You are on page 1of 32

\bHdSlc +


Tradable Water Rights Even with adequate

institutions, administrative
methods of allocating / ater
A Property Rights Approach result in ineffcientoutcr)res.

to ResolvingWaterShortages But a systemoftradabl

and PromotingInvestment waterrights
effective institLt1tons'sc'q%.
many of the problems . iat
Paul Holden administeredso!utionr'.-iilto
Mateen Thobani dealwith.

The WorldBank
Latin America and the Caribbean
Technical Departmeent
Economic Adviser's Unit
July 1996

Summary findings
In most countries, the state owns the water and hydraulic changes in water allocation in response to changing
infrastructure, and public officials decide who gets water demand for water and can stimulate investment and
rights, how the water is to be used, and how much will employment as investors are assured of access to water.
be charged for its use. But there is ample evidence that Moreover, agricultural production will become more
water allocation by administrative edict has resulted in economically efficient as output will reflect the true
costly, large-scale inefficiencies in the supply and use of scarcity of water rather than the frequently distorted
water. Secure property rights have been shown to have a prices set by administrators subject to political lobbying.
powerful positive effect on investment and efficiency, but Because of water's unique characteristics, an effective
only a few countries have tried to introduce tradable tradable water rights system is not easy to introduce and
property rights to water, thereby taking advantage of the water markets are not a panacea. But these same
allocative efficiencies of a market to assign water characteristics make administrative solutions to water
resources among users. allocation difficult - water markets rarely make them
Holden and Thobani compare administered systems of worse. Chile's experience and the demonstrated
water allocation with a system of tradable water rights. superiority of markets over administrative means of
Using an approach derived from the literature on resource allocation suggest that water markets are likely
property rights and new institutional economics, they to be a better alternative in most water-scarce countries.
argue that even with an adequate institutional To ensure an effective water market, attention should
framework, administrative methods of allocating water be paid to:
result in inefficient outcomes. Water is used wastefully. - Ensuring stakeholder participation in designing and
Public hydraulic projects are poorly conceived, implementing the new legislation.
implemented, and operated. And the'systems have failed * Deciding rules for the initial allocation of rights and
to protect the environment or make water accessible to on how new rights would be allocated.
the poor. As urbanization spreads and pressures on the * Establishing a public registry and block titling.
water supply get worse, such solutions are likely to * Setting up or strengthening water user associations.
become even more untenable. * Protecting against the development of potential
Chile's experience in water-scarce areas demonstrates monopolies.
that tradable water rights can benefit the poor and * Ensuring that trades do not infringe on the water
increase user participation on water allocation and rights of existing users.
investment decisions. They can allow rapid voluntary * Establishing appropriate environmental laws.

This paper - a product of the Economic Adviser's Unit, Latin America and the Caribbean, Technical Department - is part
of a larger effort in the department to focus attention on the effect of establishing secure property rights and using markets
to allocate resources. Copies of the paper are available free from the World Bank, 1818 H Street NW, Washington, DC
20433. Please contact Patricia Mendez, room 18-451, telephone 202-473-8893, fax 202-676-9271, Internet address July 1996. (23 pages)

the exchange
The PolicyResearchWorkingPaperSeriesdisseminatesthe findingsof workin progressto encourage of ideasabout
development issues.An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The
papers carry the names of the authors and should be used and cited accordingly. The findings, interpretations, and conclusions are the
authors'own and shouldnot be attributedto the WorldBank,its ExecutiveBoardof Directors,orany of its membercountries.

Produced by the Policy Research Dissemination Center

Tradable Water Rights:

A propertyrights approach to resolving

water shortagesand promoting investment

Paul Holden and Mateen Thobani


1. Water, whichin manycultures has an almost mystical significance,has often been

the cause of disputesboth betweenindividuals and countries. Conflictshave on occasion
escalatedinto violenceand in the case of countries, wars. The aggressionthat such disputes
has provoked rests on the life-givingand agricultural characteristicsof water. The fact that it
is essentialhas frequentlybeen used to justify heavy state interventionin the granting and
administrationof rights to the use of water. In most countries, the stateowns the water and
hydraulic infrastructureand public officials decide on who gets water rights, on the purpose
for whichthe water is to be used, and on the price to be charged for its use. However, there
is ampleevidencethat water allocationby administrative edict has resultedin large scale
costly inefficienciesin the supplyof water and in its use. Althoughthere is substantial
documentationthat secureproperty rights have a powerful positiveeffect investmnent and
efficiency(see, for example,Demsetz 1967, Alchian and Demsetz 1972,and Barzel 1989),
only a few countrieshave tried to introduce tradable property rights to water, thereby taking
advantageof the allocativeefficienciesof a market to assign waterresourcesamongusers.

2. The paper comparesadministeredsystems of water allocationwith a system of

tradable water rights. Usingan approach derived from the property rightsand new
institutionaleconomicsliteraturethe paper argues that even with an adequateinstitutional
framework, administrativemethodsof allocating water result in inefficientoutcomes.
However, tradablewater rightscombined with effective institutionssolvemany of the
problems that administeredsolutionsfail to deal with.

3. After describinghow water rights are defined, assigned, and enforcedunder

conventionalwater rightsregimes, the paper reviews the experienceof suchregimes and
evaluates on-goingmeasuresto improve water management.Next, the paper focuses on
tradable water rights, givingtheir characteristics and showingthem to be a form of property
right. It then discusseswhy and how some countries are promotingtradablewater rights
regimes and discussestheirexperience and potential. Finally, the paper reviewsconceptual
and practical problemsin establishingtradable water rights and suggestshow to design and
implementa system of tradablewater rights in order to resolve theseproblems.

Conventional Water Rights Regimes

4. In most countrieswhere water is scarce or costly to access, systemsof rights for

water use have evolvedimplicitlythrough custom or explicitlythoughbodiesof law and
regulations(or both). Thesewater rights specify how water in a river is to be divided
between alternativeuses suchas industrial use, domestic water supplyand agriculture, as
well as between individualwater users within a sector. Water rights are generallybased on a

variant or combinationof the followingthree systems: riparian rights, prior (appropriative)

rights, and public allocation(Sampath1992).

How Water Rights are Defined, Assigned and Enforced

5. Under the riparianrights doctrine, anyone who possesses land next to a flowing
river or stream may take its water as long as enough is left for downstreamusers.
Diversions of water to locationsnot adjoining the river or stream are prohibited.Such
systemstend to occur in areas where water is relatively abundantand where strict definition
of rights is not crucial (France, eastern part of the United States). In addition,even where
surface water rights are determinedby other means, countries typicallyallow groundwater
rights to accrue to those that own the land overlying the aquifer.

6. Prior rights are based on the appropriationdoctrine, under which the water right is
acquired by actual use over time. Diversions of water are permitted and quotasare allocated
to specifiedparties on a first-come,first-served basis and are subject to the "use it or lose
it" rule. This is the main systemprevailing in the arid western part of the UnitedStates-
those that establisheda beneficialuse early were given senior rights (early settlersand
farmers) over those that establishedthem later (cities).

7. Public allocationinvolvespublicly administereddistribution of water. Under this

system, public authoritiesdecide how to allocate water using guidelines or lawsestablishing
priorities and often specify the uses to which the water can be put. Most developing
countries follow variants of this approach. Although there is often a charge for wateruse
(usuallybased on size of the irrigated area), the water rights themselvesare obtainedwithout
charge, with irrigation rights linked to land.

8. Water rights are typicallydefined in one of several ways: volumetricallyas a share

of the stream or canal flow or of the water available in a reservoir or lake; or in terms of
shifts or hours of availabilityat a certain intake. In some cases, the water rights may be
defined as a combinationof the above or be conditionalupon water availability.For
instance, water going into a canal may be based on a share of the river flow whereaswater
going to individualfarmersmay be based on hours of water availableat an intakepoint.
Some rights are volumetriconly if there is a certain level of water in the river; otherwise,
they are proportional. Similarly,rights may be defined as a share of the excesswater flow
above a given stream flow (definedin liters/second) or above a certain level of waterin a
lake or reservoir (definedin cubic meters). Certain junior rights under an appropriative
rights regime may be exercisedonly if senior rights have been met. Rightsmay be
consumptiveor non-consumptive:while consumptiverights have no obligationto return any
quantityof water to a river, non-consumptiverights may face an obligationto return the
same quantityand quality of water to a specifiedlocation. Generally, only hydropower
companieshave such rights. Worldwide, 69 percent of water is used in agriculture,23
percent in industry, and only 8 percent for domestic purposes.

9. The measurement infrastructure varies from simple dividers within a stream or canal
that divert water according to established ratios to measuring devices that may continuously
record volumetric water flow and transmit the information instantly to computers at a central
monitoring station. The operation and maintenance of the water distribution system and the
enforcement of water rights is increasingly done by water user associations and communities
rather than public authorities. Similar, in cases of dispute, the water user association is
typically the first arbiter.

Experience with Administrative Methods of Water Allocation

10. The track record of such administered systems of water allocation has not been
impressive. Despite growing water scarcity and the high costs of hydraulic infrastructure,
water is typically underpriced and used wastefully, the infrastructure is frequently poorly
conceived, built, and operated, and delivery is often unreliable. At the same time, there are
high fiscal costs stemming from the construction of hydraulic infrastructure; from the
institutional bureaucracy to support the design and execution of the projects and to set and
collect water tariffs; and from the cost of operating and maintaining the system. Many large
multipurpose hydraulic projects (irrigation, hydropower, flood control, urban use, etc.) were
undertaken on political rather than economic grounds (see Box 1). Costs tend to be high

Box 1. Public Hydraulic Projects

Governmentsfrom both developedand developingcountrieshave investedheavily in

publichydraulicprojects.However,the results have oftenbeenwell belowexpectations.For
example,by the end of 1993,the Governmentof Peru had spent $3.4 billion (in constant1993
dollars)on nine coastalmultipurposeprojects. Although some of these projectshad beenin
executionfor over two decades,they had realized only 6.6 percent of their plannedexpansionin
irrigationand none of their plannedhydropowergenerationcapacity.Whilethe primary
justificationof these projectswas irrigation,the estimatedcost per hectareof these schemesat
completionrangedfrom $10,000to $56,000, even while irrigatedland in these areassells for
about $3,000(see WorldBank 1995).
Waterresourcesdevelopmentin Asian countriestypicallyaccountsfor 20 to 25 percent
of total publicinvestment.However,in Sri Lanka,the MahaweliDevelopmentProgramalone,at
its peak,absorbed6 percentof GDP and 44 percent of public investmentexpenditure!The costs
of landdevelopment,excludingheadworks,were $12,000-$15,000per hectareat 1987prices,
comparedto $3,000-$5,000in otherAsian countries.Evenwith a doublecroppedpaddy,the
economicreturnsfrom the earlier,and cheaper, projectswere foundto be lowor negative.Since
not even O&Mcosts could subsequentlybe recovered,new settlersbenefitedfrom massive
subsidiesand their spatialdistribution,if anything,aggravatedsocialtensions (see Frederiksen,
Berkoff,and Barber1993).Similarly,the performanceof Pakistan's 13,000publictubewellshas
been poor. Despitethesetubewellsreceiving 55 percent of total O&M expenditureseventhough
they accountfor only 10 percentof irrigationwater supplies,their pumpingcapacitydeclinedan
averageof 4-6 percentannually,with 20 to 45 percent of publictubewellsnot operatingat any
one time as comparedto 10percent of private tubewells(see WorldBank 1993).

because of: inappropriatedesign,stemmingin part from poor studiesdone prior to start-up;

long gestationperiodsresultingfrom funding shortfalls due to changinggovernmentpriorities
and poor capitalprogrammingand budgeting;few managerialincentivesto controlcosts;and
reported corruptionthat typicallyinvolveskickbacks from constructioncompanies.

11. In many countries,for cultural,political or religious reasons,wateruse is not priced at

all, so there is little incentiveto conserve. When it exists, water charges,are weUlbelowthe
cost of developingwaterresources.Therefore, it is not unusual to find cities in arid areas
rationingwater and foregoingpotentiallylucrative activitieswhileneighboringfarmersgrow
low-valuewater-intensivecropssuchas rice and alfalfausing inefficientirrigationtechnologies.

12. Moreover, governmentcontrol of water has favored the relativelywealthyand has

not been effectiveat ensuringaccess of the poor to water. In many cities in developing
countries it is the poor that are excludedfrom piped municipalwater and must resort to very
expensive private water truckersto meet their daily needs. A review of water vendingin
sixteen cities (World Bank 1992)shows that the unit cost of vended water is 4 to 100 times
higher than water from piped city supplies (the median cost was 12 times higher).Similarly,
influential farmers manageto get easier access to water rights, which are obtainedwithout
charge, often at the expenseof reducing availability for the poor, and for whoseuse farmers
pay only a nominalcharge.

13. Althoughpublic controlover water is thought necessaryto addressenvironmental

problems, governmentsusuallyfail to maintain water or soil quality. Unsafewater causes
water borne diseasesthat result in the deaths of 3 million people annuallyand render sick
more than a billion more. The dischargeof untreated industrial waste, the runoffof
agricultural chemicals,and poor land use practices in agriculture, forestry, and mining
causes widespreaddegradationof land and water resources (World Bank 1993).Water
logging and salinizationhave destroyed millions of hectares of fertile agriculturalsoils. In
Pakistan, extensivewater loggingand secondary soil salinizationhas resultedin an estimated
10 percent of its irrigationsystemcovering some 13.5 million hectares to be affectedby
salinity (Frederiksen,Berkoff,and Barber 1993). Sometimespublic irrigationprojects
themselveslead to salinization.Until the 1960s, the Aral Sea in Russiawas environmentally
stable with a thrivingcommercialfishery. The massive diversion of the two largestrivers in
Central Asia to expand irrigatedcotton production eventuallydried up the rivers and shrank
the lake by 66 percent. Salinityincreased, soils became waterlogged,fish spawninggrounds
dried up, and the fisherycollapsed.An ecological catastrophedevelopedas windspicked up
salt and pesticidesfrom the dry lake bed, caused salt and pesticide storms, and ruined the
productivity of farmlandover a wide area (World Bank 1993).

Measures to Resolve Water Shortages and Improve Water Use

14. In attemptingto addressthe problems described above, there is increasedattention

being paid to managementreformssuch as though better planningand changesin bureaucratic
structure and pricingpolicy.This approachis perhaps best reflectedin a recentpolicypaper

(World Bank 1993)whichtakesthe view that because of imperfectionsin watermarkets,they

shouldbe eschewedin favor of comprehensiveadministeredsolutions."At theheart of the
approachis the developmentof a comprehensiveanalyticalframeworkfor waterresources
management.Waterresourcesshouldbe managed in the context of a nationalwaterstrategy
that reflectsthe nation's social, economic,and environmentalobjectivesand is basedon an
assessmentof the country'swaterresources. The assessmentwould includea realisticforecast
of the demandfor water, based on the projected populationgrowth and economicdevelopment
and a considerationof the optionsfor managingdemand and supply,takinginto account
existinginvestmentsand those likelyto occur in the private sector" (WorldBank1993,p. 41).
Proponentsof such solutionsclaimthat nationalplans such as the one describedabovewill
solve problemsof allocationandprioritizationand will allow long term investmnentsto be
made whichwill ensurethat the demandand supply of water will remainin equilibrium.

15. Essentialto the effectiveimplementationof this form of solutionis that usersbe

charged the opportunitycost of water which "provides a measure of the scarcityvalueof water
to society, thus highlightingany cross-sectoraldifferencesin value, takingintoaccount
society's multipleobjectivesand water's multiple uses and interdependencies,"(WorldBank
1993, p. 43). Not only wouldit be difficultto estimate such prices acrossuses,regionsand
over time, it may be politicallydifficultto raise water charges to levelsthat reflectthe scarcity
value of water. Presently,water chargesbarely cover the cost of operatingandmaintainingthe
water deliverysystem, let alonethe cost of building the infrastructure.Moreover,for
irrigationwater at least, a sharp increasein water charges would imply an expropriationof
property rights since landpricesalready reflect the access to cheap water.

16. In evaluatingthe efficacyof such approaches to the allocationof waterresourcesand

in comparing them to a tradablewater rights system, care must be taken to ensurethat the
same things are being compared.The usual approach of strong advocatesof administered
solutions is to point out the existingmarket imperfections, which reducethe efficiencyor
effectivenessof water markets, and then to compare this situationwith an administrative
solutionwhich involvesa far-seeing,incorruptible, influence-freeadministrativebody that is
able to estimate alternativerates of return between water investmentsand investmentsin
other parts of the economyand then be able to design and implementthe correctpolicy. In
reality, administrativebodiesare often captured by interest groups, are not knownfor being
far-sighted, are unableto estimatefuture demands with any accuracy, are unableto set and
collect appropriatewater charges, and almost always have more imperfectionsthan the
markets that they are supposedto replace.

17. A further drawbackof the administrativeapproach is that it tends to favor large scale
investmentsover water conservation;there are few rewards for administratorsfrom
painstakingimprovementsin water efficienciesvia better pricing policies.Ratherthe
glamour of large projects and the attendantpublicity and power that they bring providefar
stronger incentives.In contrast, attempts to set prices that reflect the true cost of water
provision are unpopular.Withoutsuch prices, incentive for users to conservewater are

18. In conclusion,while the integrated water resourcesmanagementapproachhas worthy

goals, finding incorruptibleand competent administrativebodies that are able to accurately
estimate water demandand supply over time and then designand implementinvestmentsand
pricing policies effectivelywill be very difficult, especiallyin developingcountries.Despite
market imperfections,an alternativeapproach using a systemof secure and tradablewater
rights is worth seriousconsideration.

Tradable Water Rights Regimes


19. FormalMarkets:The key characteristicsof formalsecuretradablewaterrights is that

the rights are independentof land and can be traded separatelyfrom land withina legal and
institutionalframework.As such they are property rightsto water (Box2). Ideally,the water
rights shouldbe sold at freely negotiatedprices to anyonefor any purpose. However,
sometimescountriesimposerestrictionssuch as requiringthe buyer to use it for some
beneficialpurposeor that they only be sold to a public agencyat an administrativelyset price,
thereby weakeningthe propertyright associatedwith the water right (see the examplebelow
from the westernUnitedStates).There may also be other restrictionsrelatingto waterquality,
to ensuringthat a certainminimumflow in a stream or river is maintainedfor environmental
or recreationalreasonsand to protectingthe water rightsof third parties.'

20. Informalmarkets,wherebyindividualsor groups of water rights holderssell water to

other users at freelynegotiatedprices, have evolved spontaneouslyin many countriesas a
responseto the failureof public allocationof water. A 1990surveyof surface watersystemsin
Pakistan(PakistanWaterand Power DevelopmentAuthority 1990)foundactivewater trading
for irrigationwaterin 70 percentof the watercoursesstudied.In India, an estirated one-half
of the area irrigatedby tubewellsbelongs to farmers who buy water (Shah 1991).While such
marketshelp resolvewatershortages,the fact they are not supportedby existinglawslimit
such transactionsto spot salesof water or to the sale (lease)of water for a singleyear rather
than to permanentsales of water rights. The difficultyin enforcingsuchcontractshas also
tendedto confinethe transactionsto those within the samesector, oftenbetweenneighboring
farmers. The lackof long-termsecure access to water under such a systemalsodiscourages
investmentin activitiesthat require access to large quantitiesof water. Thus suchinformal
water marketscan realizeonly part of the potentialgains from trade and do not strictlyfit the
mold of tradablewaterrightsregimesthat are analyzedin this paper.

I. One of main sources of hydrological third-party effects stems from "return flows", which is the water
returned to the ground after use (e.g., the irrigation water that was not fully absorbed by the crop), and which
may infiltrate down to an aquifer that joins a water source. If another user has rights to this water and if the
upstream user were to sell all the water he received to someone whose return flow was different, the rights of
the downstream user would be diminished.

Box 2. What Are Property Rights and Why Are They Important?

A propertyrightover an asset consists of the rightto consume,earn incomefromor sell the

asset.The processof establishingpropertyrights involvesenshrininglegalownership.Property
rightsencompassbothphysicalrightswhich allow ownersto havepossessionof assetsthat they
ownas wellas legalrightswhichallow effectiverecourseto the legalsystemif thephysicalrights
are violated.Formalmethodsof establishingpropertyrights arefoundin that partof the legal
systemwhichdetermineshowpropertycan be definedand exchanged.Normallythereneedsto be
an efficientprocessfor registeringpropertyrights which includesensuringthat nobodyelse has
prior claims.Legalenforcementmechanismare importantin dealingwithrightsthat mayhave
beenviolated.However,thestrengthof ownershipis oftendilutedby constraintsarisingfromthe
natureof theasset or fromthe statewhich might place restrictionsuponits use.In addition,if the
costs of enforcingrightsarehigh becauseof inadequateprovisionof suchpublicgoodsas policing
and the legalsystem,the strengthof propertyrights will be weakened.
A key conceptthat affectsthe strengthof propertyrights is that of transactionscosts,which
are the costsof defining,protectingand exchangingpropertyrights.Theeffectsof high transactions
cost on propertyrightsand exchangehave only recentlybeenrecognizedfully(see Holdenand
Rajapatirana1995).Tradeswhichwould increasewelfaremaynot occurif theprocessoftrading is
costlyor if rightsthataretradedcannot be protected.Therefore,institutionsthat lowertransactions
costs are oftennecessaryto help in the efficient functioningof marketsand the marketfor wateris
no exception.Once rightsare defined and can be traded and transactionscostsare low, assets will
be employedin the most efficientmanner. In this way societalwelfarewill be maximized.

Countries Promoting Tradable Water Rights

21. To allow water users to secure water on a permanent basis as well as to facilitate
water leasing, some countries have begun to pass legislation to permit secure and well-
defined tradable water rights. Chile's 1981 Water Code established tradable water rights in
order to strengthen property rights, allow flexibility in water use, and empower water users
by requiring their consent to any reallocation of water. Under this law, the State grants
existing water users (farmers, industrial furms, water and power utilities) property rights to
both surface and ground water without charge. New and/or unallocated water rights are
auctioned. The water rights are separate from land and their private property status is based
on the property laws of the Civil Code. Except for a few restrictions, they can be transferred
or sold to anyone for any purpose at freely negotiated prices. As with land, market forces
determine the allocation and use of water, once assigned.

22. Water rights are acquired by being recorded in a public registry as either
consumptive or non-consumptive, temporary or permanent. Permanent consumptive rights
are defined in volumetric terms unless there is insufficient water to satisfy all water rights
holders, in which case the water is distributed proportionately. Temporary (contingent)

consumptiverights, which are particularlyuseful when there is storage availability,can only

be honored if all permanentconsumptiverights have been met. Non-consumptiverights,
used for hydropowergeneration,grant the owner the use of water as long as it is returnedto
its source at a specifiedlocationand quality. The bulk of the estimated300,000owners of
water rights hold consumptiverights, with agriculture accountingfor 89 percentof such
rights (Rios and Quiroz 1995). The monitoring, distribution, and enforcementof water
rights is carried out by wateruser associationsat the level of the river basin,underground
aquifer (for groundwater),primarycanal, and secondaryor tertiary canal. Exceptfor a few
large dams and their associatedmain canals, all hydraulic infrastructureis ownedand
operatedby water users themselves.

23. Chile's experiencewith watermarketshas been very positive. Waterusersare

particularlypleasedby the flexibilityand control over their water rights. In the arid areasnorth
of Santiago,there have been manymutually beneficial sales and leases of water, resultingin
a voluntarytransfer of water to more productiveuses (see following subsection).However,
in the high-rainfallareas south of Santiago,there have been few trades sincethe transactions
cost of registeringthe rights and conveyingthe water is greater than the gains from
transferringthe water. There are also few transactionsin the main canal of the Maiporiver
near Santiagobecause this canal uses fixed flow dividers and the cost of changingthe water
intakes for transfers is prohibitive.

24. For reasons largely unrelatedto water markets, Chile still has problemsin water use.
For example, Chile continuesto suffer from water quality problems since regulationsto its
tough 1990 environmentallaw were never issued. Also, because the obligationsof non-
consumptiverights holders to releasewater for consumptivepurposes at certain timeswere
not clearly defined, conflictsbetweenthe recently privatized hydropowercompaniesand
farmers have developedin some areas. Some shortcomings in the law have alsoenabledone
hydropowercompanyto obtain huge volumes of non-consumptiverights withoutcharge.
Despite these problems, Chile has far fewer conflictsand makes better use of its water as
compared to its neighbors.

25. Mexico. As the Mexicanagriculturaleconomybecamemore market-oriented,policy

makers realizedthat the full benefitsof the economicliberalizationcouldonly comewith
secure and tradablewater rightsthat offered the flexibility of water use to respondto changing
prices and demands.Accordingly,under the 1992 water law, and its 1994regulations,users
may converttheir existingprecariouswater rights to more secure tradable "concessions"with
a maturityof five to fifty years, with the norm being thirty years to ensure securityof the
water right. However,the rightsare not as secure as in Chile. Under the Mexican
Constitution,all water belongsto the Nation and this property right is perpetualandnon-
transferable.The law also mentionsthe possibilityof forfeiture for reasonsof publicinterestif
the water has not been used "efficiently;"or if it has not been exploitedfor threeyears.
Althoughthe rights are specifiedin volumetricterms, in practice the rightsare proportional
since the water user associationsare to allocatedeficits or surplusesproportionatelyacrossall
existingrights. The rightsfor both groundand surface water are recordedin a publicregistry.

26. By the end of 1995,the NationalWater Commissionhad processedwaterrights

applicationsaccountingfor about85 percent of volumetricwater rights. During1995,there
was widespreadleasingand sellingof both surface and groundwater rightsin Mexico'swater-
scarce regions. Even prior to 1994, water trades were common.However,they were limited
largelyto informalsales of water for a year or season,typicallybetweenneighbors.Sincesuch
trades were illegal,albeittoleratedby public authorities,permanentsales of waterrightswere
rare and little considerationwas given to third-party water rights beingaffectedby such trades.
The new law has facilitatedsuch leasingand allowedpermanentsales of waterrightswith
better protectionfor aquiferrechargeand third party rights. Most of the recentwatersales
would either not have been undertakenunder the previousregime, thereby inhibitingprivate
investment,or would have occurredsurreptitiously,therebydepletingaquifers.The bulk of
recentwater trades involvefarmers sellingto industrialusers, water companiesor more
efficientfarmers, therebyencouraginginvestmentin more productiveactivities.It has also
allowedunprofitablefarmersto reducetheir farmingdebts and to work as laborerson more
efficientfarms or to seek alternativeemployment.

27. Peru. The motivationfor Peru's water reform came with the realizationthat existing
water legislationand policies, which had caused serious problems, were poorly suitedfor a
future of tighter fiscal constraintsand weakenedpublic institutions. By 1992, following
several years of virtually no public spending for maintenanceor rehabilitationof public
irrigation structures, many irrigation systems faced a high risk of failure. Waterdelivery
became more irregular, qualitydeteriorated and water conflicts grew. There was also
widespreadwater theft. Even in areas where water was scarce, it continuedto be used
wastefully.Thus, while water was rationed in Lima, the water companyincurredhigh water
losses, and farmersjust outside Lima continued to grow low-value, water-intensivecrops. In
addition, the threat of having the State expropriate water rights for higher priority uses
discouragedmany worthwhileinvestmentsthat required assured suppliesof water.

28. To address these problems,the Ministry of Agricultureproposeda new water law

modeledalong the lines of the 1981 Chilean water code. Under the proposedPeruvianlaw,
existing water users are to be given rights to water without charge. Rightsto new or
unallocatedsurface water are to be distributed via public auction. The rights may be traded
at freely negotiatedprices provided that the trade would not reduce water availabilityto
others and that there is enough water to maintain a minimnum ecologicalflow and to maintain
the accustomedquality of life in cities and towns. Rights may also be mortgagedor leased.
The law prohibits altering water quality to the detrimentof flora or fauna; however,rather
than proposing specificsanctionsand fines, it defers to the EnvironmentalCode and
EnvironmentalAuthorityto set and enforce water quality standards. (World Bank 1995.)

29. Under the draft law, water rights are to be acquiredby being recorded in a public
Water RightsRegistry, specifying,inter alia, the flow or volume (whichmay be specifiedin
terms of percentageof stream flow or in shifts); the point at which the water will be
diverted; whetherit is for consumptiveor non-consumptiveuse and whetherit is for
permanentor temporaryuse; the point and form in which the water will be returnedto the

river system; and the amountpaid for the rights. The law also establishesa propertytax on
water rights. In contrastto the current law, the new rights do not have to be used for any
specific purpose, there are no priorities among water rights for differentpurposes,and the
water right is separatefrom the land right for both surface and ground water.

30. Passage of this lawhas been delayed partly because of Congressionalconcernson

how the initial allocationwouldbe conducted. The delay also stemmedfrom insufficient
involvementof users in the designof the proposed law and from oppositionfrom the few
that benefit from maintainingthe status quo.

31. The western UnitedStates. Because of the shortage of water in the westernUnited
States, a system of propertyrightsto water based on the prior appropriationdoctrine
evolved: those that first divertedand established beneficial use of the water obtainedprimary
rights to it. Successiveclaimantscould only obtain rights that were contingenton those with
prior rights having receivedtheir allocations.

32. Althoughwater rightsregimesvary widely between states, their commoncharacteristic

is that the uses to whichwateris put canmotbe changed withoutauthorizationof state water
authorities. Only in the case of one large project in Coloradois relativelyunrestrictedtrading
of water rightspermitted(Box3). Obtainingauthorizationto changewateruse is oftena
lengthyand costly, requiringconsentfrom the relevant governingbody afterpublichearings in
which peoplewho couldbe damagedby the change in use can object.

33. Perhaps the most extremeexample of restricting transfer betweenuses occursin

California. The agriculturalsector makes up only 4 percent of the GDP of the state yet
receives about 44 percent of the water. Environmnentaluse also is allotted44 percent while
the urban and industrialsectorreceives only 11 percent. In the agriculturalsector water
rights vary widely from inheritedsources of cheap water to water that is highlysubsidized.
The anomaliesthat these restrictionscause are extreme; water is so inexpensiveto some
users (as low as $2.50 per acre-foot)that rice is cultivated in the desert whileat the same
time some municipalitieshave built desalinization plants to supplementtheir suppliesof
water at a cost of $2,000per acre-foot.Furthermore, incentivesto conservewateruse are
perverse. In agriculturemanyfarmersare forced to operate under a "use it or lose it' rule
while in urban use the rationingthat occurs during periods of drought is based on family use
during periods of plentifulwaterwhichencourages high water use whenthere is no rationing.

34. Clearly the systemrequiresreform, yet the political complicationsthat any reform
brings are significant.Assigningto farmers the ability to simply sell their rightswould give
them millionsof dollars in windfallgains on top of the large subsidiesthat they have already
received-an unpopularresultpolitically. Many the farmers fear that once rightsbecome
transferable they would not be compensatedfor what they would be givingup. Such
problems illustratethe deficienciesof administrative solutions to the allocationof scarce
water resources. However,the legislativeand administrativeconsiderationsin reforming
laws and proceduresallowinterestedparties great latitude to influencehow the gains from

Box 3. Water Trading in Colorado's Big Thompson Project

A notable contrastto the various restricted water right regimeswhichexistin the western
United Statesis providedby the Big Thompsonscheme through which 310,000acrefeet of
water have been suppliedannuallyto users in the Northern ColoradoWater Conservancy
District. The scheme,whichbringswater from the headwatersof the ColoradoRiverthrougha
tunnel underneaththe RockyMountainsto northeastern Colorado,was partiallypaidfor by
subscribersin the waterdistrictin returnfor the right to use the water. Soonafterthescheme
became fully operational,it becameapparent that water demandvaried significantlybetween
users and areaswithinthe district.The Northern ColoradoWater Conservancytherefore
establisheda system that allowedwaterrights to be traded on a permanentbasiswith theonly
requirementsbeing "beneficialuse", no sales outside the Districtand that usersabideby the
rules of the Conservancy.A centralregistryrecords ownershipand ownershiptransfers.The
system has becomeso refinedthat a simplepostcard is used to notifythe Conservancyof a
transfer.An importantreasonfor the smooth functioningof this market is that from the inception
of this project waterusers retainrights to any return flows. Thereforewhiledownstreamusers
get the benefit of returnflowsfrom waterusers upstream,they haveno rights to them and
upstreamnusers are free to transfertheir rights without the need to compensatedownstreamusers
for their loss of water.
An extremelysophisticatedmarket has evolved for this water.Many differenttypesof
contractsare used, from straighttransfersto the purchaseand sale of optionsto water.Withinthe
ConservancyDistrictall of the complexinfrastructureis in privatehands. TheConservancy's
role is to record transactionsand to check that there is no cheatingby those takingoff water.The
systemappearsto be operatingefficientlyand although there is undoubtedlyan economiccost to
owners of waterrights not beingable to sell their water outsidethe District,withinit water
appearsto be usedat its highestvalue.

reform are distributed. In addition, the very large number of people affected by any reform
make it costly and difficult to reach a solution that is agreeable to most parties. The
California case shows that even when institutions are relatively well developed,
administrative solutions to apportioning scarce water among different groups can lead to
anomalies that defy logic and waste resources.

Advantages of Water Markets

35. Tradable water rights allow the price of water to reflect the value of its alternative
uses, which creates incentives to put it to the most productive use. For example, if farmers
were able to sell their water rights at freely negotiated prices, some might sell surplus water
to a neighboring city where it has a higher value. Often they can generate a surplus by using
more efficient irrigation techniques or by switching to less water-intensive crops. In
addition, buyers of water rights are likely to conserve water more efficiently. Most new fruit
farmers in Chile use water-saving irrigation technologies and when Chile's main water
company, EMOS, realized that it could no longer obtain water rights without charge, it
invested in a program to significantly reduce physical water losses.

36. Chile's transfer of water to more productive uses was carried out voluntarily and
without having to raise water charges (Box 4). In fact, water charges fell following the
introduction of tradable water rights. The fall occurred because this regime facilitated the
transfer to user groups of the responsibility for carrying out operations and maintenance
(O&M) activities and for setting water tariffs and because users were able to carry out O&M
activities at a much lower cost than the Government. Despite the lower water charges, the
opportunity to sell water ensures that scarce water will not be used wastefully.

37. Tradable water rights can help shift water to higher value uses in a way that is cheaper
than other alternatives that may include building expensive new hydraulic infrastructure,
confiscating water from farmers, or raising water charges substantiallyto force farmers to
conserve water. Although the conveyance infrastructure to transfer traded water must be built
if it does not exist already, the cost of building may be less than that of generating new water
rights. For many years, the city of La Serena in Chile was able to meet its rapidly growing
demand for water by purchasing water rights from farmers at a lower cost than contributing to
the construction of a dam. Farmers received an acceptable price for their water and were
induced to use more efficient irrigation techniques. A recent study evaluating the Chilean
water market experience finds that the net gains from the trading of water rights in the Elqui
Valley were about $1,000/share, roughly equal to the price of these water rights (Hearne and
Easter 1995). In the Limari Valley, the gains from trading shares in the Cogoti Reservoir were
estimated to be three times the recent transaction prices of $3,000/share-thus even after the
cost of water transfer, water was worth three times more in one use than its next best
alternative. Without markets, it would have been difficult to effect this transfer.

38. Secure water rights are particularly beneficial for smaller farmers, who have been most
vulnerable to reductions in their water allocation over time and who have few other sources of
collateral. Tradable water rights, by empowering existing users, help reduce the abuses of
administrative allocation and give assurance to poor farmers that their water availability will
not be reduced. And because of their divisibility, water lights give farmers the possibility
of mortgaging only part of their rights for small loans, rather than their entire holdings.

Box 4. What is the Price of Water?

Manyconfusethewaterchargewith the price of waterrights.Undera tradablewaterrights

regime,the waterchargeshouldequalthe O&M cost of the infrastructure,whereasthe priceof
waterrightswouldbe themarketprice for the pernanent rightto use the water.To use an analogy
from the condominiummarket,the water charge is equivalentto the condominiumfeewhereasthe
price of waterrightsis analogousto the sale price of the condominium.Underan administrative
waterrights regime,economicefficiencyrequiresthat the waterchargeshouldequalthe
opportunitycostof thewater,whichin our analogy, wouldcorrespondto the marketrentalpriceof
the condominium-usuallyseveralmultiplesof the condominiumfee.Whereasthis priceis
difficultto set and enforceadministratively,the sale or leasepriceundera tradablewaterrights
systemautomaticallyreflectstheopportunitycost of water.

Water conservationmay also help control soil salinization,which is causedprimarily by


39. It is interestingthat Chile's sustained annual growth of 6 percent in agriculture

during the 1980soccurredalthough there were no public investmentsin new hydraulic
infrastructurefrom 1975to 1990. While this was due in part to heavyinvestmentin water
infrastructurein previousdecades, the tradable water rights regimefacilitatedthe growth for
new uses of water and contributedto the rapid expansion of fruit production.Secure and
tradable water rightsassure investors that their rights will not subordinatedto those of other
users during times of shortageand that, in fact, they will be able to buy waterfrom those
with a less valuableuse for it.

40. Chile has also been successful in increasing access of the poor to potablewater.
Ninety-ninepercentof Chile's urban residents and 94 percent of its rural residentsenjoy
access to potablewater, typicallyfor 24 hours a day. This contrastssharplywith comparable
rates of coverageof 63 percentand 27 percent in 1970in Chile and with developingcountries
elsewherein the world (Rosegrantand Gazmuri 1994a). Whilethis was dueto severalfactors
such as ensuringthat regulatedwater tariffs reflect the true cost of water, allowingcompetition
among water companies(Santiagoalone has seven privatecompanies),andsubsidizingwater
consumptionfor those withlow incomes, the ability of water companiesto buy water from
farmers playeda significantrole.

41. Securewater rightsgive potential investors in new hydraulicprojectsthe confidence

that, once they obtainthe rightsto the water generated by their investment(e.g. storage
reservoirsand conveyanceinfrastructure),they will be theirsto keep or sell to others (farmers,
industry, hydropowerandwatercompanies). Therefore ongoingstateownedprojects, such as
in Peru, couldbe privatizedby sellingthe hydraulic infrastructureand unallocatedwater and
land rights associatedwiththe project, with the conditionthat buyersrespectexistingland and
water rights. A comprehensive regulatory framework, as is preparedfor the sale for public
utilities, would assistin sucha privatization.

When and How to Establish Tradable Water Rights

Dificulties in Establishing WaterMarkets

42. Despite the promisethat water markets hold, few countrieshave establishedthem
formally. The economicargumentagainst tradable water rights rests on the perceptionof
market failure whicharise because:

* There are hightransactionscosts from setting up a newlegal, regulatoryand

institutionalframework,from defining, measuring,and enforcingwaterrights, from

identifyingpotentialbeneficialtrades, and from making necessarychangesin water

intakesand conveyanceinfrastructureto effect the transfers.
• Capital requirementsmay be high and time horizonslong so that naturalmonopolies
are createdwhichrequireregulation.
* There are issues of aquiferdepletion and returnflows.
* There are public goodsaspects of flood control, pollutioncontroland disease
controlalong watercourseswhich mayjustify governmentintervention.
* There are nationalsecurityand humanitarianaspects of many waterresourceswhich
mayjustify controlby government.
* Using watermarketsmay exclude the poor from access to water.

Comparing Water Market and Administrative Solutions

43. For the reasonsmentionedabove, an effectivemarket for water will require

regulationand be more difficultto establishthan say a market for land. However,the same
characteristicsof water make it difficultto allocatewater under alternativeregimes.Even
under administrativesystemsof water allocation,the rights have to be definedin a waythat
can be measuredand the resultingallocationof water rightsstill needs to be enforced.The
conveyanceinfrastructurerequiredto effect transfersin line with prioritieshas to be built
regardlessof whetherthe prioritiesare determinedby the market or by legal and administra-
tive means (see Rosegrantand Binswanger1994). Similarly,the sameenvironmentallawsand
institutionsneededto enforceenvironmentalqualityunder an administeredregimecan operate
under a tradablewater rightsregime. The conflictsbetweenconsumptiveandnon-consumptive
rights and concernsof monopolisticpricing exist under either system. Similarly,publicgoods
aspectsand issuesof accessof the poor to water exist in both regimes. For thebulk of the
issuesthen, the questionbecomeswhich of the two approaches-tradable propertyrightsor an
administeredregime-is likelyto yield better outcomes.

44. There are reasons to believe that a water market will function better than
administeredallocationin water-scarcecountries. Because a market systemincreasesthe
value of water, there are greater incentivesfor defining water rights clearly, for improving
their measurementand enforcement,and for establishingmechanismsto resolve disputes.
Similarly, the transactionscost of identifyingpotential gains from transferringwaterwill be
lower if borne by beneficiariesthan by public authorities. The conveyanceinfrastructurethat
must be built to effect the transfer will probably be built more cost-effectivelyby the private
sector. Water user associationsand river basin councils, which must play an importantrole
under either system, have a greater incentive to become stronger and better organizedwhen
water rights are well-definedand transferable.

45. Equity concerns are often raised within the context of tradable water rights.
However, the enormousinequitiesfrom administrativeallocation of water sufferedby the
poor have been well documented.The poorer sectionsof cities frequentlyresort to
expensivewater from tanker trucks while the rich sections have piped water providedbelow
cost. Similarly,poor farmersare more vulnerableto reductions in their water rightswithout

compensationunder administrativeregimes. Allowing rights to be traded increasesthe value

of the right and its transfer to more productive purposes increasesemploymentpossibilities.
As a result, the humanitarianand equity aspects of water allocationare likelyto be better
under a market regime. In this regard, Chilean policies which subsidizewater chargesand
sewage connectionsfor the poorest sections of the communityappearto have dealt with
potential inequitiesmore effectively than those countrieswhere water is allocated

46. An argumentagainsttradablewater rights is that institutionalmechanismsfor policing

water rightsmarketsand ensuring that monopoly rents are not being earned are weak.
However, this argumentapplies a fortiori to administrativesolutions.If institutionsare not
capable of ensuringa reasonablyfunctioning market in water rights, it is hard to imagine
how they could implementfair and equitable water administration,particularlygiven the
politicalpressures to which such administrators are subject.

47. Anotherargumentagainst the establishmentof tradable water rights rests on the

externalitiesthat exist in the use of water such as those related to return flows or the
environmentand those resulting from flood control. Tradable water rights could indeed
exacerbatethese. Therefore, if these are substantial, the efficiencygains from trading rights
might be minimal or trades may have to be disallowed. At issue here is the degreeto which
externalitiesexist in water markets versus the degree to which improperpricing and
allocationdecisionsunder administrativeregimes result in the wrong incentivesand in
misallocation.In many cases, tradable water rights internalizeexternalitiesthat arise from
water being wronglypriced. The California experience illustratesjust how badly
administrativedecisionscan distort incentives.

48. Althougha tradablewater rights regimneis a promisingalternativeto administered

allocation regimes,there are upfront costs to establishingthe new legal, regulatory,and
institutionalframework.The net benefits from water trading and from havingmore secure
water rights must be larger than the transactions cost, which includesthe initial legal,
regulatoryand institutionalcosts of establishing the regime, the costs of identifyingpotential
gains from trade and any negativeexternalities, and the cost of implementingthe transfers.
The politicaland culturally viabilityof individualproperty rights to water and the
institutionalcapacityto establishthe legal and regulatory frameworkto monitor and enforce
water rights are important issues that need to be addressed.

Issues of Transition

49. While the designand implementationof tradablewater rights needs to be tailored to

specific countrycircumstances,the following general guidelinesmay be usefulin the

50. First, it is importantto build support for the passage of legislationestablishing

tradable property rights in water. It may be useful to prepare, with appropriate

modifications,a draft water law based on the experience of other countries.A vigorous
informationcampaignand debatecan then help ensure that the final designand
implementationof the legal frameworkis done in a transparent and participatorymanner.
Explainingdraft versionsof the law with a willingness to accommodatereactionis critical to
success. Farmers and otherwater users have to be made aware that their concernsand
objectionshave been consideredand dealt with. The process can also help identifyand
mobilize groups that stand to benefitthe most from the proposed legislation.

51. Second, there is a need to establish effective institutionsto draft the regulationsand
to implementthe law efficientlyand fairly. This requires ensuringthat the wateruser
associationsand public institutions,such as water registries, water councilsand watershed
authorities, are able to carry out their responsibilities and that sufficientbudgetaryresources
are devoted for their effectivefunctioning.It may be useful to contractfor technical
assistance to draft the regulationsand to strengthen water user associationsat this stage. In
addition, it is importantto ensurethat staff of the public institutionsare capable,that they
fully understandand supportthe new legislation, and that they are perceivedto be honest
and unbiased. Given the role of public institutions in the initial allocationof water rights and
in the subsequentoperationof the water market, poorly trained or corrupt employeescould
prevent the market for water rights from ever developing or functioningeffectively.

52. There are severalissuesrelated to the initial allocation of tradablewaterrights. For

existing users, it is suggestedthat water rights be granted without chargein recognitionof
the fact that some farmershave already paid for their rights implicitlyin the purchaseprice
of their land and that the governmentis unlikely to recover the capitalcostsof its investment
in infrastructure. For new and unallocatedwater rights, it is importantthat they be sold via
auctions carried out in an open and transparent manner and that a minimumreservationprice
be establishedprior to the auction.Information on prices and volumesshouldbe made
publiclyavailable, and minimalcosts charged to enter the auction. In particular,care needs
to be taken that the poor are well-informedregarding the need to registertheir rights and the
procedures for doing so. The advantagesthat the poor can enjoy from secureproperty rights
can only come if they receive the rights to begin with. In addition to wateruser associations,
the public medianeeds to be extensivelyused to ensure water rights registration.There is
also a need to clarify that wherethere are large quantities of non-consumptiverights
(hydropower, for example),they do not prejudice consumptive rights. This may require
specifyingthe volumesthat will be released each month of the year (basedon historicuse of
consumptiverights holders) and ensuringthat any consumptive rightsbetweenthe intake and
discharge points are respected.

53. Where functioningwater user associations exist, the actual allocationshouldbe a

two-step process: water rights shouldbe first assigned to the water user associationsbased
on past usage and then assignedto the individual users by the associationsaccordingto
guidelines issued by a WaterCouncil. The titles to water are registeredonly at the
individual level and not at the user associationlevel. The two-stepmethodhas two
advantagesover direct assignmentto individuals. First, it is easier for the wateruser

associationrather than the Governrmentto verify past water usage of individualfarmers.

Second, it leads to titling manyusers simultaneously.This "block titling" of waterrights
reduces unit titling costs and helps resolve conflicts. It is also importantto ensurethat
elections for the officialsof the water user associationsare conductedin a transparentand
fair manner so that if membersof the association are dissatisfied with the way it is being
run, then can removethe officialsthat are not performing satisfactorily.Whilethis will not
eliminateunjust allocationsor corruption, it will help reduce it and is still likelyto morejust
and less corrupt than when unelectedgovernment officials are makingdecisionson water
allocation and pricing.

54. For the secondstep of the initial allocation process, the guidelinesmay vary by
region, watershed, and canal. Where there already exist registered water rightsand where
there is sufficientwater to honor all water rights, it is probably sufficientto have them re-
registered in the new public registry of water rights. However, where the existingregistry
contains many overlappingproperty rights (the sum of water rights exceedsthe water
available), it would be better for the initial allocation to be based on past usageestimatedby
water user associations.In situationswhere there have existed gross abusesof water rights,
it is probably best to assignthem to communitiesbased on historic use and subsequently
proportionallyto individualsbased on irrigated land area.

Design Issues

55. The potential of water trades to infringe upon the rights or water availabilityof third-
parties needs to be well understoodand addressed. This is most likely to occur for
agricultural "return flows." If a farmer were free to sell his entire irrigationwaterto users
outside the area, users downstreamthat may have received the farmer's returnflows would
lose their water withoutcompensation(Figure 1). One way to addressthe return flow
problem is by having the wateruser associationand/or a public body suchas a watershed
authority approve requests for changesof water intake to ensure that third party rights are
not affected. Since virtuallyall sales of water outside the area will require a modificationin
water intake, this shouldin principleprotect against water sales that reducethe water
available to third parties. However,the way that this rule is enforcedcouldeither penalize
downstream farmers or stifle the market. Some alternate ways for formulatingthe
regulationsto addressthis issueare discussedbelow.

56. One option is to adopt the Chile approach where all permanentconsumptiveuse
rights are expressed as a percentageshare of water availability (eitherin a streamor
reservoir), with the shares summingto 100 percent. If, because of the return flow effect, a
sale of water rights resultsin reduced water availability, all consumptiverightsholders,
includingthe entity buying the water, would share in the reduced flow (Figure2). The
system works fairly well in Chile, where few irrigation systems have significantreturn
flows. In the case of two Chilean rivers with high return flows, the Elqui and Aconcagua,
their respectivewater user organizationshave prohibited upstreamusers from sellingtheir
water to users whose return flows would not flow back into the river. In countrieswhich

Figure 1. ReturnFlowProblemWhen UpstreamFarmersCan Sell 100%of VolumetricRights

Initial Situation
60 Fanner A With A and B returning 50%
150 o 30 s of their allotment to the water
units 90IT-- FarmerB 90 -. T-.A9
3 supply, A, B, and C can each
60 B 30 60 C receive 60 units of water, with
30 units reserved for

60 WaterCo. (buys share of Farmer A) After Sale

150 After Sale___

uni0ts 3 90 30 60 30
-A If A sells his share to a water
, FarmerB ....... "-Farmer C company, C will receive only
60 30 30 half his previous allotment.

Figure 2. How Chile Handles Intersectoral Trades

48 - WaterCo. (buys share of Farmer A) The water user association

50t1 54 A 78 30- adjusts shares until all users
unitsm15FaAe _ F have the same volume. Thus,
48 B -Farer C A's sale reduces water for B and
C by 20%.

Figure 3. Proposed Solution

33 According to published averages,

150 W WaterCo. (buys share of Farmer A) A's crop consumes 33 units of
units 117 59 88 - 30 > water(inactualityitis30units)
units>....FarmerB * -: FarmerC and so A can sell only this
58 29 58 volume to the water company.
The WUA will adjust B and C's
usage at 58 units apiece

have significantflows,the Chilean system could restrict tradesin too manyrivers or reduce the
amount of wateravailableto downstreamusers. Thus other optionsmay be preferable.

57. An alternativeformulationwould be to specify that all water rightshave both a

consumptiveand non-consumptiveportion. The consumptiveportion couldbe sold without
restriction. The non-consumptivepart could be sold if it did not depriveothers of water.
Thus, where there are no return flow issues (most transfers withinthe samewater basin for
the same use), owners wouldbe free to sell 100 percent of their water rights. Becauseof
difficulties in calculatingthe purely consumptive portion of the water on a case-by-case
basis, this approach,whichis the similar to that used in California,may not be appropriate
for developingcountries.However, it may be suitable to calculateand publishaverages for
pure consumptiveuse as a basis for the trade. These averagescould specifythe volume of
water consumedby a certain crop or activity, with owners being free to sell only this
amount. This would reducethe need for each seller to justify the consumptiveportion of the
water while giving sufficientprotection to downstream users (Figure 3). This systemwould
work equally well for surface and ground water.

58. It may be desirableto introduce a tax on the holdingsof property rights for water
whose rate is determinedsolelyon the holdings of water rights and not by the purpose for
which the water is used or for the quantity of water actually used. In this way, the tax has
some desirablepropertiessirmilarto those for land taxes: it does not distortproduction
decisions and it helps recoverpublic investment costs in infrastructure.For equity and
administrativeease, the regulationscould exempt farmers and other users that hold small
quantities of water rights.

59. To reduceconflictsbetween consumptive and non-consumptivewaterrights, the tax

should also be appliedto non-consumptiverights, although the tax rate could be lower.
Similarly the tax on contingentwater rights could be at a different(lower)rate, or be based
on the amountof wateractuallymade available. The proceeds of the tax couldbe used both
to finance watershedactivitiesof a public goods or externalitiesnature. The introductionof
the tax on water holdingsshouldcoincide with the removal of the existingtax supplementon
irrigated land so that irrigatedland is not double-taxed. There is also a good case for a
lump-sum"exit" watertariff that would be paid to the original water user associationso as
not to burden the remainingwater users if a member were to sell his or her rights outside
the association.

60. There are two areaswhere monopolies of water rights coulddevelop;in privatizing
large hydraulic projectsand in the sale of non-consumptivewater rights. To deal with the
first problem, it is crucialthat an appropriate regulatory frameworkbe developedprior to

2. This would be based on the discounted value of a stream of estimated future water tariffs. Some Mexican
water user associations obliae buyers to pay a percentage of the water tariff to the original association.

privatization.This shouldbe done in the context of each scheme, in a similar mannerto

those developed for the sale of other former public monopolies.In the lattercase, the tax on
non-consumptivewater rights and minimum reservationprices at auctions, accompanied
with regulationsdeterminingpower tariffs, should help avoid monopolies.

61. Environmentalsafeguardsmay also be needed. For most enviromnentalissuessuch

as those relating to water quality, there is no need to change standards simply becausewater
trades are now allowed. If water quality laws need tightening, it is best done independently
of the laws establishingtradablewater rights. However, it is important to ensurethat
minimum flow requirementsexist in areas where water sales could lead to desertification,
habitat could be damaged, or recreational activities threatened.

62. In areas where the extensiveuse of groundwaterpumping may lower the water table
(as in parts of Chile), it is importantthat ground water rights and use be recordedand
subject to regulation. Under most administeredsystems of water allocation, ownersof the
land above an aquifer have full rights to its water, even if their use were to result in
depletionof the aquiferand even if its extraction infringes upon surface water rights. Under
the Chilean law, there is better protectionagainst aquifer depletion by relyingmore on users
themselvesto monitor extraction. To register ground water rights, the Chilean law requires
owners to belong to a ground water users commissionthat helps monitor extraction.If the
exploitationof ground water by a user causes detriment to others who are legallyentitledto
the water, Chile's General Directorate of Water, at the request of one or more of the
affectedparties, may establishtemporary and proportional reductions in volumetricrights
and bar new exploitation.The law also establishes an area of protection in whichthe
installationof similar works (e.g. pumps) is banned.


63. This paper arguesthat as comparedto administrativemethodsof waterallocation,

secure and well-definedtradableproperty rights to water in water-scarceregionsare likeiyto
improve wateruse. The experiencewith administrativewater managementsystemshas not
been impressive.Water is used wastefully,public hydraulicprojects are poorly conceived,
implementedand operated,and the systemshave failed to protect the environmentor make
water accessibleto the poor. As urbanizationincreasesand pressures on watersuppliesand
governmentbudgetsgrow, solutionsbased on such approachesare likelyto becomemore
difficult. In principle, and reflectingthe Chilean experiencein water-scarceareas,tradable
rights can benefitthe poor and increaseuser participationin water allocationandinvestment
decisions.They can allowrapid and voluntarychangesin water allocationin responseto
changingwaterdemandsand stimulateinvestmentand employmentas investorsare assuredof
their access to water. In addition,the economicefficiencyof agriculturalproductionwillbe
enhancedas output will reflectthe true scarcity of water rather than the frequentlydistorted
prices set by administratorssubjectto politicallobbying.

64. However,water marketsare not a panacea. And becauseof water's unique

characteristics,an effectivetradable water rights systemis not easy to introduce.There are
high transactionscosts. In addition, an unregulatedwater market also couldlead to
environmentalproblemsand monopolies.It could also result in under-investmentin activities
that may be sociallybut not privatelyprofitable. However,these same characteristicsmake
administrativesolutionsto water allocationdifficult. As shownby the experienceof Chile,
water marketsrarely makethem worse. The challengeis to decidewhen watermarketsare a
better alternativeandto enact legislationto establishthem. The designand implementationof
the legislationshouldinvolveall those that have a stake in how water is used. It shouldalso
pay particularattentionto the initial allocationof water rights, to the creationand maintenance
of a water rightsregistry,and to ensuring that the rightsof third parties are respected.As with
any system,public authoritieswill also need to designand enforceenvironmentallaws and to
subsidizethose high-returnactivitieswhere the benefitsaccrueto personsthat are not sharing
in their costs.

65. One strikingaspect of the debate on water marketsversus administrativemethodsof

water allocationis the lack of empiricalevidenceregardingmany of the key issues. This is not
surprisingsincetransactionscosts and institutionalconsiderationsare notoriouslyhard to
quantify. Nevertheless,in a debate of such importancethe absenceof data is hampering
reasoneddiscussion.Someefforts are underway to remedythis problembut muchmore needs
to be done.3 Further investigationof water markets that are currentlyworkingis requiredas
well more work on the inadequaciesand shortcomingsof administeredsolutionsto problemsof
water scarcity.Giventhe widespreadperceptionof failureof governmentsin developing
countriesto solveproblemsof scarcity by administrativefiat, many peoplefeel that the burden
of proof on the superiorityof an admninisteredapproachis on its advocates.Additional
evidencewill allowthe alternativesto be compared and evaluatedmore objectively.

3. Whilesomestudiessuch as Hearneand Ester (1995) have attemptedto measurethe gains fromwatertrades,

such studieshavenot attemptedto comparethe institutionaland investmentaspectsof market-basedversus


Aichian, Armen, and Harold Demsetz. 1973. "The Property Rights Paradigm."Journalof
EconomicHistory 31(1): 16-27.
Barzel, Yoram. 1989. EconomicAnalysis of Property Rights. New York: Cambridge
Coase, Ronald H. 1960. "The Problem of Social Cost." The Journal of Law and Economics
(3). 1-44.
Demsetz, Harold. 1967. Ownership,Control, and the Firm. Oxford: BasilBlackwell.
Frederiksen, Harald, JeremyBerkoff,and William Barber. 1993. "WaterResources
Managementin Asia." World Bank Technical Paper 212. Asia Technical
Hearne, Robert R., and K. WilliamEaster. 1995. "Water Allocationand WaterMarkets:
An Analysis of Gains-from-Tradein Chile." World Bank, Agricultureand Natural
Resources Department,Washington,D.C.
Holden, Paul, and SarathRajapatirana.1995. Unshacklingthe Private Sector;A Latin
American Story. Directionsin Development, World Bank, Washington,D.C.
Pakistan Water and Power DevelopmentAuthority. 1990. "Trading of Canal andTubewell
Water for IrrigationPurposes." P&I Publications 358 (December).
Rios, Monica, and Jorge Quiroz. 1995. "The Market for Water Rights in Chile:Major
Issues." World Bank,TechnicalPaper Number 235, Washington,D.C.
Rosegrant, Mark, and Hans Binswanger.1994. "Markets in Tradable WaterRights:
Potential for EfficiencyGains in Developing Country Water ResourceAllocation."
World Development22 (November):1613-25.
Rosegrant, Mark W., and RenatoGazmuri. 1994a. "ReformingWater AllocationPolicy
through Markets in TradableWater Rights: Lessons from Chile, Mexicoand
California." EPTD DiscussionPaper 6 (October). Washington,D.C.: International
Food Policy ResearchInstitute(IFPRI).
.1994b. "EstablishingTradableWater Rights: Implementationof the MexicanWater
Law," Draft mimeoavailablefrom authors.
Sampath, Rajan K. "Issues in IrrigationPricing in DevelopingCountries." World
Development20 (July):967-77.

Shah, Tushaar. 1991. "ManagingConjunctiveWater Use in Canal Commands:Analysisfor

Mahi Right Bank Canal, Gujarat." In Ruth Meinzen-Dickand Mark Svendsen,eds.,
Future Directionsfor Indian Irrigation. WashingtonD.C.: IFPRI.
World Bank. 1993. WaterResourcesManagement. A World Bank Policy Paper.
Washington, D.C.: World Bank.
. 1995. "Peru: A User BasedApproach to Water Managementand Irrigation
Development." Report Number 13642-PE. Washington, D.C.: World Bank.
Policy Research Working Paper Series

Title Author Date for paper

WPS1611 Economic Analysis for Health Jeffrey S. Hammer May 1996 C Bernardo
Projects 37699

WPS1612 Stock Market and Investment: The Cherian Samuel May 1996 C Samuel
Signaling Role of the Market 30802

WPS1613 Does Public Capital Crown Out Luis Serven May 1996 E Khine
Private Capital? Evidence from India 37471

WPS1614 Growth, Globalization, and Gains Thomas W Hertel May 1996 A Kitson-\Nalters
from the Uruguay Round Christian F. Bach 323947
Betina Dimaranan
Will Martin

WPS1615 Issues in Measuring and Modeling Martin Ravallion June 1996 P Sader
Poverty 33902

WPS1616 Transient Poverty in Rural China Jyotsna Jalan June 1996 P Sader
Martin Ravallion 33902

WPS1617 Why is Unemployment Low in the Simon Commander June 1996 L Alsegaf
Former Soviet Union? Enterprise Andrei Tolstopiatenko 36442
Restructuring and the Structure
of Compensation

WPS1618 Analytical Aspects of the Debt Stijn Claessens June 1996 R. Velasquez
Problems of Heavily Indebted Enrica Detragiache 39290
Poor Countries Ravi Kanbur
Peter Wickham

WPS1619 Capital Flows to Latin America: Sara Calvo June 1996 M Gomez
Is There Evidence of Contagion Carmen Reinhart 38451

WPS1620 Bank Insolvencies: Cross-country Gerard Caprio, Jr. July 1996 B. Moore
Experience 38526

WPS1621 The Sustainability of African Debt Daniel Cohen July 1996 S. King-Watson

WPS1622 Capital Control Liberalization and Ross Levine July 196 P Sintim-Aboagye
Stock Market Development Sara Zervos 38526

WPS1623 Environmental Degradation and Deon Filmer July 1996 S. Fallon

the Demand for Children: Searching Lant Pritchett 38009
for the Vicious Circle

WPS1624 Structural Adjustment, Ownership Luca Barbone July 1996 C Pelegrin

Transformation, and Size in Polish Domenico Marchetti, Jr. 85067
Industry Stefano Paternostro

Title Author Date for paper

WPS1625 Restructuringand Taxationin SimonCommander July 1996 L. Alsegaf

TransitionEconomies AndreiTolstopiatenko 36442

WPS1626 Partnersor Predators?The Impact JeffreyD. Lewis July 1996 N. Mensah

of RegionalTrade Liberalizationon ShermanRobinson Q4-058

WPS1627 TradableWater Rights:A Property Paul Holden July 1996 P. Mendez

Rights Approachto ResolvingWater MateenThobani 38893
Shortagesand PromotingInvestment