You are on page 1of 27

CULTUREAND CONGRUENCE:THE FIT BETWEENMANAGEMENT

PRACTICESAND NATIONALCULTURE

KarenL. NewmanandStanleyD. Nollen*


Georgetown University

Abstract.Thefinancialperformance of EuropeanandAsianworkunitsof
one multinationalcompanyis examinedas a functionof the congruence
betweenmanagementpracticesand nationalculture.Using Hofstede's
five nationalculturedimensionsand analogousmanagementpractices,
we findthatworkunitfinancialperformanceis higherwhenmanagement
practicesin the workunitare congruentwiththe nationalculture.

Until recently,the dominance of American managementtheory led to the


belief that "one size fits all," that a good managerin the U.S. will also be a
good managerin othercountries,and that effectiveU.S. managementpractices
will be effective anywhere.This view is now being supplanted with the
knowledge that managerialattitudes, values, behaviors,and efficacy differ
across national cultures. There is no one best way to manage a business.
Differencesin nationalculturescall for differencesin managementpractices.
This simple idea is surprisinglydifficultto accept. The currentdrive toward
globalization makes it harder. Globalization leads to standardization.Big
Macs are the same aroundthe world, but McDonald'smanagementpractices
should not be. Just as Big Macs in Moscow are status and luxurywhile Big
Macs in New Yorkare utilitarian,requisitemanagementpracticesdifferacross
cultureseven whenproductsdo not.
The fact that managementpracticesshould not be universalis illustratedby
examples with which most managers are familiar. Pay-for-performance
schemesarepopularand workquitewell in the U.S. and U.K. but are less used

*KarenNewmanis Professorof Managementat GeorgetownUniversity'sSchool of Business.


Her researchinterests include organizationalchange in Central and East Europe, ethical
work climates,high performancework groups, and managerialcareers.Her work has been
published in such journals as the Academy of Management Journal, Industrial Relations,
Public Administration Review, and Human Relations. Professor Newman earned her Ph.D.
from the Universityof Chicago,GraduateSchool of Business.StanleyNollen is Professorof
Managementat the GeorgetownUniversitySchool of Business.His researchincludeswork
on the problemsand prospectsfor businessin emergingmarketeconomiesof CentralEurope
and Asia; high performancework groups;and studiesof the processof forminginternational
trade policy in the U.S. Congress.His articles have appearedin IndustrialRelations,Inter-
national Organization, Human Relations, the Journal of Labor Research, and JIBS.
This researchwas supportedby the participatingcompanyand by the GeorgetownUniversitySchool of
Business,whosecooperationand assistancewe gratefullyacknowledge.CarlaInclan,KasraFerdows,Rob
Grant, and JohnyJohanssonprovidedvaluablefeedbackon early drafts.Dawn Brand,Byron Crossen,
HollyKramer,andSoominOmprovidedresearchassistance.Twoanonymousreviewersprovidedinvaluable
guidance.All viewsexpressedas well as errorsof fact or interpretation
are the authors'responsibilities.
Received: November 1995;Revised: May & September 1996; Accepted: September 1996.

753

Palgrave Macmillan Journals


is collaborating with JSTOR to digitize, preserve, and extend access to
Journal of International Business Studies ®
www.jstor.org
754 JOURNAL OF INTERNATIONALBUSINESS STUDIES, FOURTH QUARTER 1996

and not so successfuloutsidethe Anglo countries.This is not just an issue for


Americancompanies.It worksboth ways.Qualitycirclesare widelyused and
effectivein Japanbut have not deliveredthe same performanceresultsin the
U.S. despiteno lack of trying.
The thesis of this research is that multinational enterprisesneed to adapt their
managementpractices to the national cultures in which they operate in order to
achieve high business performance. Work units that are managed consistent
with nationalculturalexpectationswill be betterperformingthan work units
whose managementpracticesdo not fit the nationalculture.
There has been no definitiveempiricaltest of this thesis to date. The main
limitationof the literatureis that it contains only a few studiesthat actually
measuredperformanceoutcomesobjectively.Those studiesthat did wereable
to consideronly one or two differentmanagementpracticesand one or two
cultural dimensionsin two or three countries.In this study we attempt to
providea more completeempiricaltest.
We analyzefinancialperformanceoutcomesfor work units in eighteencoun-
tries on threecontinents.We describethe fit betweena varietyof management
practicesin theseworkunits and severaldimensionsof the nationalculturesin
whichthey operate.Wefind that in most cases,workunitswhosemanagement
practicesfit betterwith the nationalculturehavehigherfinancialperformance
than workunits wherethe fit is not as good. The messageto managersis clear:
Adapt your managementpractices away from the home country standard
towardthe host countryculture- Whenin Rome,do as the Romansdo.

NATIONAL CULTURE AND MANAGEMENT PRACTICES


National cultureis definedas the values, beliefs and assumptionslearnedin
earlychildhoodthat distinguishone group of people from another[Beckand
Moore 1985, Hofstede 1991]. This definition is consistent with Hofstede's
notion of national culture as softwareof the mind [1991]and with Jaeger's
"commontheories of behavioror mental programsthat are shared"[1986:
179]. National culture is embeddeddeeply in everydaylife and is relatively
imperviousto change.
Child's[1981]observationthat nationalculturewas woefullyunderdeveloped
conceptuallyfor comparativeresearchhas been addressedin recentyearswith
severalattemptsto conceptualizeand measuredifferencesin culturesamong
nations and to relateculturaldifferencesto differencesin managementprac-
tices. The most well-knownexamplesincludethe internationalsurveyresults
reportedin Trompenaars[1993],Hofstede [1991, 1980],Laurent[1983, 1986],
and Haire,Ghiselliand Porter[1963].Thereis ampleempiricalevidencethat
nationalculturesvary and that a varietyof managementpractices,including
strategic decisionmaking[Schneiderand DeMeyer 1991], leadership style
[Dorfmanand Howell 1988;Puffer 1993],and human resourcemanagement
CULTURAL FIT WITH MANAGEMENT PRACTICES 755

[Luthans,Welshand Rosenkrantz1993]differby nationalculture.The ques-


tion is whetherthese differencesamong managementpracticesand national
culturesmatterto workplaceperformance.We arguethat they do.
National cultureis a centralorganizingprincipleof employees'understanding
of work, their approachto it, and the way in which they expectto be treated.
National culture implies that one way of acting or one set of outcomes is
preferableto another.Whenmanagementpracticesare inconsistentwith these
deeply held values, employees are likely to feel dissatisfied,distracted,un-
comfortable,and uncommitted.As a result,they may be less able or willingto
performwell. Managementpracticesthat reinforcenationalculturalvaluesare
more likely to yield predictablebehavior [Wrightand Mischel 1987], self-
efficacy[Earley1994]and high performance[Earley1994]becausecongruent
managementpracticesareconsistentwith existingbehavioralexpectationsand
routines that transcendthe workplace.Employees are not distracted from
work performanceby managementpracticesthat ask them to behavein ways
that are consistentwith extantnationalculturalvalues.
The competitive advantage derived from correctly adapted management
practicescomes from alignmentbetween key characteristicsof the external
environment(national culture in this case) and internal strategy,structure,
systems, and practices [Burns and Stalker 1961; Chatman and Jehn 1994;
Powell 1992; Prescott 1986]. No single managementpractice is superior to
another with respect to performanceoutcomes. Instead, the congruence
betweenmanagementpracticesand the characteristicsof the nationalculture
produce better performanceoutcomes. (See also Wilkins and Ouchi [1983],
Denison [1990],Kotter and Heskett [1992],Denison and Mishra [1995] for
relateddiscussionsof the economicvalue of corporateculture.)For managers
of multinationalcorporations,the implication is that adaptation to local
culturalconditionsis necessaryto achievehigh performanceoutcomes.
In this research,we examine empiricallythe effect of congruencebetween
Hofstede's [1980, 1991] five national culture dimensions and analogous
managementpractices.Hofstede [1980],and Hofstede and Bond [1988]have
identifiedwork-relateddimensionsalong whichnationalculturesvary:power
distance, uncertaintyavoidance,individualism,masculinity,and long-term
orientation.Hofstede'sempiricalresultshave been replicatedat the national
level [Shackletonand Ali 1990;Chow,Shieldsand Chan 1991]and his cultural
frameworkhas been accepted as important and reasonablefor describing
differencesamong nations [Triandis1982].At the same time, little empirical
research has used the Hofstede dimensions to investigate the efficacy of
differentmanagementpractices on performancein differentculture groups
[Sondergaard1994].We addresseach dimensionin turn.
Power Distance
Powerdistanceis the extentto whichpeople believethat powerand statusare
756 JOURNAL OF INTERNATIONALBUSINESS STUDIES, FOURTH QUARTER 1996

distributedunequally and accept an unequal distributionof power as the


proper way for social systems to be organized. In organizations,power
distanceinfluencesthe amount of formal hierarchy,the degreeof centraliza-
tion, and the amount of participationin decisionmaking.Companiesin high
powerdistancecountries,such as those in East Asia and Latin Europe,tend
to be more centralized and have less employee participationin decision-
making.
Participationis an importantdimensionof work unit management.Denison
and Mishra [1995] found among firms in the U.S. (which has low power
distance) that "involvement"was one of four cultural dimensions that
delineatedorganizationsand that those organizationswith more employee
involvementwere more efficient and faster growing than others. Similar
results are found in other large-scalestudies of managementpracticesand
financial performance among U.S. firms [Hansen and Wernerfelt 1989;
Denison 1990].
The efficacy of participationin high power distance cultures is doubtful.
Employeesin high power distance cultures are likely to view participative
managementwith fear, distrust and disrespectbecause participationis not
consistentwith the nationalculture.Managerswho encourageparticipationin
these countries are likely to be seen as weak and incompetent.Morris and
Pavett [1992], for example, found participativemanagementto be related
positivelyto performancein the U.S. but not in Mexico. More authoritarian
managementpractices were effective in Mexico but not in the U.S. Jaeger
[1986] argues that managementinitiatives such as team building are not
effective in high power distance cultures because employeesfrom different
levels in the organizationare not comfortableinteractingface-to-face in a
group.
HI: In low power distancecountries,more participativework units
will be higherperformingthan less participativework units. In
high powerdistancecountries,less participativework units will
be higherperformingthan moreparticipativeworkunits.

UncertaintyAvoidance
Uncertaintyavoidanceis the extent to which people are threatenedby un-
certain, unknown, or unstructuredsituations.In organizations,uncertainty
avoidance is manifested by the clarity of plans, policies, procedures,and
systems.Relianceon clearprocedures,well-knownstrategies,and well-under-
stood ruleshelpsemployeesreduceuncertaintyand cope with theirdiscomfort
with unknownsituations.
This dimensionhas probablybeencriticizedmorethan any otherof Hofstede's
dimensions.Some have arguedthat it may be an artifactof the time during
whichit was developed(the 1960sand 1970s).Hofstedehimselfacknowledged
CULTURAL FIT WITH MANAGEMENT PRACTICES 757

that the linearnotions of right,wrongand truthimplicitin the dimensionare


characteristicof Westernthought and are not seen as strongly in Eastern
philosophies[Hofstede 1991].Nevertheless,the dimensionis consistentwith
the workof others.
Laurent[1986]reporteda nearlymonotonicrelationshipbetweenuncertainty
avoidanceand the proportionof managersthat agreewith the statement,"It is
important for a manager to have at hand precise answers to most of the
questionsthat his subordinatesmay raise about their work."He also found
that managersfrom the high uncertaintyavoidancecountriesof LatinEurope
are more likely to see the organizationas an authoritystructurethan others
[Laurent1983].
Denison and Mishra's [1995] "consistency"and "mission"diniensions of
corporatecultureare relatedto uncertaintyavoidance,the formerwith respect
to predictabilityand the latterwith respectto vision and direction.However,
they reported a weak but positive relationship between consistency and
performancein U.S. firmsand a strongpositiverelationshipbetweenmission
and performancein U.S. firms,contraryto the expectedcongruenceeffect(the
U.S. is a low uncertaintyavoidancecountry).
Schneiderand DeMeyer[1991]found LatinEuropeanmanagers,comparedto
those from lower uncertaintyavoidancecountries,tended to respond more
forcefullyand extremelyto environmentaluncertaintyand threat, and were
more likelyto respondto such threatswith majororganizationredesignsand
large-scaletraining programs.Uncertainty was perceived as a crisis more
among these managersthan others, their responseswere more extreme,and
their proposed solutions tended to involve a structuringthat would reduce
uncertainty.
Crozier's[1964]classicstudyof Frenchorganizationsis illustrativeof the effect
of uncertaintyavoidanceat the organizationallevel. Croziernoted that French
firmsare more hierarchicaland more rule-governedthan Americanfirms,yet
no moreor less effective.This is the congruenceargument.Whatworkswell in
France,a high uncertaintyavoidancecountry,does not work as well in the
U.S., a loweruncertaintyavoidancecountry.Rules,as integratingand control
mechanisms,are more efficaciousin Francethan in the U.S. [Slocumand Lei
1993].WhileFrenchemployeespreferthe certaintyof rules,Americansprefer
the discretionthat goes with ambiguity.
H2: In high uncertaintyavoidancecountries, work units in which
rules and direction are well-definedwill be higher performing
than work units in which rules and direction are not well-
defined. In low uncertaintyavoidancecountries,work units in
which rules and direction are not well-definedwill be higher
performingthan work units in which rules and direction are
well-defined.
758 JOURNAL OF INTERNATIONALBUSINESS STUDIES, FOURTH QUARTER 1996

Individualism-Collectivism
Individualism-collectivismis the extentto whichidentityderivesfromthe self
versus the collectivity. The individualism-collectivismdimension is an
important way of differentiatingamong national cultures [Kluckholnand
Strodtbeck 1961; Triandis 1989]. Individual cultures are loosely coupled.
Individualsare expected to look out for themselves and their immediate
families.Status derivesfrom individualaccomplishment.Collectivecultures
rely on membershipin groups- social classes, communities,religions, or
extendedfamilies- for identityand status.People are protectedby the group
and are expectedto act in the group'sbest interests.The Anglo countriesof
Britain, Australia, New Zealand, Ireland, Canada, and the U.S. are very
individualistic cultures. East Asian countries such as Taiwan, Korea,
Singapore,and Hong Kong are verycollectivecultures.
In organizations,individualismis manifested as autonomy, individual re-
sponsibilityfor results,and individual-levelrewards.Collectivemanagement
practicesemphasizeworkunit solidarityand team-basedrewards.
Thereis some empiricalsupportfor the importanceof this dimensionat work.
Individualism-collectivismis the dimension most likely to be replicatedin
studiesby people other than Hofstede [Sondergaard1994].Though Denison
and Mishra [1995] did not identify an analogous dimension, their sample
contained only U.S. companies, so they might not have found enough
variabilityon the dimension for it to discriminateamong U.S. companies.
Morris,Davis and Allen [1994]found that entrepreneurial attitudesincreased
as individualismincreasedin the U.S. (up to a point) whilejust the reversewas
found in Portugal,a more collectivesociety.Earley's[1994]study of training
and performanceis particularlyinstructive.Comparingthe U.S., Hong Kong
and China, he found that individuallybased trainingled to improvedself-
efficacyand higherperformancefor U.S. managerswhilegroup-basedtraining
led to improvedself-efficacyand higherperformancefor Chinesemanagers.
H3: In individualisticcountries, work units in which individual
responsibilityis emphasized will be higher performingthan
workunits in whichindividualresponsibilityis not emphasized.
In collectivistcountries,workunits in whichindividualresponsi-
bility is not emphasizedwill be higher performingthan work
units in whichindividualresponsibilityis emphasized.

Masculinity-Femininity
Masculine cultures are characterizedby doing and acquiring rather than
thinkingand observing,similarto the "orientationtowardactivity"dimension
identified by Kluckholn and Strodtbeck [1961]. Masculine cultures value
achievementand abhor failure while feminineculturesvalue affiliationand
view failureas much less important.MasculinecountriesincludeJapan,the
CULTURAL FIT WITH MANAGEMENT PRACTICES 759

U.S. and the Germaniccountries.Femininecountriesare typifiedby Nordic


countriessuch as Denmark,Norwayand Sweden.
In the workplace,this dimensionis reflectedin merit-basedopportunitiesfor
high earnings,recognition,advancement,and rewards,characteristicsalmost
takenfor grantedin the U.S. More femininemanagementpracticesemphasize
the quality of interpersonalrelations and quality of working life issues
commonlyassociatedwith Scandinaviancompanies.
The researchon performance-contingent rewardsin the U.S. is indicativeof
the efficacy of congruencebetween managementpracticesand a masculine
national culture. There is no doubt that performance-contingentrewards,
properly allocated, produce higher performancein U.S. work units. (See
Lawler[1990]for a thoroughdiscussionof performance-contingent rewards.)
Managementby objectivesis another managementtechnique that is better
suited to masculinecultures than feminine cultures [Jaeger1986; Hofstede
1991].
H4: In masculine countries, work units with more merit-based
rewardpracticeswill be higherperformingthan workunits with
less merit-basedrewardpractices.In feminine countries,work
units with less merit-based reward practices will be higher
performing than work units with more merit-based reward
practices.

Long-Term versusShort-Term Orientation


Hofstede'slast dimensionrefersto a country'stime orientation,long term or
short term [Hofstede and Bond 1988; Hofstede 1991]. This dimension was
added after Hofstede'soriginalIBM researchin an effort to capturedimen-
sions that might be particularlyrelevantin Asia. It is also noteworthythat the
studythat yieldedthis dimensionalso yieldedthreeof the otherfour.The only
one that was not found was uncertaintyavoidance.
Long-term-orientedcultures are characterizedby patience, perseverance,
respect for one's elders and ancestors, and a sense of obedience and duty
towardthe largergood [Hofstede1991].This dimensionis similarto the "time
orientation"culturedimensionidentifiedby Kluckholnand Strodtbeck[1961].
Long-term countries are found in Asia, including Hong Kong, Singapore,
Taiwan,and Japan.
Managementpracticesconsistentwith a long-termculturalorientationinclude
providing long-term employmentand solving problems for the long term
ratherthan making "quick fixes."Because this dimension is relativelynew,
there is no empiricalevidence of the efficacyof a long-term orientationon
performance.Nevertheless,we expect the congruencehypothesisto apply to
this dimensionas with the others.
760 JOURNAL OF INTERNATIONALBUSINESS STUDIES, FOURTH QUARTER 1996

H5: In long-term-orientedcountries,work units with a longer term


outlook will be higher performing than work units with a
shorter term outlook. In short-termoriented countries,work
units with a shortertermoutlook will be higherperformingthan
workunits with a longerterm outlook.

METHODS AND DATA


Data for this study come from 176 work units of one large U.S.-based
corporation.The work units are located in eighteen European and Asian
countries.Most workunitshavebetweenten and one hundredemployees,with
a mean size of fifty-five.The functionsof the workunits are sales,serviceand
support;there are no manufacturingunits. Product lines range from high-
technology computers and integrated systems to routine office products.
Customersrange from large banks and governmentagenciesto small retail
shops.Someemployeesaremarketingrepresentatives who call on customersto
sell products,some employeesaretechnicalspecialistswho provideserviceand
repairs to customers at their place of business, and others are systems
engineerswho create solutions to potential customers' problems.No em-
ployeesin staff functionssuch as financeor humanresourcesare in the study.
Nearlyall of the employeesarecitizensof the countryin whichthey work,not
expatriatesfrom the U.S. The only position routinelyheld by U.S. expatriates,
and only in largecountries,is Directorof Finance(not includedin this study).

Defining Work Units


The workgroupthat is the unit of analysismustbe aggregatedat a meaningful
level to measure perceptions of management practices and work unit
performance[Glick 1985].Employeesin a workunit shouldbe ableto interact
with each other,identifywith the unit, and hold some meaningfuloutcomesin
common. Our work units are geographicallyproximate(everyonein a work
unit works in the same place), functionallyhomogeneous,and have five or
more employees.In most cases they are departmentsor sections of a depart-
ment. Examples are field engineering,systems engineeringand marketing.
Membersof a workunit reportto a bonus-eligiblemanagerwho is the lowest
level manageraccountablefor financialresults,usuallythe departmenthead.
The methodologicaljustificationfor workunit level of aggregationis informed
by Blalock [1979],Jones and James[1979]and Glick [1985].Two criteriaare
used:differencesamongworkunits,and similaritywithinworkunits.On both
criteriaworkunit aggregationis supported.

Measures of Work Unit Performance


Work unit performanceis measured in three ways. Two of the ways are
financialperformancemeasures:returnon assets (ROA),and returnon sales
(ROS).ROA is net income beforetaxes dividedby assets.ROS is net income
beforetaxes dividedby revenue.These are both measuresof profitabilityand
CULTURAL FIT WITH MANAGEMENT PRACTICES 761

efficiency. The third performance measure is the work unit manager's


performancebonus as a percent of base salary.The bonus is based on the
degreeto whichthe managermeetsROAand revenuegoals (40%each) and on
the degree to which the manager meets individual performanceobjectives
other than ROA and revenue(20%).Bonus is more subjectivethan ROA and
ROS and takes into account circumstancesthat the purelyfinancialperform-
ance measuresdo not. These data are taken from the company'sfinancial
reportingsystem. All threemeasuresof performanceare normed againstthe
performanceof the workunit'sfunctionin the companyoverallby subtracting
the function'smean performancefrom each work unit's performance.All
performancemeasuresreferto 1989.
Measures of Management Practices
Weuse datafromthe firm'sregularemployeeattitudesurvey,aggregatedat the
work unit level, to measuremanagementpractices.The surveywas developed
for the firm's needs, not these research questions, so measurement of
managementpracticesis sometimesless than ideal. Each unit in the studyhad
a survey response rate exceeding 65%. The anonymous surveys were
administeredin 1988 and early 1989 and thereforeoccur in time before the
measureof performance,lending credibilityto causal arguments.Questions
wereaskedin a modifiedLikertformatwith a five-pointscalerangingfrom (1)
very positive,or stronglyagree,to (5) very negative,or stronglydisagree.We
reversedthe attitudescalesso that higherresponsevaluesalwaysindicatemore
positiveattitudes.Differentmanagementpracticesaremeasuredto correspond
to each of Hofstede'sfive cultural dimensions.The survey instrumentwas
culled for items that reflectedHofstede'sdimensions,based on his conceptual
argumentsand relatedwork of others.
Employeeparticipationin decisions is the managementpracticewe use that
correspondsto powerdistance.Hofstede'sconstructionof the powerdistance
dimension itself depended in large part on employees' perceptions of
managers' decisionmaking styles. Others have studied participation as a
managementpracticethat might differin efficacyacrosscultures[Morrisand
Pavett1991].We measureparticipationwith a two-itemscale:How wouldyou
rate the effort made by management to get the ideas and opinions of employees
like yourself, and, Employees where I work can get a fair hearing for their
complaints.The reliabilitycoefficientfor this scale is .91.
Clarity about policies and direction corresponds to uncertainty avoidance (e.g.
Laurent[1986];Slocum and Lei [1993]).Clarity about policies is measured
with a two-item scale: I understandpromotionpolicies and practices . . ., and, I
understand salary policies. . . . Reliability for this scale is .86. Clarity about
directionis measuredwith another two-item scale: How wouldyou rate this
companyon having a clear sense of direction?and How wouldyou rate your work
uniton havinga clearsense of direction?Reliabilityfor this scale is .79.
762 JOURNAL OF INTERNATIONAL BUSINESS STUDIES, FOURTH QUARTER 1996

Emphasis on individual contributions in the work unit is relevant to the


culturaldimension(e.g. Earley [1994]).Emphasis
individualism-collectivism
on individualcontributionis measuredwith a single item, We respect the
individuality of each employee andfoster an environmentin which employees'
creativity and productivity are encouraged, recognized, valued, and rewarded
(agree-disagree).This item is taken verbatim from the company'smission
statement.It is noteworthythat the mission of this U.S.-basedfirm included
such a statement,reflectingthe high level of individualismin the U.S.
Use of merit-basedrewardsin the work unit is relevant to the masculine-
feminine cultural dimension. In Hofstede'sconstructionof this dimension,
opportunitiesfor high earningsand advancementwere key components.We
measure merit-based rewards with a three-item scale: The better my
performance, the larger my salary increase will be, The better my performance,
the greater will be my opportunityfor promotion to a betterjob, and, How would
you rate . .. your opportunityfor advancement to higher leveljobs? Reliability
for this scale is .85.
Long-term problem-solving and employment security are relevant to the long-
term versus short-term culture dimension [Hofstede 1991]. Long-term
problem-solvingis measuredwith a single item: Whena problemoccurswhere
I work, we try tofix it in a way thatprevents it from occurringagain (rather than
just using a temporary 'quickfix'). Employment security is measured with a
single item: This company makes a genuine effort to keep people employedeven
underadverse business conditions.

Measure of National Culture


Using the culture dimensions identified by Hofstede [1980], countries are
dichotomized at theoretically meaningful points, based on the analysis
reportedby Hofstede[1980, 1991]and others[Ronenand Shenkar1985].For
example,long-termversusshort-termorientationis dichotomizedso that the
high (long-term)groupcontainsAsian countriesonly,all of whichhavescores
of 48 or greater.The low (short-term)groupcontainsEuropeancountriesonly,
all of which have scores of 44 or less. This yields a smallerlong-termgroup
than short-termgroupbecauseof the compositionof our sample.If we wereto
make the two groups more equal in size, we would have to include the
Netherlandsin the long-termgroup.This wouldnot be consistentwith extant
theoreticalwork on this dimension[Hofstede1991;Hofstedeand Bond 1988]
and would make the long-term group less homogeneous. Table 1 gives
descriptivestatisticsfor the workunits in the study,includingthe dichotomies
for each culturedimension.
Control Variables
Factorsaside from managementpracticesdeterminework unit performance,
especially the resources availableto the work unit and the nature of the
CULTURAL FIT WITH MANAGEMENT PRACTICES 763

E _0 otooNot colVco 0

r- 0O

co CO

3 O~~~C
urN0
4? c' Id d Cd Iq
D c' oq
0^t)
cs uo ) cq cd cdLo |ASlJ
-
ct

0f 0 g ON V2

a 5~~~~~~~~~~~~~~0

EJ Ic ISo

c0(D aD
C LOLO4T C -O
= N N 11) 04O= . o,. I~~~~~~~
CO 0) Al v 0 ru
i

l B0 C 1 88:~~~~~~~~~~~~~~~~

D r oO=n~~~

Z | S QcE
E c > tz B.- m X- a 10-8

8 < < ma E
O z U UE z z X X X ? , M ?9~~~~~~~~~~~~~~~~~~
764 JOURNAL OF INTERNATIONAL BUSINESS STUDIES, FOURTH QUARTER 1996

marketplace it faces. Three such variables are included in the analysis as


controls.
Labor resources are measured as the amount of experience in the work unit,
specifically the proportion of the work unit's employees with four-to-twenty
years of experience with the company. This variable comes from a categorical
response question on the attitude survey. We use four-to-twenty years to
capture a work force that is fully trained but not yet retiring. We expect that
work units with an experienced work force will have higher performance than
work units with either an inexperienced or an aging work force.
Capital resources are indicated by capital productivity, measured as revenue
divided by assets. These data are taken from the firm's financial reporting
system. Work units that have higher capital productivity, given labor resources,
should have superior financial performance.
Market conditions are measured by the rate of change of gross domestic
product (GDP) in the country in which the work unit is located for the
industry in which the work unit operates. For example, for work units whose
customers are banks, we use change in GDP originating in the financial
institutions industry. For work units with customers in many industries, we use
overall country GDP growth. Work units in more rapidly growing markets are
more likely to increase their revenue, and in turn to earn higher net income and
achieve higher profitability. Sources for these data are official statistics from
the OECD and the Asian Development Bank.
Unobserved Variables.In cross-sectional analysis, the value of the dependent
variable may be affected by unobserved, unmeasurable, or omitted in-
dependent variables. If omitted variables are correlated with included
independent variables, then the estimated effects of those variables on the
dependent variable will be biased [Jacobson 1990; Mauro 1990]. To account
for the effects of omitted variables on the relationship between management
practices and work unit performance, we include prior work unit performance
as a control variable. Prior performance is measured as return on assets in
1987, return on sales in 1987, and manager's bonus in 1987, two years prior to
current performance and one or more years prior to the measurement of work
unit climate.
Means, standard deviations, reliability coefficients (Cronbach's alpha), and
zero-order correlations among continuous variables are shown in Table 2.
We use analysis of covariance to test our hypotheses. For each national culture
dimension, countries were placed in either a high or low category as described
above. Management practices were dichotomized at the mean value of the
variable for all work units. An F-test indicates whether the interaction of
national culture and management practices affects performance significantly
overall. T-tests are used to test each hypothesis explicitly.
CULTURAL FIT WITH MANAGEMENT PRACTICES 765

LO E
2
t CDCD,iv~~~~~~~~~~~~~~~~~c
0n
ChCh'oCz~~~~~~~~~~~~4
QO(
CON c
CM Ch oo o CM ~~~~~~~~~~~~~~~o
~~~~~~~~~~~~~~~~~~~~~cooc

ooc\j co co c\j 3

@~~~~~~~~ 0

lla~~~~~~~~~~~~~~c
(O a'ot LO 't o.o )

N a I Q X O U O U U \5 .\ O L CO- I
.> O~~~~~~~~~~~~~
.: . . . . co0

C) Q o
t
0~~~~~~~~~~~ CO CM0 D 0 D E DC )s oOO
cnrc c s o

co CO C
CMco cocs

co
LU~ ~ ~ |6

co - t |X=>

O CDZ
ci *C\j 'It (D 0 0 (0 'It CO LO f,, LOi
766 JOURNAL OF INTERNATIONAL BUSINESS STUDIES, FOURTH QUARTER 1996

RESULTS
The results for each culture dimension are shown in Tables 3 through 8. Mean
performance figures, adjusted for the effects of the control variables, are shown
in the top half of each table. Results are also graphed in the bottom half of
each table with the exception of uncertainty avoidance (Table 8). If the
congruence hypothesis is supported, the graphs will resemble an inverted "U".
The congruent cells in the tables are the top right and bottom left cells.
Performance is hypothesized to be highest in these cells.
The hypotheses that work unit performance is higher when management
practices are congruent with national culture are supported for four of the five
culture dimensions. Work units in which managers fit their practices to the
values of the country's culture in which they operate had higher returns on
assets, higher returns on sales, and in some cases higher bonuses than work
units in which the fit was lacking. The exception to this finding occurs in the
case of uncertainty avoidance, where the hypothesis was only partially
supported (discussed below).
In particular, work units in low power distance cultures were higher per-
forming if they were more participative (Table 3). Conversely, work units in
high power distance cultures were higher performing if they were less
participative. The effect of fit on financial performance was stronger in high
power distance work units than in low power distance work units. This finding
is consistent with previous empirical results from studies in the U.S. (which is
a low power distance country) in which the effect of participation on
performance is equivocal [Miller and Monge 1986; Wagner and Gooding 1987,
Dyer and Reeves 1995].
Among individualistic national cultures, performance was higher when mana-
gers emphasized individual employee contributions (Table 4). In collectivist
cultures, however, performance was higher in work units with less individual
employee emphasis. The importance of fit between management practice and
national culture was greater in collective national cultures than in individual
national cultures.
The congruence hypothesis also was supported for the masculinity-femininity
dimension of national culture (Table 5). Work units in more masculine cultures
were higher performing if they made more use of merit-based rewards for pay
and promotion, and conversely, work units in more feminine cultures had
higher performance if they made less use of merit-based rewards. For this
culture dimension, the importance of fit with management practices is roughly
the same for both masculine and feminine national cultures.
Congruence between long-term versus short-term orientation in the national
culture and management practices also was important to good financial
performance (Tables 6 and 7). In particular, in short-term-oriented cultures,
CULTURAL FIT WITH MANAGEMENT PRACTICES 767

TABLE3
Effect of Congruence between Power Distance and Participation
Power Distance (PD)

Low High

ROA = -10.28 ROA = 6.24


ROS = - .72 ROS = 2.44
Low Bonus = - 2.03 Bonus = 1.48
(n = 37) (n = 62)
0

ROA = - 2.20 ROA = -14.17


High ROS ==- .49 ROS =- 3.41
Bonus - .69 Bonus = - .65
(n 76) (n = 15)

tl,2 ROA = 1.36* t3,4 ROA = 2.26** FROA = 6.94***


tl,2 ROS = .13 t3,4 ROS = 2.22*** FROS = 3.68*
tl,2 Bonus = 1 35* t3,4Bonus = 1.48* FBonus= 3.95**

Y1 Power Distance and Participation Y2


10- - -
--- 3

.. . . . . . . . . . . . . . . .... . . . . ...XA . . . . . . . . . .- 2
5 - 0t

10 g------/
: - - - -- -- -0
-5 - .-.-.-.-..- -- --- ----- ---- --1

--2
-
-10_ . - --- - -3

-15 - 1 -4
PD Lo Partic Lo PD Lo Partic Hi PD Hi Partic Lo PD Hi Partic Hi

ROA (Y1) ---ROS (Y2)

Congruence occurs in the top right and bottom left cells which are hypothesized to have higher
performance than the other cells. The number of observations in each cell (n) applies to the Bonus
analysis; n varies slightly for the other performance variables because of missing data. The F-statistic
tests the overall hypothesis that performance is affected by congruence between culture and
management practice; the t-tests refer to differences between performance means for work units
within the same culture group.
*p <.10; **p <.05; ***p <.01
768 JOURNAL OF INTERNATIONALBUSINESS STUDIES, FOURTH QUARTER 1996

TABLE4
Effect of Congruence between Individualism-Collectivism and
Individual Emphasisa
Individualism-Collectivism

Individual Collective

uW
ROA = -18.35 ROA = 8.44
E ROS =- 1.15 ROS = 3.87
) Low Bonus = - 2.32 Bonus = 1.64
(n 37) (n=47)

CsL _ ROA = - 7.74 ROA = - 6.72


E ROS =- .61 ROS =- 2.67
i High Bonus = - 1.73 Bonus = - 1.52
(nc =22) (n = 56)

tl,2 ROA = 1.46* t3,4 ROA = 2.74*** FROA = 7.82***


tl,2 ROS = .26 t3,4 ROS = 4.20*** FROS= 7.54***
tl,2 Bonus = .44 t3,4 Bonus = 3.22*** FBonus = 5.02

Individualism-Collectivism and Individual Emphasis


Y1 Y2
10- / -4

-.- -X, ........


O----.---.- -- -- 2

0-
- - - ---
-- - - ---- ----
O -5 -2\

ROA(Y1) - -- ROS(Y2)

aAustraliaand New Zealandare excluded because questions were not asked on theirsurveys.
Congruence occurs in the top right and bottom left cells which are hypothesized to have higher
performancethan the other cells. The numberof observations in each cell (n) applies to the Bonus
analysis;n varies slightlyfor the other performancevariablesbecause of missing data. The F-statistic
tests the overall hypothesis that performance is affected by congruence between culture and
management practice; the t-tests referto differences between performancemeans for work units
withinthe same culturegroup.
CULTURAL FIT WITH MANAGEMENT PRACTICES 769

TABLE5
Effect of Congruence between Masculine-Feminine and
Merit-Based Rewards
Masculine-Feminine

Masculine Feminine

ROA = - 8.64 ROA = 2.55


ROS = - .73 ROS = 1.46
e Low Bonus = - .97 Bonus = 1.02
3I.O. (n = 57) (n = 39)

m
ROA = 3.77 ROA =- 8.74
X) ROS = .86 ROS =- 1.63
s High Bonus = .95 Bonus = - 2.25
(n= 62) (n =32)

tl,2ROA =2.28** t3,4ROA = 1.42* FROA = 6.02**


tl,2 ROS = .98 t3,4 ROS = 1.33* FROS= 2.74*
tl,2 Bonus= 2.11 ** t3,4 Bonus = 2.66*** FBonus = 11.52***

and MeritRewards
Masculinity-Femininity
Y1 ...Y2
4- 1.5

2- -/\0.5

< -2- ' \% $0O


0 0
cc ~~~~~~~~~~~~~~~~~~CD

-6 \~
-8- / -1.5

-1 0- l-2
Masc-LoMerit Masc-HiMerit Fem-LoMerit Fem-HiMerit
-- ROA(Y1) --- ROS (Y2)

Congruence occurs in the top right and bottom left cells which are hypothesized to have higher
performancethan the other cells. The numberof observations in each cell (n) applies to the Bonus
analysis;n varies slightlyfor the other performancevariablesbecause of missing data. The F-statistic
tests the overall hypothesis that performance is affected by congruence between culture and
management practice; the t-tests refer to differences between performancemeans for work units
within the same culture group.
*p<.1O;**p<.05; ***P< .01
770 JOURNAL OF INTERNATIONALBUSINESS STUDIES, FOURTH QUARTER 1996

TABLE6
Effect of Congruence between Long-Term Orientation and Approach
to Problemsa
Long-TermOrientation
Long-term Short-term

CO
E ROAa = 3.05 ROA = 10.51
X Quick ROS = 3.70 ROS = 1.79
.0 fix Bonus = 1.04 Bonus = .54
0. (n = 18) (n = 19)
0)

C'
C3 . _ _ _ __ _ _ _ ._ __ _ __ . _

co
0

Perma- ROA = 14.39 ROA = - 9.50


Perma- ROS = 6.63 ROS =- 1.93
Q en Bonus = 1.94 Bonus = - 1.76
CL fix (n = 10) (n = 45)

tl,2 ROA = 1.20 t3,4 ROA = 2.79*** FROA = 7.09**


tl,2 ROS =1.14 t3,4ROS = 1-97** FROS =4.43**
t1,2 Bonus = .51 t3,4 Bonus = 1 90** FBonus= 2.30

Y1 Long-TermOrientation and Problem Solving Y2


15 - -8

10 t-\ 6

0 0

-5-- \ \ -0

Long-Quick Long-Permanent Short-Quick Short-Permanent


- ROA (Y1) --- ROS (Y2)

aHomogeneity of variance assumptions are met (p > .05) using Cochran's C except for here.
Congruence occurs in the top right and bottom left cells which are hypothesized to have higher
performance than the other cells. The number of observations in each cell (n) applies to the Bonus
analysis; n varies slightly for the other performance variables because of missing data. The F-statistic
tests the overall hypothesis that performance is affected by congruence between culture and
management practice; the t-tests refer to differences between performance means for work units
withinthe same culturegroup
* .10;** 05;*P<.01
CULTURAL FIT WITH MANAGEMENT PRACTICES 771

TABLE7
Effect of Congruence between Long-Term Orientation and
Employment Security
Long-Term Orientation

Long-term Short-term

ROAa = 6.39 ROA = 5.29


ROS = 5.02 ROS = 2.47
Low Bonus = 1.85 Bonus = -1.05
(n=20) (n=20)

'& ROA = 19.84 ROA =-1 0,01


E
u Hh ROS = 7.58 ROS = - 2,65
Hgh Bonus = 1.37 Bonus = - 1,02
(n =) (n =44)l

tl,2 ROA =.78 t3,4 ROA = 2.46** FROA = 4.62*


tl,2 ROS = .82 t3,4 ROS = 3-32* FROS = 5.1 2*
tl,2 Bonus=.25 t3,4 Bonus= .02 FBonus= .05

Long-TermOrientationand EmploymentSecurity
Y1 Y2

Is I--- ....... -- ------------- -- 6

10 ..... -- - - - - - - 4
,/ .......... L.............

<A5 -. ..8 R -10

......- -- -- I......... . ....... ............. . .... ..........0

within ae~ utregop


-10Y
-N Y2 -
*P<.10 ** < 205;** 0
-15- l1l-4
Long-LoSecure Long-HiSecure Short-LoSecure Short-HiSecure
EROA -7.5
(Y1) ROS (Y2)

aWomogeneityof varianceassumptions are met 1 >.05) using Cochran'sC except for here.
Congruence occurs in the top right and bottom left cells which are hypothesized to have higher
performancethan the other cells. The numberof observations in each cell (n) applies to the Bonus
analysis;n varies slightlyfor the other performancevariablesbecause of missing data. The F-statistic
tests the overall hypothesis that performance is affected by congruence between culture and
management practice; the t-tests refer to differences between performancemeans for work units
772 JOURNAL OF INTERNATIONAL BUSINESS STUDIES, FOURTH QUARTER 1996

TABLE8
Effect of Congruence between Uncertainty Avoidance and
Management Practices
Uncertainty Avoidance
High Low

ROAa = -7.22 ROA = 1.24


Unclear ROS
= -1.56 ROS = 3.53
Unclear Bonus = .02 Bonus = - .31
C(n=48) (n=44)

0.
0

RA = 2.44 ROA = .32


ROS = .04 ROS = .27
Clear Bonus = - .66 Bonus =- .04
(n=45) (n=52)

tl,2 ROA = .86 t3,4ROA = .28 FROA = .66


tl,2 ROS =1.01 t3,4ROS = 2.34*** FROS = 5.63**
tl,2 Bonus = .63 t3,4Bonus = .25 FBonus = .40
Uncertainty Avoidance
High Low

ROAa = -6.75 ROA = -2.80


Unclear ROS = -2.01 ROS = 2.44
Bonus = - .36 Bonus = - .16
? (n = 61) (n = 33)

a0 __ _ .__ _ _ _ _ _ _ _ _ _ _

Z6W

ROA = - 1.08 ROA = 2.18


o ROS = 1.54 ROS = .77
Clear Bonus = - .39 Bonus = - .27
(n =32) (n = 63)

tl,2 ROA = .89 t3,4ROA = 1.86** FROA = .01


tl,2 ROS = 1 95** t3,4 ROS = 1.02 FRoS = 4.44**
tl,2 Bonus = .02 t3,4Bonus = .10 FBonus= .00

aHomogeneity of variance assumptions are met (p > .05) using Cochran's C except for here.
Congruence occurs in the top right and bottom left cells which are hypothesized to have higher
performance than the other cells. The number of observations in each cell (n) applies to the Bonus
analysis;n varies slightlyfor the other performancevariablesbecause of missing data. The F-statistic
tests the overall hypothesis that performance is affected by congruence between culture and
management practice; the t-tests refer to differences between performancemeans for work units
withinthe same culturegroup.
CULTURAL FIT WITH MANAGEMENT PRACTICES 773

work units that stressed"quickfixes"to problemsoutperformedwork units


that adopted a more permanentsolution to problems.Furthermore,work
units with less employment security outperformedwork units with more
employmentsecurityin these short-termoriented cultures.Our results also
suggestedthat work units in long-termorientedcultureswerebetterperform-
ing if they had the opposite managementpractices- permanentsolutions to
problems,and a stress on employmentsecurity- but the performancediffer-
enceswerenot statisticallysignificant,perhapsbecauseof smallsamplesizes in
these cells.
The hypothesisthat congruencebetweenuncertaintyavoidancein the national
cultureand the clarity of managementpracticesis importantfor high work
unit performanceis only partiallysupported(Table 8). Work units in high
uncertaintyavoidancecultureswhose employeesreporta more clear sense of
directionor more clear policies have better financialperformancethan work
units in these cultures with a less clear sense of direction. However, the
expectedresultsare not found in low uncertaintyavoidancecultures.In such
culturesthere are no consistent performancedifferencesbetween congruent
and noncongruentworkunits.

DISCUSSION AND CONCLUSIONS


We find supportfor the thesis that businessperformanceis betterwhen
managementpracticesarecongruentwith nationalculture.Workunitsthat are
managedconsistentwith the valuesof the externalcultureare more profitable
than work units in which the fit is less well achieved.These resultscontribute
to the small body of empiricalevidence supportingthe importanceof con-
gruencebetweenmanagementpracticesand nationalculture.This studybuilds
on the work done by Hofstede,whose analysiswas also based on marketing,
sales and customerserviceemployeesof one largecompany(he also collected
data from manufacturingplants, but did not use them in his cross-country
comparisons[Hofstede 1980, p. 41]). Our study is more comprehensivethan
other recentempiricalworkbecausewe examineall five of Hofstede'scultural
dimensions(not just one or two), using data fromeighteencountries(not just
two or three),and we controlfor priorperformanceand exogenouseconomic
factorsthat could accountfor differencesin performance.
The findingshave an importantimplicationfor managers:managementprac-
tices should be adaptedto the local cultureto be most effective.Managers'
effortsto encourageemployeeparticipationmight improvethe profitabilityof
workunits in countrieswith a cultureof low powerdistance,such as the U.S.,
but more employeeparticipationis likelyto worsenprofitability,not improve
it, in countrieswith high power distance,such as Latin Europeanand East
Asian countries.An emphasison individualemployeecontributionsis likelyto
improve profitabilityin individualisticcountries, such as the U.S., but to
worsen profitabilityin collective countries, including those in East Asia.
774 JOURNAL OF INTERNATIONALBUSINESS STUDIES, FOURTH QUARTER 1996

Substantialuse of merit-basedpay and promotionshouldimproveprofitability


in masculinecountries,such as Anglo and Germaniccountries,but worsenit
in feminine countries,such as Nordic countries.A short-termapproachto
problem-solvingand a small sense of employmentsecurity apparentlyare
good for profitabilityin short-termorientedcountries,such as the U.S., while
long-term approacheswork better in long-term-orientedcountries, such as
those in East Asia.
Corporateinitiativesthat arecreatedat headquartersand promotedworldwide
run the risk of conflictingwith unreceptivenational cultures.In particular,
management practices that are favored in the U.S., such as employee
participation,individualresponsibility,merit-basedrewards,and short-term
approaches,are likelyto be unwisein some othercountriesthat are culturally
unlike the U.S. A mismatchbetween work unit managementpracticesand
national cultureis likely to reduceperformance.The issue of fit is especially
salient for U.S. managersbecausethe U.S. has quite extremevalues for both
individualismand long-termorientationculturaldimensions.By implication,
the more different the host country culture is from the company'shome
countryculture,the more the companywill need to adapt.If we take national
cultureto be relativelyimmutable,the onus is on managersto adapt.
The companyfromwhichthese data weretakenis not generallyconsideredto
havea strongcorporateculture.In fact, one of the reasonsfor the surveywas
to assessa recentattemptto developa strongercorporateculture.If the com-
pany in question had a strong corporate culture, the congruence effects
betweennationalcultureand managementpracticesmightbe weaker.Perhaps
the strongcorporateculturewould attenuate,but not eliminate,the influence
of nationalculture,consistentwith Adler's[1997]observationthat organiza-
tional culturedoes not erasenationalculture.Unless hiring,developmentand
promotion practicescan all transcendnational culture to reinforcea very
strong corporateculture [Evans 1992], companiescan raise performancein
foreign affiliates by adapting their management practices to the affiliate
country'snationalculture.
Ourresultssuggestthat companiesare well advisedto takenationalcultureas
a given and adjusttheirpracticesaccordingly.This also meansthat American
firms, for example, should not try to become more like Japanesefirms or
Germanfirms.Whilethereis muchto be learnedfromexemplarymanagement
practicesin other cultures,the differencesbetweencultureslimit the transfer-
ability of managementpracticesfrom one to the other. The more collective,
long-termorientationof Japanesenationalculturearguesin favorof manage-
ment practicesthat are very effectivein Japanbut that havenot achievedsuch
high performanceresultsin the U.S.
In our results the uncertaintyavoidancedimension is a special case. Our
findingswere not consistent:clarity of directionand policies aided perform-
CULTURAL FIT WITH MANAGEMENT PRACTICES 775

ance in high uncertaintyavoidancecountries,though some effects were not


statisticallysignificant.In low uncertaintyavoidancecountrieswe found no
discernablepatternof results.
These equivocalresultsmay be understoodin three ways. First, uncertainty
avoidance may be an artifact of the time during which Hofstede did his
originalresearch.He measuredit in part as the amount of stressexpressedby
employees.He interpretsthe dimensionas "thelevel of anxiety. . . in a parti-
cular society in the face of an uncertainfuture"[1991:112].This definitionis
differentfrom the other definitionsof his dimensionsbecause it contains an
implicit reference to time - an uncertain future. Measuring uncertainty
avoidance in the late 1960s and early 1970s may have resulted in higher
uncertaintyavoidancein Europebecause of the cold war and the still-vivid
memoriesof WorldWarII than might be the case today.Second,uncertainty
avoidancemay simply be irrelevantas a concept in countrieslabeled as low
uncertaintyavoidance.It could be that clarity of policies and directionis a
good managementpractice, regardlessof national culture, consistent with
Denison and Mishra's[1995]findings.Third,Hofstedeand Bond [1988]found
that uncertaintyavoidancedid not emergeas a dimensionthat discriminated
among countries,using a new instrumentbased on a Chinese values study.
Perhapsthe dimensionis not relevantfor Asians [Hofstede1991].Our results,
taken together with Hofstede and Bond's results, suggest that uncertainty
avoidancemay not be a usefuldimensionfor comparativeresearch.
Methodologically,this study benefitsfrom having independentdata sources
for managementpracticesand work unit performance.Performanceis not a
subjectiveperception,colored by individuallikes and dislikes.The study also
benefitsfromthe measurementof managementpracticespriorto the measure-
ment of workunit performance.This lends credibilityto causal argumentsby
allowingchangesin managementpracticesto affect subsequentperformance.
If managementpracticespersist over time, the lag is not criticalempirically.
However, if work unit management turns over, then the timing of data
collectionis important.
This studyis weakenedby less-than-perfectmeasuresof independentvariables.
The items we used to measuremanagementpractices came from a survey
designedfor anotherpurposeand are not replicatedin the empiricalliterature.
The findingsare consistent,the independentvariableshave face validity,and
the scales haveconvergentvalidityand acceptableinter-itemreliability.How-
ever, a better study design would include items designed for the purpose of
measuringthe managementpracticesimplied by Hofstede'sdimensionsand
anchored in existing empirical literature.While mindful of Dorfman and
Howell's [1988] admonition that management practices should not be
measuredwith the sameitems Hofstedeused to derivehis culturedimensions,
thereis ampletheoreticaland empiricalliteraturetg
to guidethe measurementof
appropriatemanagementpractices.
776 JOURNAL OF INTERNATIONALBUSINESS STUDIES, FOURTH QUARTER 1996

For example,powerdistanceshould be matchedwith employeeparticipation


and decentralization.We accomplishedthe formerbut not the latter.Manage-
ment practicesthat are congruent with the individual-collectivedimension
might include relianceon group incentive and performancesystems;group
decisionmaking;and family-likeemploymentpracticessuch as employment
for life and the extentto whichthe companyis the sourceof socialand cultural
activitiesfor employeesas well as work.The masculine-feminine dimensionis
nicely capturedby practicessuch as individualgoal-setting,merit-basedpay,
merit-basedpromotion, and opportunity for feedback. Finally, long-term
orientationis the most difficultbecauseit is the newestof the dimensionsand
the least familiarto Westernresearchers.Hofstede [19911identifiedfour prin-
ciples of Confucian teaching that form the foundation for the long-term
dimension.These are acceptanceof unequal relationships(power distance);
the family as prototypefor all organizations(collectivism);treatingothers as
you would want to be treated(the Golden Rule);and a collection of virtues
including working hard, being thrifty, being patient, and persevering.Per-
severanceand patienceare probablythe key characteristicsof the long-term
orientationthat are differentfrom all other dimensions.These are the virtues
we measured,thoughnot as well as we would haveliked.
A final limitationof this study is its relianceon data from employeesof one
company.While this usefullycontrols for both some companyand industry
effects,it limits the generalizabilityof the findings.A multiplecompanystudy
includingmanufacturingemployeeswould add robustness.Sucha studycould
also addressLaurent's[1983]surprisingfindingthat culturaldifferenceswere
more pronounced among foreign employees working in the same multi-
nationalcompanythan among employeesworkingfor differentcompaniesin
theirhome countries.
Futureresearchalso might explorethe simultaneousrelationshipsamong two
or more Hofstede dimensions and managementpractices.Hofstede often
discussesnationalcultureas a cross-tabulationof two dimensions.In his 1991
book, he hypothesizesa whole set of managementpracticesthat ought to be
moreor less prevalentin fourcountryclustersdefinedby the cross-tabulation
of
powerdistanceand uncertaintyavoidance.Empiricaltests of his impliedinter-
actions are an important next step in our understandingof cross-cultural
differencesin the efficacyof managementpractices.Perhapsuncertaintyavoid-
anceis a usefuldimensionwhenconsideredsimultaneously with powerdistance.
Researchon theseand otherinteractionswill contributeto a moresophisticated
knowledgebase for companiesas they try to matchtheirmanagementpractices
with the demandsof the extantnationalculturesin whichthey operate.
NOTES
1. Differencesamong units are tested with analysisof varianceusing the Kruskal-Wallistest
because we have unequal work unit sizes and variances.Results indicatedifferencesamong
CULTURAL FIT WITH MANAGEMENT PRACTICES 777

work units in mean ratingson the managementpracticesvariables.The strengthof within-


unit characteristicsis ascertainedin three ways: agreementwithin group on perceptionof
management practices (eta2), reliability of group mean (Spearman-Brownformula for
intraclasscorrelations,unequal variancesnotwithstanding),and averagestandarderrorsof
the means for managementpractices variables.The intraclass correlations are small, as
expected,rangingfrom .08 to .20. The Spearman-Brownintraclasscorrelationsare between
.74 and .92, an acceptablerangeaccordingto Glick [1985].The standarderrorsare in the .20
to .40 range.
REFERENCES
Adler, Nancy J. 1997 (third edition). Internationaldimensionsof organizationalbehavior.
Cincinnati,Ohio:South-Western CollegePublishing.
Beck,BrendaE.F.& LarryF. Moore.1985.Linkingthe host cultureto organizationalvariables.
In P.J.Frostet al., Organizational
culture,335-54. BeverlyHills,Calif.:Sage.
Blalock,HerbertM., Jr. 1979(secondedition).Socialstatistics.New York:McGraw-Hill.
Burns,Tom& GeorgeM. Stalker.1961.Themanagement
of innovation.London:Tavistock.
Chatman,JenniferA. & Karen A. Jehn. 1994. Assessingthe relationshipbetween industry
characteristicsand organizationalculture: How different can you be? Academy of
ManagementJournal,37: 522-53.
Child, John. 1981. Culture, contingency and capitalism in the cross-nationalstudy of
organizations.In L.L. Cummings& B.M. Staw, Researchin organizationalbehavior,3:
303-56. New York:JAI.
Chow, Chee W, Michael D. Shields & Yoke Kai Chan. 1991. The effects of management
controlsand nationalcultureon manufacturing
performance.Accounting,Organizations,
and
Society,16:209-26.
Crozier,Michel.1964. Thebureaucratic Chicago:Universityof ChicagoPress.
phenomenon.
Day,DavidV.& ArthurG. Bedeian.1991.Predictingjob performanceacrossorganizations:
The
interactionof work orientationand psychologicalclimate. Journalof Management,17:
589-600.
Denison, Daniel R. 1990. Corporatecultureand organizational
effectiveness.New York:John
Wiley.
& Aneil K. Mishra. 1995. Towarda theory of organizationalcultureand effectiveness.
Science,6: 204-23.
Organization
Dorfman, Peter W & Jon P. Howell. 1988. Dimensions of national culture and effective
leadershippatterns:Hofstede revisited.In R.N. Farmer& E.G. McGoun, Advancesin
international
comparativemanagement, 3: 127-50.New York:JAI.
Dyer,Lee & ToddReeves.1995.Humanresourcestrategiesand firmperformance:Whatdo we
know,wheredo we need to go? InternationalJournalof HumanResourceManagement,6:
656-70.
Earley,P. Christopher.1994. Self or group?Culturaleffects of trainingon self-efficacyand
performance.Administrative ScienceQuarterly,
39: 89-117.
Evans, Paul A.L. 1992. Managementdevelopmentas glue technology. Human Resource
Planning, 15: 85-105.
Glick, William H. 1985. Conceptualizingand measuringorganizationaland psychological
climate.Academy of Management Review, 10: 601-16.
Haire,Mason,EdwardE. Ghiselli& LymanW.Porter.1963. Culturalpatternsin the role of the
manager.Industrial Relations, 2: 95-117.
778 JOURNAL OF INTERNATIONALBUSINESS STUDIES, FOURTH QUARTER 1996

Hansen, Gary S. & Birger Wernerfelt. 1989. Determinants of firm performance: The relative
importance of economic and organizational factors. Strategic Management Journal, 10:
399-411.

Hofstede, Geert. 1980. Culture'sconsequences.Beverly Hills, Calif.: Sage.

1991. Culturesand organizations. London: McGraw-Hill.

& Michael Harris Bond. 1988. The Confucius connection: From cultural roots to
economic growth. OrganizationDynamics, 16: 4-21.

Jacobson, Robert. 1990. Unobservable effects and business performance. Marketing Science, 9:
74-95.

Jaeger, Alfred M. 1986. Organization development and national culture: Where's the fit?
Academy of Management Review, 11: 178-90.

Jones, A.P. & Lawrence R. James. 1979. Psychological climate: Dimensions and relationships of
individual and aggregated work environment perception. Organizational Behavior and
Human Performance,23: 201-50.

Kluckholn, Florence R. & Frederick L. Strodtbeck. 1961. Variations in value orientations.


Westport, Conn.: Greenwood.

Kotter, John P. & James L. Heskett. 1992. Corporateculture andperformance. New York: Free
Press.

Laurent, Andre. 1983. The cultural diversity of western conceptions of management.


International Studies of Management and Organizations,13: 75-96.

. 1986. The cross-cultural puzzle of international human resource management. Human


Resource Management, 25: 91-102

Lawler, Edward E, III. 1990. Strategic pay. San Francisco: Jossey-Bass.

Luthans, Fred, Dianne B. Welsh & Stuart A. Rosenkrantz. 1993. What do Russian managers
really do? An observational study with comparisons to U.S. managers. Journal of
InternationalBusiness Studies, 24(4): 741-61.

Mauro, Robert. 1990. Understanding L.O.V.E. (left out variables error): A method for
estimating the effects of omitted variables. Psychological Bulletin, 108: 314-29.

Miller, Katherine I. & Peter R. Monge. 1986. Participation, satisfaction, and productivity: A
meta-analytic review. Academy of Management Journal, 29: 727-53.

Morris, Michael H., Duane L. Davis & Jeffrey W Allen. 1994. Fostering corporate
entrepreneurship: Cross-cultural comparisons of the importance of individualism versus
collectivism. Journal of InternationalBusiness Studies, 25(1): 65-90.

Morris, Tom & Cynthia M. Pavett. 1992. Management style and productivity in two cultures.
Journal of InternationalBusiness Studies, 23(1): 169-79.

Powell, Thomas C. 1992. Organizational alignment as competitive advantage. Strategic


Management Journal, 13: 119-34.

Prescott, John E. 1986. Environments as moderators of the relationship between strategy and
performance. Academy of Management Journal, 29: 329-46.

Puffer, Sheila M. 1993. A riddle wrapped in an enigma: Demystifying Russian managerial


motivation. European Management Journal, 11: 473-80.
CULTURAL FIT WITH MANAGEMENT PRACTICES 779

Ronen, Simca & Oded Shenkar. 1985. Clustering countries on attitudinal dimensions: A review
and synthesis. Academy of Management Review, 10: 435-54.
Schneider, Susan C. & Arnoud DeMeyer. 1991. Interpreting and responding to strategic issues:
The impact of national culture. Strategic Management Journal, 12: 307-20.
Shackleton, Viv. J. & Abbas H. Ali. 1990. Work-related values of managers: A test of the
Hofstede model. Journal of Cross-CulturalPsychology, 21: 109-18.
Slocum, John W & David Lei. 1993. Designing global strategic alliances: Integrating cultural
and economic factors. In G.P. Huber & WH. Glick, Organizational change and redesign,
295-322. New York: Oxford University Press.
S0ndergaard, Mikael. 1994. Hofstede's consequences: A study of reviews, citations and
replications. OrganizationStudies, 15: 447-56.
Triandis, Harry C. 1982. Review of culture's consequences. Human Organization,41: 86-90.
. 1989. The self and social behavior in differing cultural contexts. Psychological Review, 96:
506-20.
Trompenaars, Fons. 1993. Riding the waves of culture. Chicago: Irwin.
Wagner, John A., III & Richard Z. Gooding. 1987. Shared influence and organizational
behavior: A meta-analysis of situational variables expected to moderate participation-
outcome relationships. Academy of Management Journal, 30: 524-41.
Wilkins, Alan L. & William G. Ouchi. 1983. Efficient cultures: Exploring the relationship
between culture and organizational performance. Administrative Science Quarterly, 28:
468-81.
Wright, J.C. & W Mischel. 1987. A conditional approach to dispositional constructs: The local
predictability of social behavior. Journal of Personality and Social Psychology, 53: 1159-77.

You might also like