This document provides a summary and review by Elinor Ostrom of several books that examine the relationship between rational choice theory and institutional analysis in political science. Ostrom summarizes that the books aim to push the boundaries of rational choice theory to better explain empirical phenomena while incorporating an understanding of institutions and cultural traditions. However, she also notes there are productive disputes when scholars let disciplinary lenses bound their vision. Overall, Ostrom argues the field is moving toward complementarity between rational choice theory and other analytical tools in political science.

1. Rational Choice Theory and Institutional Analysis: Toward Complementarity
The Cement of Society: A Study of Social Order by Jon Elster; Solomonic Judgements:
Studies in the Limitations of Rationality by Jon Elster; Political Choice and Social Structure:
An Analysis of Actors, Interests, and Rationality by Barry Hindess; The Common Good:
Citizenship, Morality, and Self-Interest by Bill Jordan; Rediscovering Institutions: The
Organizational Basis of Politics by James G. March; Johan P. O ...
Review by: Elinor Ostrom
The American Political Science Review, Vol. 85, No. 1 (Mar., 1991), pp. 237-243
Published by: American Political Science Association
Stable URL: http://www.jstor.org/stable/1962889 .
Accessed: 24/05/2012 11:17
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
American Political Science Association is collaborating with JSTOR to digitize, preserve and extend access to
The American Political Science Review.
http://www.jstor.org

2. RATIONALCHOICETHEORYAND
INSTITUTIONAL
ANALYSIS:
TOWARDCOMPLEMENTARITY
ELINOROSTROM
IndianaUniversity
TheCementof Society:A Studyof SocialOrder.ByJonElster.New York:Cambridge
UniversityPress,1989. 311p.$44.50 cloth, $15.95paper.
SolomonicJudgements:
Studiesin the Limitationsof Rationality.By JonElster.New
York:Cambridge
UniversityPress,1989.232p. $12.95paper.
PoliticalChoiceandSocialStructure:
An Analysisof Actors,Interests,andRationality.
By BarryHindess.Brookfield,VT:EdwardElgar,1989.216p. $42.95.
The CommonGood: Citizenship,Morality,and Self-Interest.
By BillJordan.Oxford:
BasilBlackwell,1989.190p. $49.95 cloth, $16.95paper.
Rediscovering
Institutions:
TheOrganizational
Basisof Politics.ByJamesG. Marchand
JohanP. Olsen. New York:FreePress.227p. $24.95.
NestedGames:RationalChoicein Comparative
Politics.ByGeorgeTsebelis.Berkeley:
Universityof CaliforniaPress,1989.274p. $28.00.
Empirical methodshave been tra-
ditionallypresentedas the foundationon
which to build the new politicalscience.
Its practitionersstudied empiricalphe-
nomenain the searchfor generallaws to
serve as a basis for a positive science.
Economicsis a differenttype of science,
basedon the powerof deductivetheories
derivedfroma minimumnumberof basic
assumptions about the individual and
how individuals are related to one
anotherand a physical world. It is not
easyto integratethesetwoways of under-
standingthe complex,multileveled,insti-
tutionalworlds in which we live. How-
ever,a growingnumberof politicalscien-
tistshavetriedto bridgethegap.Rational
choicetheoristshave faced the challenge
of explainingthe complex workings of
mostpoliticalinstitutions.
All thesebookspointto thefruitfulness
of the encounterbetweenpowerfultheo-
ries for predictingequilibriain tightly
specified situations and the empirical
worldof politicalscience,wheresomesit-
uations are rigid and unchangingwhile
others are fluid and dynamic. These
booksalsohighlighttheunproductive
dis-
putesthatcanoccurwhenscholarsletdis-
ciplinaryorideologicallensesboundtheir
vision. ThebooksforeshadowwhatIper-
ceive to be the likely long-termoutcome
of thisintellectualconfrontation,namely,
thattherationalchoiceapproach'(1)will
undergo progressive reformation and
developasa majoranalyticaltool usedby
mostpoliticalscientistsand(2)willnot be
theexclusiveanalyticaltool usedbypolit-
icalscientists.Whetherthisoutcomedoes
occur will dependon furtherconfronta-
tion between theory and empirical
researchand an avoidanceof overcom-
mitmentto a singlemodeof analysis.
What one readsin Elster'stwo books
and Tsebelis'sNested Gamesis an effort
to pushthe boundariesof rationalchoice
theoriesas far as possibleto explainem-
piricalphenomenawhileincludinga well-
AMERICAN POLITICALSCIENCEREVIEW
VOLUME 85 NO. 1 MARCH 1991

3. American Political Science Review Vol. 85
developed understandingof institutions
and cultural traditions. March and
Olsen'sRediscoveringInstitutionsstands
on multiplepillarsto examinebotha logic
of consequentiality
(rationalchoice)anda
logic of appropriateness.
They stressthe
latterbutfrequentlyrelyon theformerin
theireffortto bridgemultipleapproaches.
Whiletheydo notentirelysucceedintheir
endeavor,the work illustratesthe grow-
ing interestof politicalscientistsin build-
ing thesebridges.
Thesebooks-and the genuineflower-
ing of work drawingon a rationalchoice
approachto the study of political and
social institutions-give us hope that the
real breakthroughs
in the social sciences
may comein politicalscienceratherthan
in the disciplineswherecommitmentto a
single approachhas almost reachedthe
point of religiousdogma. Further,politi-
cal action based on an empiricallyin-
formedrationalchoicetheorymaygener-
ate resultsthat are far differentfromthe
"neworthodoxy"criticizedso vociferous-
ly in Jordan'sCommonGood.
Irecommend
forreadingby mostpolit-
ical scientists the books by Elster,
Tsebelis, and March and Olsen. In my
estimate, the value added by reading
Hindess's Political Choice and Social
Structureand Jordan'sbook is less ob-
vious. Theirworks illustratethe current
debate among political scientists about
the role and usefulnessof rationalchoice
theory and the need for a more institu-
tionalapproach.
Amongthemanyissuesraisedby these
volumes, threeare at the core of current
debatesaboutthe futuredirectionsof the
discipline:the limits of rational choice
theories(1) as tools for normativeanaly-
sis, (2) as tools forempiricalexplanation,
and (3) the role of history, institutions,
and culturaltraditionsin rationalchoice
explanationsof politicalbehavior.
The Limitsof RationalChoice
Theoriesin NormativeAnalysis
Rationalchoice theoriesare primarily
normative theories. A rational choice
theory "tellsus what we ought to do in
orderto achieveour aimsas well as pos-
sible. Itdoesnot, in thestandardversion,
tellus whatouraimsoughtto be"(Elster,
SolomonicJudgements,
3). Inotherwords,
a rationalchoicetheorycanbe viewedas
a "theoryof advice"thatinformsindivid-
uals or, potentially,collectivitiesof indi-
viduals,abouthow bestto achieveobjec-
tives-whatever thesemay be.
In this normativerole, a theory must
treat all individualsas attemptingto be
rational.Byassumingthateveryoneis ra-
tional (i.e., takes actions conformingto
those predictedby the theory)a theorist
specifiesthebeststrategythateveryactor
couldadoptin situationscharacterized
by
particularstructures.The concept of an
equilibriumrequiresan assumptionthat
everyoneadoptstheirbeststrategyaspre-
dicted by the theory and that once at
equilibriumno actoris unilaterallymoti-
vated to adoptanotherstrategy.
Elsterarguesthat all theoriesof social
behaviorimplicitlyassumeindividualra-
tionality. To engagein any communica-
tion or debateis to presumethat others
are rational. Otherwise, why bother
spending time sharing informationand
arguing?Any effortto try to understand
others is an effort to learn about their
goals and how they view the opportuni-
tiesandlimitsof thesettingsinwhichthey
find themselves.While presumingratio-
nality,Elsteralsoanalyzesthelimitsof ra-
tional choice theoriesin orderto under-
standwherethey can be usedwith confi-
dence and where applying a theory is
questionable. In contrast, March and
Olsen argue that behavior is primarily
rule-governedrather than consequence-
governedand thata logicof appropriate-
ness is the way to theorizeabouthuman
behavior:"Politicsis organizedby a logic
238

4. RationalChoiceTheoryand Institutions
of appropriateness.Political institutions
are collections of interrelatedrules and
routinesthatdefineappropriate
actionin
termsof relationsbetweenrolesandsitua-
tions. The process involves determining
what the situationis, what role is being
fulfilled,andwhat the obligationsof that
role in the situationare"(p. 160). March
and Olsenfurtherarguethat "inour cul-
turewe justifythe action (appropriately)
by a logicof consequentiality"
(p. 162).In
thisview, individualsjustifychoicesusing
a means-endscalculus,ratherthan con-
ducting means-ends analysis prior to
choice.Actionis "basedon a logic of ap-
propriateness,and justification[is]based
on a logic of consequentiality."
Whenra-
tionalchoicetheoryis viewedas a source
of ex post facto justificationbut not as a
logic of choice, obvious tensions and
hypocrisiesresult.
As aninstitutionalanalyst,Icanhardly
arguewith Marchand Olsen'sstress on
theimportanceof rule-governed
behavior
in all socialcontexts.Stressingtheimpor-
tanceof rules,however,does not require
denying the importance of calculated
choice by relegating analyses of the
reasonsfor choiceto after-the-fact
justifi-
cationof previouslytakenactions.Rather
than conceptualizing rule-governed
choice as more importantthan rational
choice,a generalapproachwouldattempt
to explainhow bothrulesandanticipated
consequencesaffect behavior and out-
comes. In a Westernculturalcontext (to
usetheirreferent)
rulesrarelyinformindi-
vidualsasto theexactactionsto betaken.
Rather,some actions are ruled in (per-
mitted), and others are ruled out (for-
bidden).Choicesmadeby rule-following
individualsarefromwithinthesetof per-
mittedactions. In a rule-governedsitua-
tion, theactionsthatareruledout should
not (andarenot)considered
inthelogicof
choice by most individualsin most cir-
cumstances.2That is what is meant by
rule-governed.Choicefromamongthose
actionsthat are ruledin cannotbe made
on thebasisof institutionalrules.3Choice
from within the alternativesallowed by
the rulesof thegamemustbe understood
differently from the determination of
whatis or is not ruledin. Gametheoryin
extensiveformhasprovedto be a power-
ful tool in applyinga logicof consequen-
tiality to choices within constraintsde-
finedby the rulesof a game.
A key stepinjoiningthelogicof appro-
priatenesswitha logicof consequentiality
is to accept permitted actions as the
"defaultcondition"for rule-governed
be-
havior. Whenpermittedactionsarecon-
sideredto be the defaultcondition,rule-
governed behavior is based on actions
beingpermitteduntila ruleforbidsor re-
quires an act. When prohibitedactions
areconsideredto bethedefaultcondition,
rule-governedbehavior is based on ac-
tions beingforbiddenuntila rulepermits
ormandatesanact. Differentdefaultcon-
ditions lead to fundamentallydifferent
systemsof lawandto differentpatternsof
behaviorwithina society. Elster's
distinc-
tion betweena principleof legalityand a
positive conceptionof the law illustrates
the differencesbetweenthesetwo default
conditions.A principleof legality is de-
finedas follows: (1) an individualaction
is permittedunlessthereexistsa law that
expresslyand unambiguouslyforbidsit,
and (2) state interferenceis forbidden
unlessthereexistsalawthatexpresslyand
unambiguouslyauthorizesit; whereasa
positive conceptionof the law is defined
as follows:(1)an individualactionis per-
mittedif thereexistsa law thatexpressly
authorizesit, and (2) the state has the
right to interferein all unauthorizedac-
tions even if they are not expresslyfor-
bidden(SolomonicJudgements,
197).
Ifa positiveconceptionof lawis theac-
cepted default condition, March and
Olsen'slogicof appropriateness
is thepri-
mary mechanismfor individualchoice.
Oncerulesaredetermined,littleroomfor
individualinitiativeis left.A positivecon-
ception of the law has been the founda-
239

5. American Political Science Review Vol. 85
tion for social orderin many societiesin
the world. Evenin these societies,how-
ever, the difficultyof creatinga complete
and coherentpositive conceptionof the
law leaves room in the intersticesof the
law (and in relationshipto covert or il-
legalbehavior)forextensiveuseof a logic
of consequentiality.Whena principleof
legality is used as the defaultcondition,
many fewer situationsare narrowlyde-
fined by rules and considerableroom is
left open for consequentialthinkingafter
learningthebasicrulesof thegame.Thus,
recognizingthe importanceof rules and
socialnormsis not inconsistentwitha ra-
tional choice interpretation
of individual
action within the constraintsof a rule-
ordered set of relationships. We will
returnto this themebelow.
The Limits of Rational Choice
Theories As a Positive Theory
To use rationalchoice as a theory of
advice,onehasto useit to predictandex-
plainactions.Thus,thenormativeaspects
of a theoryare intricatelyrelatedto the
positive aspectsof a theory.If a rational
choice theory is indeterminateor inade-
quatein predictingthebehaviorof others
in particularsituations,its only use as a
normativetheoryis in utopianthought.If
othersdo not behaveas predictedby the
theoryorif therearemanyequilibria
pre-
dicted,the best strategyto adoptmay be
unknown or quite different from that
recommendedby theory. Indeterminacy
(lackof a uniqueequilibrium)
andinade-
quacy (lackof empiricalsupportfor pre-
dictions)aretwo limitationson theappli-
cabilityof rationalchoicetheoriesto nor-
mative and positive analysesextensively
explored by Elster and by March and
Olsen.
In Solomonic JudgementsElstercon-
cludesthat the "standard
version"of ra-
tionalchoicetheoryis most applicablein
choice situationsinvolving one actor or
many actors confronting problems of
mediumsize(p. 27). Wherethereareonly
a small number of actors and where
uncertainty,complexity, nonrecurrence,
and duration make for loosely defined
problems, this theory is limited in its
capacityto predict.Elsterdoes not go as
far as March and Olsen, however, in
referring
to somegroupprocessesas hav-
ing the structuresof a "garbage
can."
Nonrecognition
of thelimitsof a theory
can lead to hyperrationalitydefined as
"the failure to recognize the failure of
rational-choicetheory to yield unique
prescriptionsor predictions"(Solomonic
Judgements, 17; emphasis original).
Hyperrationalityoccurs when a scholar
(1)triesto eliminateuncertaintyin analy-
sis of an inherentlyuncertainsituation;
(2)reframeshis or herview of a situation
to include the dimensionsthat lead to
determinateanswersand ignores others
that generateindeterminacy;and (3) at-
tempts to assess the best action in each
situationratherthan "lookingfor more
general rules that cover many similar
cases"(p. 26).
As hisparadigmcaseof indeterminacy,
Elsterdevotes Chapter3 of Solomonic
Judgements
to an analysisof custodydis-
putes where two competentparentsare
engagedin a protractedconflict. In this
inherently indeterminate situation, he
arguesthat one option rarelyconsidered
-the useof a lotteryto decidecustody-
may produce better outcomes than a
hyperrationalinvestment of substantial
resourcesin attemptsto determine
theex-
actbenefitsandcostsof assigningcustody
to one parent.
InChapter4 of SolomonicJudgements,
Elsterextends his criticismof forms of
hyperrationalityto the faith of many
scholarsandpoliticiansin comprehensive
social engineering.The concept of an
identifiable"commongood"soughtafter
as a collectiveobjective(arguedfor with
poetic force by Jordan)is given a pretty
strong drubbingby Elster.Elsterpoints
240

6. RationalChoiceTheoryand Institutions
out that individualsfrequently"do not
know what they want; or do not know
what they know; or fail to do what they
havedecidedto do"(p. 181).Buteachin-
dividualhasa will oranegothatattempts
to organize the potentially fragmentary
parts of that individual: "Societies,by
contrast,haveno centre"(p. 181).Conse-
quently,tryingto view societiesthrough
thelensof a unitaryactormodelcanpro-
ducegrosslycounterintuitive
results.
The limits of rational choice theories
appearnot only when the unitaryactor
model is projected onto societies as a
whole but also when narrowmodels of
thetheoryareusedto predictandexplain
the behaviorof real individualsin com-
plex situations. The essential working
partsof anindividualmodelof choicein-
volve assumptionsmade about (1) how
actors value outcomes (and the actions
leadingto theseoutcomes),(2)how actors
acquireandprocessinformation,and (3)
whatformulasareusedby actorsto select
their own strategies. Very specific
assumptionsabout valuation, informa-
tion, andcalculationmadein the models
used by most rational choice theorists
yield usefulpredictionsand explanations
in some situations, particularly those
most closelyresemblinga perfectlycom-
petitive market allocatinghighly divisi-
ble, easily transferable,privatelyowned
goods with few externalities. Rational
choicetheorieswerefirstdevelopedto ex-
plainactionsandoutcomesinthesehighly
constrained
situationsconstitutedby a set
of institutions.4
Insituationsresembling
a
spot commodity market, the following
frequently
usedassumptions
arerelatively
good firstapproximations
giventhatpar-
ticularinstitutionalsetting: an actor (1)
canmakea completepreferenceordering
over the alternativesavailable; (2) has
completeinformationabout the relevant
data for making this decision; and (3)
maximizespersonal benefits minus per-
sonalcosts. Becauseof the largenumber
of actors, the institutionalrules, and the
attributesof the goods assumedin the
model, an equilibriumoutcome (market
clearingpriceandquantity)is likely, and
what any actor does will not affectthat
equilibrium.In such situations, no one
can improve the outcome by following
advice other than that providedby ra-
tionalchoicetheory.
Thefurtherfromthestarkinstitutional
environmentof a competitivemarketora
similarly constraining institution, the
moreapparentarethelimitsof a narrow-
ly focused model of rational choice.
Assumingcompleteinformationwhenin-
dividualsfaceanextremelackof informa-
tionabouttherelevantstrategiesandout-
comes is not likely to produceas useful
explanationsof behavioras assumingun-
certaintyaboutthe exactvaluesandrela-
tionshipamong some variables.Tsebelis
demonstratesthat it is not necessaryto
keep the perfectinformationassumption
to makerationalchoiceexplanationsand
thatthepowerof an explanationmay in-
creaseby assuminguncertainty
whenthat
is a strongelementof the situationfacing
actors.Manyof theexcitingformulations
madeby rationalchoicetheoristsinvolve
the useof learningmodelsandevolution-
ary models to supplementtheir under-
standingof behaviorin highlystructured
and stablesituations.
In Cementof Society, Elsterexamines
many of thesekindsof situationsas well
as thosewherea narrowinterpretation
of
rationalchoice leads theoriststo predict
highly undesirableoutcomes. Elster'sef-
fort is to show how rules and norms
enablerationalindividualsto cope more
effectively with these situations. Recent
worksby FritzScharpf(both1990)com-
plementthe basic thrusttakenby Elster
but addressthe limits of rationalchoice
fromtheconcernsof anempirical
scientist
wantingto findthemostpowerfultheory
to explainempiricalphenomena.Scharpf
concludes that game-theoreticalmodels
can be used to explain the outcomes in
many of the gamesthat realactorsplay.
241

7. American Political Science Review Vol. 85
Tsebeliscertainlydemonstrates
this to be
the casein his book.
Jordan,Hindess,andMarchandOlsen
criticize the rational choice approach
from the outsideand urgealternativeap-
proaches.Elstercriticizesfrom the inside
and urges complementaryapproaches.
Tsebelispushesthe boundariesof the ap-
proachoutwardandappliesit successful-
ly to explain quite complex behavior.
These books thus represent a healthy
scientific debate about a core issue in
political science:How are we to under-
standhumanbehavior?Thisis wherethe
great political philosophersbegan their
analyses,andthisiswherethedebatecon-
tinuesinto the twenty-firstcentury.The
synthesis likely to come from this con-
frontationwill involve a view of human
action that is both rule-governed
and ra-
tional. To be rule-governed,the rational
individual must know the rules of the
gamesinwhichchoicesaremadeandhow
to participatein the craftingof rules to
constitutebettergames.Thus,history,in-
stitutions, and cultural traditions will
play a moresignificantrole in the evolu-
tion of rational choice theories in the
futurethanthey have in thepast.
History, Institutions,and
CulturalTraditions
All of thesebooksstresstheimportance
of history, institutions, and culture.
Hindess and Jordanargue that rational
choice theoryinadequatelyaddressesthe
context in which individuals make
choices. From the perspective of the
nineties,some of the earlierwork in the
rationalchoice traditionjustifythis criti-
cism.Tsebelis'bookofferssubstantial
evi-
dence that the earlierefforts to develop
"institution-free"
modelsarenot the only
way rationalchoicetheorieswillbedevel-
oped and appliedin politicalscience.By
focusing directly on nested games,
Tsebelisexaminessituationswherepoliti-
cians must play in simultaneousgames
(e.g., legislativeand electoralarenas)or
wherethe game for choosingthe rulesis
playedbeforeplayingthegamesin which
rulesconstrainand strategiesarechosen.
He uses a theory of nestedgamesto ex-
plain why BritishLabourparty activists
replacetheirMPs and lead theirpartyto
electoral defeat; why Belgian political
elites initiate conflict and mobilizelarge
numbersof citizensfor particularistic
in-
terests;why Belgianinstitutionshave so
consistently involved federal principles
assigningjurisdiction
overdifferentissues
to the groupsor territorialunits that are
most concernedabout these issues; and
why all Frenchpoliticalparties(including
the Communist)adopt a particularcom-
bination of competitiveand cooperative
strategieswithin the context of French
electoralinstitutions.All of theaboveand
the many otherempiricalanalysesin the
book are explainedto a greateror lesser
extent in the comparativepolitics litera-
tureusinga wide varietyof differentap-
proaches.WhatmakesTsebelis'contribu-
tion significantis his furtherdevelopment
of a rationalchoicetheoryto explainsuch
a wide diversityof empiricalphenomena
thatthey do not appearon the surfaceto
have a similar underlying explanation.
Further,he uses the theory to explain
many puzzles that have been left unex-
plainedin previousstudies.
Books like these show us that a real
convergenceis occurringasmorepolitical
scientistspresumethatindividualsarera-
tional and searchfor institutionalstruc-
tures to help explain behavior that ap-
pears to be irrationalupon first inspec-
tion. Forbehaviorin repetitivesituations
wherethe stakesarehigh, assumingirra-
tionality leads to an explanation of
desperationand an intellectualtrap. We
learnby confrontingfailuresratherthan
glossingover them.To offercoherentra-
tional choiceexplanationsof complexin-
stitutionalbehavior,however, requiresa
deepunderstanding
of thelogicof institu-
242

8. RationalChoiceTheoryand Institutions
tions and institutionalchoice. Thus, ra-
tionalchoiceandinstitutional
analysisare
likely to be essentialcomplementsin the
political science of the twenty-firstcen-
tury.
Notes
1. By rational choice approach, tradition, or
framework
Imeanallworkthatisbasedonmethod-
ologicalindividualism
andassumesthatindividuals
compare
expectedbenefitsandcostsof actionsprior
to adoptingstrategiesforaction.By rationalchoice
theoriesI meanthemorespecificassumptions
made
by a scholaraboutthe type of information,valua-
tion, andcalculationinvolvedin individualchoice.
WhenIreferto a modelof a rationalchoicetheory,I
meana specificformalrepresentation
of a theory.
Multiplemodelsexist for any theory, and several
theoriesare usuallyconsistentwith an approach.
Thedistinction
I usebetweenapproach,theory,and
modelis theresultof manydiscussions
withVincent
Ostromand other colleaguesat the Workshopin
PoliticalTheoryandPolicyAnalysis.
2. The difference between rules devised by
humansand physicallaws, however,is that rules
canbebroken.Thus,individuals
do haveto decide
whetherthey wish to follow the rulesrelatedto a
particular institution or not. When following rules
produces better consequences for an individual than
not following rules, little calculative effort needs to
be devoted to a decision. The key difference between
the position that a rational choice theorist would
take and that of March and Olsen is that the former
presumes that all individuals will at some time be
sorely tempted to break some rule due to the very
high payoffs that could result. Those who adopt an
assumption of universal opportunism posit that only
the expected cost of external sanctions keep individ-
uals in line. Both Elster and Tsebelis presume only
that individuals can be tempted rather than that in-
dividuals do behave in a universally opportunistic
manner.
3. Such choices are frequently the result of per-
sonal commitments to particular modes of action.
The term rule is applied by some scholars to per-
sonal commitments as well as to socially shared
commitments. I prefer to use the term strategy for
the plan of action that an individual adopts and rule
for the shared prescriptions with which actions
must, must not, or may, be taken.
4. The simple phrase set of institutions (assuming
law and order) is used as a necessary condition for
allowing a market to work. While the phrase is short
and simple, the set of institutions creating property
rights, monitoring buyers' and sellers'behavior, and
enforcing rules is an incredibly complex, multilevel
institutional arrangement in which competitive
markets are nested.
ElinorOstromis Co-Directorof Workshopin PoliticalTheoryandPolicyAnalysis,
and Professorof PoliticalScience,IndianaUniversity,Bloomington,IN 47405.
243