme'scurrent disadvantagescan be cor-
rectedbybettereducatingthe publicandintroducingchecks
and balances,includingcost controllimits and incentives
aimedat both the supplyand demandsides.Revenuecaps
have been imposed on the public hospitals, which also
have to maintain at least two-thirdsof their hospitalbeds
for subsidizedpatients. Besidesencouragingthe prudent
use of Medisave,financialcounselingof patients and fam-
ily members prior to admissionand other such measures
(includingmedicalaudits) are being institutedto prevent
abuse and to maintain standards.To discourageunneces-
saryhospitalization,Medisavehas been extendedto cover
moreambulatorycare (suchasdaysurgeryand certainexpen-
siveoutpatient services).Althoughthe Medisaveaccount
coversthe acutehospitalizationneeds of the typicalfamily,
it is insufficientto covermajor catastrophicillnesses.Such
illnessesrequiregreaterrisksharingand financingthat insur-
ance could better provide for.
Medishield
Sincethe 1950sthe issue of health insurancehas cropped
up in variousdiscussionsabout its manyshortcomingsand
lackof applicability
forSingapore(Ministryof Culture1982).
In the deliberations that led to the formulation of the
NationalHealth Plan andthe Medisavescheme,the health
insuranceoption was dismissedbecause of the negative
experiencesof other countries' health insurancesystems.
Amongthe mainweaknesseswere:
* The illusionof a free serviceat the point of consump-
tion, encouragingoveruseand escalatingcosts.
* Lackof incentivesforthe consumerto staywelland for
the provider to economizesince they are reimbursed
for anyutilization.
Thus the Medisaveschemewasconceivedto avoidthe pit-
fallsof third-partyreimbursementsystemsfinancedfrom
insurancepremiumsor taxes (Ng 1988).
MEDICAL SAVINGS ACCOUNTS AND HEALTH CARE COVERAGE IN SINGAPORE
253
After Medisavewas implemented,however,it became
apparentthat therewas stilla needforcoverageof majorill-
nessesrequiringexpensiveandprolongedtreatment.Tocater
to suchneeds,a low-costnationalcatastrophicillnessinsur-
ance scheme,knownas the Medishieldscheme,wasintro-
ducedon 1July1990.Toavoidthe problemsassociatedwith
prepaidinsurance,thereis a systemof deductiblesand coin-
surance.Allmembersof the CentralProvidentFund (CPF)
who are citizensor permanent residentsare automatically
coveredunlessthey opt out of the scheme.Participationof
CPF members'immediatedependents is voluntary.Non-
CPF membersmayoptintothe schemebycontributing toward
Medisave.CPF membersand theirdependentsare insured
up to the age of 70. To encouragepeople to participatein
the scheme,premiumsare low and affordable,and varyby
agegroupto minimizecross-subsidy. Annualpremiumsrange
fromS$ 12forpeoplebelow30 to S$ 96 forpeoplebetween
60 and 65, to S$ 132 for people in the 65-70 age group.
Premiumscan be paid from Medisavefunds.
Reimbursementis on actualexpensesincurredunderthe
plan (upto a limit)lessthe initialdeductibleof S$500-1,000,
whichis borne by the insured. The copaymentis 20 per-
cent,andthereareannualandlifetimeclaimlimits.Medisave
can be used to pay the deductibles and copayments.
Deductiblesare intentionallykept high to avoidexcessive
demandformedicalservices.(Deductibleswere peggedat
the levelwhere only 10 percent of hospitalizationswould
be eligiblefor Medishieldclaims.)
Medifund
Despite the widespread coverage of Medisave and
Medishield,a smallnumberof Singaporeanslackadequate
savingsor familysupportto payforhealth care.Thisgroup
includes older cohorts of low-incomeindividualswithout
familieswhohaveinsufficientorno providentfund accounts.
The idea of settingup a largeendowmentfund to help fill
this financialgapwasbroachedin 1991,and parliamentary
approvalwas granted in January 1992.Accordingto the
MedicalEndowmentAct,the governmentwilldepositgrants
in specialaccountsfor publichospitalsto defraypart or all
of the billsincurredby eligiblepatients.
Medifundwas given start-up capital of S$ 200 million
and is supplementedby S$ 100 milliona year from bud-
getary surplus.A MedifundAdvisoryCouncilwas formred
to adviseon the use of income derived from the endow-
ment.Each hospitalhas appointeda Medifundcommittee
to approve paymentsbased on the guidelinesestablished
by the council.
SinceApril 1993needypatientshavebeen able to apply
for partial or full waivers of their medical fees through
medicalsocialworkers.Priorto this, waiversand subsidies
had to be absorbedby individualhospitals through gov-
ernment subventions from general tax revenues. Thus
Medifundfurtherrelievesthe dependencyon taxationas a
means of healthcare financing.
Recenthealthpolicies
Demographic,epidemiological, social,andeconomictrends
indicate that the tax-based health care financingsystem
will be strained even further in the years to come. The
main objectives of health care financing policies are to
strengthen the safetynet to protect people againstrising
healthcostsas wellas to controlcosts.Medisaveformsthe
first layerof this safetynet, based on compulsorysavings
and a familysupportsystem.Medishieldprovidesthe next
layerof the safetynet, coveringmajor chronicconditions
requiringlong-termand high-costhealth care.Medifundis
an additionallayer,financedthrough an endowmentfund
createdfrom surplusrevenue.This multilayeredfinancing
systemshouldbe able to withstandthe increasingburden
of health care costs as wellas provide a greater degree of
socialsecurityin the yearsahead.Througha multipronged
approachusing a public-privatemix of health services,a
varyinglevel of subsidyand cost sharing,and Medisave,
Medishield,and Medifund, it is hoped that a more con-
trolledincreasein health expenditurewillbe achievedin
line with the rate of inflationand economicgrowth (Kwa
1989).
A ReviewCommitteeon NationalHealth Policieswas
appointed in 1991 to reviewfuture policydirections for
the country.Thefirstpart of its report,presentedin October
1991,emphasizedhealth promotionand diseasepreven-
tion as the basic philosophyguiding Singapore'shealth
carepolicies(Ministryof Health 1991).Its mainreport,pre-
sented in February1992,recommendedpriorityareas for
actionand measuresfor improvinghealth care while con-
INNOVATIONS IN HEALTH CARE FINANCING
254
trollingcosts. Suggestionswere made on staffingneeds,
medicalspecializationand training,and professionalstan-
dards and qualityof care. The recommendationsfocused
on healthcare financing,definingthe government'srolein
financing health care and managing health care costs
(Ministryof Health 1992).
The govemmentaccepted the recommendationsof the
reviewcommittee,and a MinisterialCommitteeon Health
Policieshas finalizedthe courseof actionto be implemented.
Aparliamentarypaper on healthcare,presentedin October
1993,set out the government'sapproach to controlling
healthcare costs,in order to keep basichealth care afford-
ableto allSingaporeans.Amongthe cost-containmentmea-
sures:
* Defininga good basic medicalpackage
- Regulatingthe supplyof doctors and hospitals
- Regulatingsubvented hospitalsthrough revenue caps
and subsidies
* Encouraginggreater cost sharingin medicalinsurance
and employmentbenefits
* Controllingpricesin the privatesector
- Coordinatingmedicalresearchand development
* Coordinatingmedicaleducationandtraining(Singapore
1993).
Althoughmany of these policiesare beingimplemented,it
is too earlyto evaluatetheir full impact.
Conclusion and Recommendations
Can Singaporecontinueto ensurethat its healthfinancing
systemwilldevelopalongabalancedpath to achieveequity,
efficiency,and cost-effectiveness
in deliveringhealthcare?
Givenpast experiencewith the traditionaltax-fundedsys-
tem, as well as with the Medisave and the Medishield
schemes,what other modificationsare necessary?Current
interest seemsto focus not on whether alternativemeth-
ods of health care financingcan play a greater role, but
whethernewformsof fnancingcouldbe successfully imple-
mented to contain costs.
Anyviablefinancinginstrumentmust be able to balance
supplyand demandon a sustainablebasis,minimizemoral
hazardandabuse,avoidlabormarketdistortions,andresolve
distributionaleffectsacrosssocialgroups(suchas intergen-
erationalincometransfers).Againstthesecriteriathe checks
and balancesbuilt into Singapore'shealth care financing Whenresourcesarepooledamongfamilymembersor com-
systemseemto provideadequatesecurity.Givencontinued bined with insurancemechanismsfor catastrophiccover-
robust economicgrowth,healthcare spendingas a shareof age,compulsorysavingsare a potentiallypowerfulway to
GDP may appeardeceptivelylow at present. As the econ- finance health care and old age securityon a sustainable
omymatures,pressurescan be expectedfrom employersto basis(allowingforlifetimerisk-adjustedcontributions)with-
lowerhealthcareconsumptiongivenrisingwagecosts. out distortingeconomicgrowth.
Amongthe keyissuesthat willcontinueto be addressed
in the healthsectoris the needto meetthe ever-rising
expec- References
tationsof an aging,affluentpopulationfor the latest med-
icaltechnology.Can the push toward achievingexcellence Fong,N.P, andK.H. Phua.1985."Utilisation andExpenditure
in health care and developingSingaporeas a regionalhi- onMedicalServices in aLocalCommunity." Singapore
Medical
Journal26(2):131-38.
tech medicalcenterbe balancedwith more sociallyaccept- Kwa,S.B.1989.'Financing theHealthCareSysteminSingapore."
ablemeansof costsharing,whiledampeningthe inflationary Singapore Medical Association
Newsletter 20(8):1-2.
effectsof rapidgrowth?In the futurecost-containment poli- Ministry of Culture.1982."Health-ACrucialConcern.Excerpts
ciesaimedat both supplyand demandwillhaveto be imple- fromMinisterialSpeeches on Health1980/82."Information
mented judiciouslythroughoutthe health care system. Division, Singapore.
Althoughelementsof Singapore'shealth financingsys- Ministryof Health.1983."BluePaperon the NationalHealth
Atemhmanouh
belentioSnarely transferabletohealthfinantries,sPlan." Singapore.
tem maynot be entirelytransferableto other countries,the . 1991."Healthy Famnily,
HealthyNation."Reportof the
systempoints to generallessonsfor health care financing ReviewCommittee onNationalHealthPolicies. Singapore.
strategies.The concept of compulsorysavingshas many - . 1992."Towards BetterHealthCare."MainReportof
advantagesovertraditionalfinancingmethods,but it requires theReview Committee onNationalHealthPolicies.Singapore.
several prerequisites to work. Among them: Ng, A. L. 1988. "The MedisaveScheme."In Proceedings of the
* Ahighleveloflabor forceparticipationin formalemploy- SeminaronEconomicIssuesinHealthCare.Singapore: National
University of Singapore.
ment at sustainablegrowthrates. Phua,KaiHong.1986."Singapore's FamilySavingsScheme."
An effectivesystemof payrollcollectionand contribu- WorldHealth (May):11-12.WorldHealthOrganization,
tion, aswellas efficientfundmanagementand payment. Geneva.
* Awell-developedcomputerizedinformationsystemlink- . 1987."SavingforHealth."WorldHealth Forum8(1):
ingpersonalmedicalsavingsaccountswith hospitalsand
ingpesonl
edialsavng acouts it hopialsan 38-41.
_____. 1991."Privatization andRestructuringof HealthServices
healthcareproviders,in additionto securityfeaturesand in Singapore." Occasional Paper5. InstituteofPolicyStudies,
accountingcontrols. Singapore.
These requirementsmay not be possiblein largecoun- Seow,Adeline,andH. P Lee.1994."FromColonyto CityState:
trieswherethere are wide disparitiesin labor force partic- Changesin HealthNeedsin Singapore from1950to 1990."
ipation rates and where there would be difficulties in JournalofPublicHealthMedicine16(2): 149-58.
enforcing mandatory contributions from employers on a
Singapore.1993."Affordable
HealthCare."WhitePaperpre-
setdoPalmn,Cd.1.Sgpr.
sentedto Parllament,
Cmd.16.Singapore.
regularbasis.Butsavingsschemesare morepolitically
accept- WHO (WorldHealthOrganization).1994.'Health System
able,especiallyiftheyarepersonalizedaccountswith appre- Reforms."Reportof a symposiumsponsoredbythe WHO
ciatinginterestandareportabledespiteemploymentchanges. WesternPacificRegional
Office,Wellington,
NewZealand.

MEDICAL SAVINGS ACCOUNTS AND HEALTH CARE COVERAGE IN SINGAPORE

255
I
Distributors of COLOMBIA
Ltda.
Inloenlace
HAITI
CulureDiffusion
ITALY
Licosa SPA
Sansoni
Commissionaria
NEWZEALAND
EBSCO NZLtd.
ROMANIA
Compani SA.
DeLibrariiBucuresti
SWITZERLAND
PayotServiceInsltutionnel
Libraine
World Bank Carrera6No.51-21
Aereo34270
5, RueCapois
C.P257
ViaDucaDiCalabda,
Caselba 552
Postale
1/1 PrivateMailBag99914
NewMarket Bucharest
no.26,seclor3
Sir.Lipscani C6tes-de-Montberion30
1002Lausanne
Apartado
Tel:(4121)341-3229
Publications SantalfdeBopta,D.C. Poa-au-Prince 50125Firenze Auckland
Tel:(649)524-8119
Tel:(401)6139645
Fax:(401)3124000 Fax:!4121)341-3235
Pricesand credit terms zaryfrom Tel:(571)285-2798 Tel:(5091)3 9260 Tel:(55)645-415
country to country. Consult your Fax:(571)285-2796 Fax:(55)641-257 Fax:(649)524-8067
HONGKONG,MACAO licosa@Itbcc.ft
E-mail: RUSSIAN FEDERATION ADECO EdisonsTechniques
VanDiermen
localdistributor beforeplacing an
order COTEDIVOIRE Asia2000Ltd. UA.httpf/www.Ibc.itlcosa NIGERIA <VesMir>
Isdatelstvo Ch.de Lacuez41
Center Africaines Sales&Circulation
dEdilionetde Diffusion Department PressLimied
University Pereulok
Sa,Lolpachniy CH1807 Bbnay
Seabird House,usa1101-02 JAMAICA ThreeCrowns Jericho
Building Moscow 101831 Tel:(4121)9432673
ARGENTINA (CEDA) Fax:(4121)9433605
o0icinadelLibrolntemacional 04 B.P541 22-28Wyndham Street,Central IanRandle Ltd.
Publishenrs PrivateMailBag5095 Tel:?7095)9178749
Av.Cordoba 1877 Abidjan 04 HongKong 206Ol HopeRoad Ibadan Fax:(7095)9179259
1120Buenos Aires Tel:(225)246510;24 6511 Tel:(852)2530-1409 Kingston6 Tel:(23422)41-1356 TANZANIA
Tel:(541)815-8354 Fax:(225)250567 Fax:(852)2526-1107 Tel:909-927-2085 Fax:(23422)41-2056 SINGAPORE, TAIWAN, Oxford Press
Uniuersuty
E-mail:saleseasIa2000.mom.hk Fax:809-977-0243 MYANMAR,BRUNEI Maktaba Street
Fax:(541)815-8156 NORWAY Asahgate Pte.Ltd.
Asia Pacdic
Publishing POBox5299
CYPRUS URL:hrapJwww.asia2000.com.hk E-mailirpl@colis.com
AUSTRALIA, FIJIPAPUA NEWGUINEA Center for Research
Applied NICInfoA/S 41 KaliangPuddingRoadC04-03 DaresSalaam
SOLOMON ISLANDS, VANUATU, AND Cyprus College INDIA JAPAN BookDepartment GodenWheel Building Tel:(25551)29209
6,Diogenes Street,Engomi Ltd.
AlliedPublishers Easlem BookSeNice Postboks6512Enterstad Singapore349316 Fax:(25551)46822
WESTERN SAMOA
D.A.Inlom1albn Services PO.Box2006 751MountRead 3-13Hongo3-chome, Bunkyo-ku N-0606Oslo Tel (65)741-5166
Nicosia Madras -600 002 Tokyo 113 Tel:(4722)974500 Fax.(65)742-9356 THAILAND
0.48Whiehonrse Rvie
Mcrhami3132 Tel:(3572) 44-1730 Tel:(9144)852-3938 Tel:(813)3818-0061 Fax:(4722)974545 E-mail:ashgate8asianonned.com BooksDistribution
Central
Vidoria Fax:(3572)46-2051 Fax:(9144)852-0649 Fax:(813)318-0864 306SilomRoad
Tel(61(392107777 E-mail:orderssvo-ebos.co.jp PAKISTAN SLOVENIA Bangkok10500
Fax:(68)392107799 CZECH REPUBLIC INDONESIA MirzaBookAgency
URL:hblpeWwwe.bekkoame.orjp/-svt-ebs Gospodarski Goup
VestnikPublishing Tel(662)235-5400
E-mail:service
lidadirectoomra National Center
Inlormalion Pt.IndiraLimRid 65,Shahrah-e-Quaid-e-Azam Dunajskacestas Fax:(662)237-0321
URL:hop://www.dadirect.omrn.au prodejna, 5
Konviktska JalanBorobudur 20 KENYA Lahore 54000 1000Ljubljana
CS- 11357Prague1 PO.Box181 AfricaBookService(E.A.)Ltd. Tel:(9242)7353601 Tel:(38601)133 0347.1321230 TRINIDAD& TOBAGO,
AUSTRIA Tel:(422)2422-9433 Jakarla10320 Quaran House,MfanganoStreet Fax:(9242)7565283 Fax (39661)1338030 ANDTHECARRIBBEAN
Fax:(42 2)2422-1484 Tel:(6221) 390-4290 PO.Boa45245 E-mail:belicd@gvestnik.si Systematics Unh
Studies
Gerold andCo.
26 URL:htpI1atwww.nisucl Fax:(6221)4214289 Nairobi Press
OdordUniversity 9Waft Street
Webthburggasse Curepe
A-1all Wien Tel:(2542) 223641 5 BangaloreTown SOUTHAFRICA,BOTSWANA
Tel:(431)512-47431-0 DENMARK IRAN Fax:(2542)330272 SharaeFaisal Forsingle
titles: WestIndies
Trinidad.
Fa:x(431)51247-31-29 SamfundsLtteratur KetabSaraCo.Publishers POBox13033 OdordUnivershy Atrica
PressSouthem Tel.(809)662-5654
URL:ht1pJvww .gqrold.co/at.onIine Rosenoems AlIe11 KhaledEsiamboli Ave., KOREA, REPUBLIC OF Karachi-75350 PO.Box1141 Fax:(609)662-5654
Daejon TradingCo.Ud. Tel:(9221)446307 CapeTown000 E-mailtobe@tnidad net
DK-1970 C
Frederiksberg 6thStreet
BANGLADESH Tel:(4531)351942 Kusheh No.8
Delafrooz PO.Box34,Youida Fax.(9221)4547640 Tel.(2721)45-7266
MicroIndustfiesDevelopment Fax:(4531)357822 PO.Box15745-733 706SeounBldg E-mail:oup@oup.khi.erum.com.pk Fax:(2721)45-7265 UGANDA
Tehran 44-4Yoouido-Dng,Yeongchengpo-Ku Ltd
Gustro
A.ssistnce Sci;ety(MIDAS) Building
POBox9997,Madhvani
S,Rociety
Houssistae 16 IDA EGYPT, ARABREPUBLIC OF Tel:(9821)8717819;8716104 Seoul PakBookCorporation orders:
Forsubscription
Fax:(98 21)8712479 Tel:(822)785-1631/4 AzizChambers 21 Service
Subscription
Intemational 16/4JinjaRd.
Plot
Hnuse5,Road16 AlAhranm Agency
Distibution E-mail:ketab-sara@neda.net.ir Fax:(822)784-0315 QueensRoad PO.Box41095 Kampala
Danmon rea AlGalaaStreet
TelDa8k2a329427 Cairo Lahore Craighall Tel:(25641)254763
Fax:(8802)811188 Tel:(202)5784083 KowkabPublishers MALAYSIA Tel:(9242)6363222;6360885 Johannesburg2024 Fax:(25641)251468
Fax:(202)5784833 PO.Box19575-511 Cooperative
UniversiftyofMalaya Fax:(9242)6362328 Tel:(2711)880-1440
BELGIUM Tehran Bookshop,Limated E-mail:pbc@brain.net.pk Fax:(2711)880-6248 UNITED KINGDOM
JeanDeLannoy TheMiddleEastObserver Tel:(9821)258-3723 PO.Box1127 E-mail:iss@is.co.za Ltd.
Microinfo
41,Sherr4
Street Fax:(9821)258-3723 Baru
JatanPantai PERU PO.Box3
Av.du Roi202 SA
Desarrollo
Ednorial SPAIN Afon,Hampshire GU342PG
1000Brassels Cairo 59700KualaLumpur
Tel:(202)393-9732 IRELAND Tel:(603)756-5000 Apartado3824 Mundi-Prensa S.A.
Libros, England
Tel:(322)538-5169
Fax:(322) 538-0841 Fax:(202)393-9732 Govemment Agency
Supplies Fax:(603)7554424 Lima1 Castello 37 Tel:(441420)
(8486
Olig antSolkIhair Tel:(5114)295380 28001Madrd Fax.(44 1420)89889
BRAZIL FINLAND 4-5HarcourtRoad MEXICO Fax:(5114)296628 Tel:(341)431-3399 E-mail:wbank@ukminio.demonco.uk
INFOTEC Fax:(341)575-3998 URL:http./www couk
.microinso
Publicac0es Ltda
TocnicaIntemacionais aleeminen Kiriakauppa Dublin2
RuaPeixotoGomida, 209 PO.Bax128 Tel:(3531)661-3111 Av.SanFemando No.37 PHILIPPINES E-mail:libreria@mundiprensa.es
FIN-00101 Helsinki Fax:(3531)475-2670 Col.ToielloGuerra Intemational CenlerInc.
Booksource URL:http:lwww es/
.mundiprensa VENEZUELA
01409SanPaulo, SR
14050Mexico,D.F 1127-A St.
Antipolo Libros,S.A.
Tecni-Ciencia
Teol:5511)259-6644 Tel:(3580) 1214418 Mundi-PrensaBarcelona CentroCuidadComercialTamanco
Fax:(5511)258-6990 Fax:(3580) 121443S ISRAEL Tel:(525) 624-2800 Barangay, Venezuela
E-mail:akatilaus8stockmann.fi Yozmot Ltd.
LRerature Fax:(525)624-2822 MakatiCity Conselde Cent,391 NivelC2
E-rmail:
postmaster(ptimol.ir
URL:hftp:#www.akateeminen.com/ PO.Box56055 E-mail:infotec@4n.net.mx 6507
Tel:(632) 8966501:6505; 08009Barcelona Caracas
URLE htipr//www .uol.br Tel:(343)488-3492 Tel(582)9595547:5035;0016
ual.br
URL:*hItp://wwm Street
3YobnanHasandlar URL:ht1p://an.neo.mx Fax:(632)8961741
TelAvivw61560 Fax(343)487-7659 5636
Fax:(582)F959
CANADA FRANCE E-mail:barcelona@mundiprensa.es
Renouf Co.Ltd.
Publishing WordBankPublicatibos Tel:(9723)5285-397 NEPAL POLAND
5369Canotek Road 66,avenued1alna Fax:(9723) 5285-397 Everest Services
MediaInlematlonal (P)Ltd. Intemational Service
Publishing ZAMBIA
75116Pars GPOBox 5443 Ul.Piekna31137 UnverstyofZambia
Bookshop,
University
Ottawa,OntarioK1J9,3
Tel (613)745-2665 Tel:(331)40-69-30-56/57 R.O.YIntemational Kathimandu 00-677Warzawa SRILANKA,THEMALDIVES GreatEastRoadCampus
Fax:(33 1)40-49-3086 POBox13056 1el:(9771)472152 Tel:(482)628-089 LakeHouse Bookshop PO.Box32379
Fax:(613)745-7660
TelAviv61130 Fax:(9771)224431 Fax:(482) 621-7255 100,SirChrtampalam Mawatha
Gardiner Lusaka
renoufbooks.com
order.deptO
E-mail:
URL:hoprDe .dwwrenoudbooks.com GERMANY Tel:(9723) 5461423 books%ips@ikp.atm.com
E-mail: pl Colombo 2 Tel:(2601)252576
Fax:(9723)5461442 NETHERLANDS Tel:(941)32105
URL:htip:/Iwww.ipscg.wawpVips/export/ Fax.(2601)253952
UNO-Verlag
CHINA PoppelsdoderAllee00 royil@netvision.net.il
E-mail: DeLindeboorrilnOr.Publikates Fax (941)432104
& Economic
ChinaFinancial 53115Bonn PO.Box202 PORTUGAL E-mail:LHL@si.ianka.net ZIMBABWE
House Tel:(49228)212940 Palestinian Cool
Authorily/Middlo 7400AEHaaksbergen LivraitaPortugal Longmnan )Pvl.(Ltd.
Zimbabwe
Publishing
8, DaFoSi DongJie Fax:(49228)217492 Services
IndexInfornation Tel:(3153)5744044 Apartado2081 SWEDEN Toude Road,Ardbennie
Jerusalem Faa:(3153)572-9296 RuaDo Camno 70-74 AB
Wennergren-Williams PO.BoxoST125
nBaio
nPO.B.19502

Tel:(810)63-257 GREECE Tel:(9722)6271219 E-mail:lindeboodlwoddonline.nl 1200Lsbon PO Box1305 Southenon

6401-7385
Faa:(8610) PapasonoiouS.A. Fax:(9722) 6271634 URL:hup:/Awww .woddonline.nl-lindebooTel.(1)347-4982 S-17125 Solna Harare
35,StoumaraSIr. Fax:(1)347-0264 Tel:(468) 705-97-50 Tel:(2634)6216617
Fax:(460) 27-00-71 Fax.(2634)621670
10082 Athens %2ur9
Tel:(301)364-1826 E-mail.mail@wwi.se
Fax:(301)364-8254
Recent World Bank Discussion Papers (continued)
No. 331 CaseStudiesin War-to-Peace
Transition:The Demobilizationand Reintegrationi
of Ex-Combatantsin Ethiopia,Namiibia,
and Uganda. Nat J. Colletta, Markus Kostner, Ingo Wiederhofer, with the assistance of Emilio Mondo, Tairni
Sitari, and Tadesse A. Woldu
No. 333 Participationin Practice:TheExperienceof the WorldBankand Other Stakeholders.Edited by Jennifer Rietbergen-
McCracken
No. 334 Managing Price Risk in the Pakistan Wheat Market. Rashid Faruqee and Jonathan R. Coleman
No. 335 PolicyOptionsfor Reform.of ChineseState-OwnedEnterprises.Edited by Harry G. Broadman
No. 336 Targeted Credit Programs and Rural Poverty in Bangladesh. Shahidur Khandker and Osman H. Chowdhury
No. 337 The Roleof FamilyPlanningand TargetedCreditProgramsin DemographicChangein Bangladesh.Shahidur R.
Khandker and M. Abdul Latif
No. 338 Cost Sharing in the SocialSectorsof Suzb-Saharan
Africa:Impacton thePoor.Arvil Van Adams and Teresa
Hartnett
No. 339 Putblicand PrivateRolesin Health:TheoryandFinancingPatterns.Philip Musgrove
No. 340 Developingthe NonfarmSectorin Bangladesh: LessonsfromOtherAsian Couintries.Shahid Yusuf and Praveen
Kumar
No. 341 BeyondPrivatization:The SecondWaveof Telecommutnications
Reformsin Mexico.Bjbrn Wellenius and Gregory
Staple
No. 342 EconomicIntegrationand TradeLiberalizationin SouthernAfrica:Is Therea RoleforSouth Africa?Merle Holden
No. 343 Financing Private Infrastruicture in Developing Couintries.David Ferreira and Karman Khatami
No. 344 Transportand the Village:Findingsfrom African Village-LevelTraveland TransportSutrveysand RelatedStudies.Ian
Barwell
No. 345 On the Roadto EU Accession:FinancialSectorDevelopmentin CentralEutrope.
Michael S. Borish, Wei Ding, and
Michel Noel
Africa:Findingsfroma Seven CountryEnterpriseSurvey.
No. 346 Structural Aspectsof Manufacturingin Sutb-Saharan
Tyler Biggs and Pradeep Srivastava
No. 347 Health Reform in Africa: Lessonsfrom Sierra Leone. Bruce Siegel, David Peters, and Sheku Kamara
No. 348 Did ExternalBarriersCautsethe Marginalizationof Sub-SaharanAfricain WorldTrade?Azita Amjadi
Ulrich Reincke, and Alexander J. Yeats
No. 349 Sutrveillance
of AgriculturalPriceand TradePolicyin LatinAmericaduring MajorPolicyReforms.Alberto Valdes
No. 350 Who BenefitsfromPublicEducationSpendingin Malawi:Resultsfromthe RecentEducationReform.Florencia
Castro-Leal
No. 351 From UniversalFoodSuibsidiesto a Self-TargetedProgram:A CaseStudy in ThinisianReform.Laura Tuck and Kathy
Lindert
No. 352 China'sUrbanTransportDevelopmenit
Strategy:Proceedingsof a Symposiuzmn
in Beijinig,Noveember
8-20, 1995.
Edited by Stephen Stares and Liu Zhi
No. 353 TelecommuinicationsPoliciesfor Sub-Saharan Africa. Mohammad A. Mustafa, Bruce Laidlaw, and Mark Brand
No. 354 Saving across the World: Puizzles and Policies. Klaus Schmidt-Hebbel and Luis Serven
No. 355 Agricuilture and German Reuinification. Ulrich E. Koester and Karen M. Brooks
No. 356 Evaluiating Health Projects:Lessonsfrom the Literature. Susan Stout, Alison Evans, Janet Nassim, and Laura Raney,
with substantial contributions from Rudolpho Bulatao, Varun Gauri, and Timothy Johnston
No. 357 Innovationsand Risk Taking:TheEngineof Reformin LocalGovernmentin LatinAmericaand the Caribbean.
Tim Campbell
and InvestmenfCompanies.Anjali Kumar, Nicholas Lardy, William
No. 358 China'sNon-BankFinancialInstituitions:Trust
Albrecht, Terry Chuppe, Susan Selwyn, Paula Perttunen, and Tao Zhang
No. 359 The Demandfor Oil Produtctsin Developing Countries. Dermot Gately and Shane S. Streifel
of a Conference
No. 360 PreventingBankingSectorDistressand Crisesin LatinAmerica:Proceedings hieldin
Washington, D.C., April 15-16, 1996. Edited by Suman K. Bery and Valeriano F. Garcia
No. 361 China: Power Sector Regutlation in a Socialist Market Economy. Edited by Shao Shiwei, Lu Zhengyong, Norreddine
Berrah, Bernard Tenenbaum, and Zhao Jianping
Union, and Other Coluntries.
No. 362 TheRegtulationofNon-BankFinancialInstitittions:The United States,the Eutropean
Edited by Anjali Kumar with contributions by Terry Chuppe and Paula Perttunen
No. 363 FosteringSuistainableDevelopment:TheSectorInvestmentProgram.Nwanze Okidegbe
No. 364 IntensifiedSystems ofFarmingin the Tropicsand Sutbtropics.
J.A. Nicholas Wallis
THE WORLD BANK

1 ISI I SIT-cct. .\\.

\\.likliYtlEml I ).(. 2(1433.l S \

ILIcph4)flc: 2012-477-1234

IFc'simic: 202-477-o3,)I
ICl\: NCI (04145\\(I(I.I)I, NIs
\I( I 248423n \\MlRl.11)111\Nk

\\\SIlI (IOI(

\,1
\i.l.\\.1h: IIIt p: ENN
X\ altIII-,l.I-8
IT-,,,II) 3
I '-,l:1:11'al \ ml,l-, d(,, ll15:,i 1-/ ,'

ISBN 0-8213-3964-8