Institutions:RulesorEquilibria?AvnerGreifandChristopherKingston1Introd - PDF document

Institutions:RulesorEquilibria?AvnerGreifandChristopherKingston1IntroductionInrecentyears,theinterestinrationalchoiceanalysisofinstitutionshasreceivedsubstantialimpetusfromanaccumulatingbodyofevidencedemonstratingtheimportanceofasociety’sinstitutionsindeterminingitseconomicoutcomes.Econo-metricstudieshaveuncoveredcorrelationsbetweeninstitutionalvariablessuchasthesecurityofpropertyrights,theruleoflaw,andtrust,andeconomicandpoliticaloutcomesincludinglevelsofproduction,saving,andcorruption.Historicalstudieshaverevealedtherolethatinstitutionsplayedinlong-runtrajectoriesofindustrialandcommercialdevelopment.Studiesofthedevelopingworldandofcountriestransitioningfromsocialismhaverevealedthechallengesinvolvedincreatingwell-functioninginstitutions,thebenetsthatcanbeobtainedwheninstitutionalchangeandeconomicreformaresuccessful,andthedangersthatensuewhentheyarenot.Stimulatedbythesedevelopments,theconceptualframeworksemployedbyscho-larsstudyinginstitutionshavealsobeenevolving,asoldframeworkshavebeenadaptedandnewframeworkshaveemergedtoexploreoldandnewquestionsabouthowinstitutionsfunction,howtheychange,andhowtheyaffecteconomicbehaviorandoutcomes.Therational-choiceapproachtoinstitutionalanalysisdoesnotrequireustoassumethatpeoplearealways‘rational’,orthatinstitutionsarechosenrationally.Rather,itholdsthatarational-choiceperspectiveenablesustogenerateatheorywithempiri-callyrefutablepredictionsabouttheinstitutionsthatcanprevailinagivensituation.A.Greif(DepartmentofEconomics,StanfordUniversity,Stanford,CA94305,USAe-mail:avner@stanford.eduC.KingstonDepartmentofEconomics,AmherstCollege,AmherstMA01002,USAe-mail:cgkingston@amherst.edu Forexample,LaPortaetal.(),KeeferandKnack(),EasterlyandLevine(Forexample,Milgrometal.(),North(),Greif(Forexample,Roland(),Aoki(),Qian(N.SchoeldandG.Caballero(eds.),PoliticalEconomyofInstitutions,DemocracyandVoting,DOI10.1007/978-3-642-19519-8_2,Springer-VerlagBerlinHeidelberg2011 Thisinvolvestwokeyquestions:rst,howinstitutionsareselectedandsecond,howpeoplearemotivatedtofollowinstitutionalizedpatternsofbehavior.Onestrandofthoughtwithintherational-choiceapproachtoinstitutionalanalysis,theso-called‘institutions-as-rules’approach,emphasizestheimportanceofatheoryofselectionofinstitutions,whileanemergingalternativeapproach,the‘institutions-as-equilibria’lineofanalysisemphasizestheimportanceofatheoryofmotivation.Theinstitutions-as-rulesapproach,followingNorth(,p.3),identiesinstitutionsas“therulesofthegameinasociety”,includingboth“formal”rulessuchasconstitutionsandlawsenforcedbythestate,and“informal”constraintssuchas“codesofconduct,normsofbehavior,andconventions”,whicharegenerallyenforcedbythemembersoftherelevantgroup(North,,p.36).Manykindsofformalrulesareselectedthroughacentralizedprocessofbargainingandpoliticalconictbetweenindividualsandorganizationswhoattempttochangetherulesfortheirownbenet.Inothercases,formalorinformalrulesmaybeselectedinadecentralizedwaythroughevolutionarycompetitionamongalternativeinstitutio-nalforms.Ineithercase,theinstitutions-as-rulesviewholdsthatinstitutionsareultimatelybestunderstoodfromafunctionalistperspectivethatrecognizesthattheyareresponsivetotheinterestsandneedsoftheircreators(althoughthereisnoguaranteethattherulesselectedwillbeefcient).Withintheinstitutions-as-rulesview,theenforcementoftherulesisconsideredasadistinctissuefromtheformationandcontentoftherulesthemselves.Enforcingtherulesinvolves“enforcementcosts”.Theformalandinformalrules,togetherwiththeir“enforcementcharacteristics”constitutetheinstitutionalstructurewithinwhichinteractionsoccur.Thus,theinstitutions-asrulesapproachemploysarational-choiceperspectivetostudytheformationofinstitutions,butatheoryofmotiva-tion–explainingpeoplefollowparticularrulesofbehavior–isnotintegratedintotheanalysis.Agrowingbodyofrecentresearchoninstitutionsplacesatheoryofmotivationatthecenteroftheanalysis,andtherebyendogenizesthe“enforcementoftherules”,bystudying‘institutions-as-equilibria’.Thisperspectivefocusesonhowinteractionsamongpurposefulagentscreatethestructurethatgiveseachofthemthemotivationtoactinamannerperpetuatingthisstructure.Togiveasimpleexample:intheUnitedStates,people(nearlyalways)driveontheright-handsideoftheroad.Thisregularityofbehaviorgeneratesexpectationsthatmotivatethebehavioritself:peopledriveontherightbecausetheyexpectotherstodoso,andwishtoavoidaccidents.Ofcourse,itisalsoa“rule”thatonemustdriveontheright.However,manyalternativetechnologicallyfeasiblerules(forexample,womendriveontherightandmenontheleft)wouldgenerateexpectationswhichwouldfailtomotivateapatternbehaviorconsistentwiththerule:thatis,suchpatternsofbehaviorarenotequilibria,andeveniftheywereformallyspeciedasa“rule”wewouldnotexpectthemtoemergeasinstitutions,becausethe“rule”wouldnotbeself-enforcing.Foreveryonetodriveontheright,however,isoneoftwopotentiallyself-enforcing“rules”whichcouldemerge(orbeenacted)asanequilibrium.Thecrucialpointisthatwhilea“rule”mayserveasacoordinationdevice,itisfundamentallytheexpectedbehaviorofothers,ratherthantheruleitself,which14A.GreifandC.Kingston motivatespeople’sbehavior.Asimilarlogiccanbeusedtoexamineeconomic,political,andsocialinstitutionseveninsituationsinvolvingspecializedactorsandmorecomplexformal“rules”.Fromtheinstitutions-as-equilibriaperspective,itisalwaysultimatelyexpectationsaboutthebehavioroftheotheractors(includingthoseinspecializedenforcementrolessuchaspolice,judges,etc.)thatcreatetheinstitutionalconstraintswhichmoldpeople’sbehavior,andallsuchbehaviormustthereforeultimatelybeexplainableendogenouslyaspartoftheequilibrium.Despitetheirdifferences,theinstitutions-as-rulesandinstitutions-as-equilibriaapproacheshavemuchincommonandarebestviewedascomplementsratherthansubstitutes.Bothseektoadvanceapositiveanalysisofthenon-technologicaldeter-minantsoforderandregularitiesofhumanbehavior.Recentadvancesintheliteraturecombineelementsofthetwoperspectives.Thischaptersurveysthesedevelopmentsandhighlightspromisingdirectionsforfutureresearch.Aswewilldiscuss,theinsti-tutions-as-rulesframeworkhasbeenfruitfullyappliedtoshedlightontheemergenceandfunctioningofavarietyofinstitutions,includingcommunities,organizations,andpoliticalandlegalinstitutions.However,wewillarguethatbyendogenizingtheissueofenforcement,theinstitutions-as-equilibriaapproachenablesamoresatisfactorytreatmentofseveralkeyissues,includingpromotingourunderstandingofprocessesofinstitutionalchange.2InstitutionsasRules:ConceptualIssuesAsdiscussedabove,themostcommonlyciteddenitionofinstitutionsisthatadvancedbyDouglassNorth:institutions“aretherulesofthegameinasociety,ormoreformally,arethehumanlydevisedconstraintsthatshapehumaninteraction”(North1990,p.3).Institutionsincludebothformalrules,whichareexplicit,writtenrulessuchaslawsandconstitutions,andinformalconstraintssuchasconventionsandnorms.InNorth’stheory,formalrulesarecreatedbythepolity,whereasinformalnorms“areapartoftheheritagethatwecallculture”(p.37)andthereforeimpervioustodeliberatehumandesign.Thefocusoftheanalysisisthereforeonformalrules,namely,rulesthatareexplicitlyandintentionallycreated.Toillustratetheinstitutions-as-rulesapproach,considertheframeworkdevelopedbyOstrom(2005),whoenvisagesahierarchywithseverallevelsofrules:“operationalrules”whichgovernday-to-dayinteractions;“collective-choicerules”,whicharerulesforchoosingoperationalrules;“constitutionalrules”(rulesforchoosingcollective-choicerules);“metaconstitutionalrules”(rulesforchoosingconstitutionalrules);andatthehighestlevel,thebiophysicalworld(p.58).Thatis,eachlevelinthis Forarecentdiscussion,seeGreif().KingstonandCaballero()surveytheoriesofinstitutionalchange.North(,p.47)envisagesasimilarhierarchywithfourlevelsofformalrules:constitutions,statuteandcommonlaws,specicbylaws,andindividualcontracts.Institutions:RulesorEquilibria?15 hierarchyofrulesconsistsofrulesthatgovernhowrulesatthelowerlevelarecreated.Forexample,constitutionalandcollective-choicerulesprovidethestruc-turethatgovernsthechoiceofoperationalrules.Higher-levelrulesarealsomoredifcultandcostlytochange.Whentheyperceivethatexistingrulesgoverningtheirinteractionsatonelevelareunsatisfactory,individualsaredrivento“shiftlevels”andtrytochangetherules.Apoliticalbargainingprocessensues.Eachindividualcalculatestheirexpectedcostsandbenetsfromanyproposedinstitutionalchange,andaninstitutionalchangecanoccuronlyifa“minimumcoalition”necessarytoeffectchangeagreestoit.Whatconstitutesa“minimumcoalition”isdeterminedbythehigher-levelrules;forexample,inademocracy,amajoritywouldconstituteaminimumwinningcoalition;inadictatorshipthedictatoralonemightconstituteaminimumcoalition.Therefore,thesetofrulesthatultimatelyemergeswilldependontheperceivedinterestsoftheactorsinvolvedinsettingtherules,ontheabilityofvariousinterestgroupstoactcollectivelytomaketheirinterestscount(Olson),andonthehigher-levelrulesthatdeterminehowthoseindividualinterestsareaggregated.Thereisnoguaranteethatthisprocesswillleadtotheselectionofefcientrules.Inmanycases,thosewithpoliticalpowermaytrytoselectrulestogeneratedistribu-tionalbenetsforthemselves;thatis,tomaximizetheirwelfareratherthanthatofsocietyasawhole.Toexplainwhysocieties“choose”inefcientinstitutions,how-ever,itisnotsufcienttonotethatthegroupsinpowerhaveintereststhatdivergefromtherestofsociety.Ifaninstitutionalchangecouldincreaseefciencyandeconomicoutput,whycannotthebeneciariesofthechangeagreetoredistributethegainstocompensatethelosers?Acemoglu(2003)arguesthatthekeyproblemiscommitment:thepowerfulcannotcrediblycommitnottousetheirpowerfortheirownbenetastheopportunityarises,andothergroupscannotcrediblycommittocompensatethepowerfulforgivinguptheirpower.Asaresult,thesetofbargainswhichcanbestruckisrestrictedtothosebargainswhichcanbesustainedasequili-briumoutcomes(Fearon2007;Greif19982006).Becausethereisnoexternalauthoritytoenforceinter-temporalbargains,politicallypowerfulgroupsmayblockchangesthatwouldbebenecialoverall,orimposeinefcientchangesthatbenetthemselvesattheexpenseofothers.Fundamentally,therefore,asatisfactoryunder-standingoftheseaspectsofinstitutionalchangerequiresarecognitionthattheprob-lemisnotjustchoosingnewrules,butthemorerestrictiveproblemofengineeringamutuallybenecialshifttoanew,self-enforcingequilibrium.Wewillreturntothisissuelater.Asecond,complementarystrandofthoughtwithintheinstitutions-as-rulesapproachviewsthedevelopmentofrulesasanoutcomeofevolutionarycompeti-tionamongalternativeinstitutionalforms.Alchian()arguedthatcompetitivepressureweedsoutinefcientformsoforganizationamongrmsincompeti-tivemarkets,becausermsthatdevelopmoreefcientorganizationalformswillbemoreprotable,andtheuseoftheserulesandformsoforganizationwillthere-foretendtospreadthroughgrowthorimitation.Demsetz()extendedtheevo-lutionaryargumenttothedevelopmentofproperty-rightsrules,hypothesizingthattheserulesdevelopandadjustasaresultof“legalandmoralexperiments”which16A.GreifandC.Kingston “maybehit-and-missprocedurestosomeextent”,butwhichonlyproveviableinthelongruniftheygenerateefcientoutcomes.Hayek(1973)arguesthatgroupsororganizationsthat,byaccidentordesign,developlessefcientruleswillnotsur-vivecompetitionwithgroupsthatdevelopmoreefcientrules.Therefore,throughgroupselection,ruleswillevolvetowardsoptimality.TheevolutionaryapproachndsitsmostprominentmodernexpressioninOliverWilliamson’s“Transactionscosteconomics”(TCE).Accordingtothisview,‘trans-actioncosts’ariseinmanytransactionsbecauseoftheboundedrationalityandopportunismofthetransactingparties(Williamson2000).Dependingontheattri-butesofaparticulartransaction,somesetsofrules(‘governancestructures’)willleadtomoreefcientoutcomesthanothers.Thetransactions-costeconomicsapproachassumesthatthemostefcientinstitutionalforms(thosewhich‘minimizetrans-actionscosts’)willemerge.So,forexample,ifachangeinproductiontechnologyrendersexistinginstitutionsinefcient,thenovertime,new,moreefcientinstitu-tionalformswillemergetoreplacethem.Althoughthepolitical-designandevolutionaryapproachesenvisagequitediffer-entprocessesfortheselectionofrules,thetwostrandsofresearcharebestviewedascomplementary.Bothtreatinstitutionsassetsofrules(or“governancestructures”);andbothfocusonhownewrulesareselectedratherthanhowtheyareenforced.Differentinstitutionsareassociatedwithdifferent“transactioncosts”,including“monitoringcosts”and“enforcementcosts”,butthenatureofthesecostsisnotpartoftheanalysis.Theconceptof“transactionscosts”iswidelyusedinNewInstitutionalEconom-ics.Thetermisgenerallyusedverybroadlytoincludethecostsofndingtradingpartners,negotiatinganddrawingupcontracts,monitoringcontractualpartners’behaviorandenforcingagreements,andothercostsincurredinanefforttodene,measureandenforcepropertyrightsoragreementstoexchangepropertyrights.Transactioncostsmayalsoincludethecostsofpoliticalactivity,bargaining,legalaction,andsooninvolvedindeliberateeffortstocreatenewrules,thecostsofinefciencyresultingfromcommitmentproblemsandotherformsofpoliticaltrans-actioncosts,aswellasallthecostsinvolvedinsettingup,maintainingandchangingthestructureofrulesandorganizations,andmonitoringtheactionsoftheagentsgovernedbythoserules.Inshort,anydifferencebetweenthevalueofoutputgene-ratedintherealworld,wherearealtransactionisgovernedbyrealinstitutions,andanimaginedworldwithoutanyagencyproblemsorinformationasymmetries(andthereforeaworldinwhichnogovernanceisrequired),includinganydeviationfromrst-bestproductionandexchange,canbecalleda“transactioncost”.Despitethisbreadth,theconceptof“transactioncosts”hasachievedwideacceptanceasananalyticaltoolinthetheoreticalliteratureoninstitutions,particu-larlywithintheinstitutions-as-rulesframework.Theusefulnessoftheconceptis Williamsonreferstothisasthe“discriminatingalignment”hypothesis.Thus,forWilliamson,“Theoverallobjectoftheexerciseessentiallycomesdowntothis:foreachabstractdescriptionofatransaction,identifythemosteconomicalgovernancestructure”(Williamson,p.234).Institutions:RulesorEquilibria?17 thatitprovidesameasureofinstitutionalefciency.However,theuseoftrans-actionscoststerminologyriskscloudingtheissueofenforcement.Toillustrate,consideranagencyrelationshipbetweenamanagerandtheworkerswithinarm.Thesaleoftheagent’slaborservicesinvolvesafundamentalproblemofexchange:thedecisionofwhethertoworkhardismadebytheagent,butitaffectsthewelfareoftheprincipal.Giventhisfundamentalagencyproblem,differentinstitutionswillgiverisetodifferentpatternsofbehavior.Theexplicitandmeasurabletransactionscostsinsuchasettingmightincludethecostsofhiringamanagertomonitortheworkersandmeasuretheirperformance,aswellasthecostsofdesigninganorgani-zationsoastoenablethismonitoringtooccur,choosingaproductionprocesswhichfacilitatessuchmonitoring,installingsurveillanceequipment,andthelegalcostsofnegotiatingemploymentcontracts,andsuingorringashirker;andsoon.Inaddition,ifintheenditprovestoocostlytomotivatetheworkertoactasshewouldinarst-best(zerotransactionscost)world,thentheresultinginefciencywouldbeanother(implicit)transactioncost.Butwhiletheconceptof“transactionscosts”canserveasahandyshorthandtodescribehowwelltheseproblemsaresolved,allofthese“costs”ultimatelyderivefromtheagencyproblemsandinformationasymmetrieswhichgiverisetothefun-damentalproblemofexchangeinthe(potential)transactionofinterest.Byseparat-ingthe“costs”ofrunningtheeconomicsystem–monitoring,enforcement,andsoon–fromthesystemitself,theinstitutions-as-rulesapproachcloudstheissueofpeopleactastheydo,andbecomesapooranalyticalsubstituteforanaccountofhowbehaviorisactuallymotivatedwithinalternativeinstitutionalregimes,noneofwhichwillapproximatethezero-transactions-costideal.Thatis,theproblemofdesigningefcientinstitutionsisnotfundamentallyaproblemofchoosingrulessoastominimize“costs”,butaproblemofaligningincentivesinawaywhichgene-ratesthemaximumpossiblebenet,givenafundamentalproblemofexchange.Higherefciency(oralowertransactioncost)isadesiredoutcomeofasuccessfulsolutiontothisproblem,butitisnottheproblemitself,andfocusingontransactionscostsasacatch-allminimandrisksmaskingtheessenceoftheproblem,whichisoneofaligningincentives.3InstitutionsasRules:Applications3.1CommunitiesandNetworksCommunityenforcementreferstoasituationinwhichbehaviorwithinagroupisgovernedby“rules”whichareenforcedbythemembersofthegroupthemselvesratherthanaspecialistthird-partyenforcer.Oneviewholdsthatthesekindsofinformalrulesarebesttakenaspartofaxed,exogenously-givenculturalheritage(Williamson).Otherauthors,however,considerthatinformalrulescontinu-allyadaptandevolve.Forexample,basedonhisstudiesofcattlefarmersinShastacountyandNewEnglandwhalers,RobertEllickson()hypothesizesthatgroups18A.GreifandC.Kingston withinwhichinformation(gossip)circulateseasilyandinformalpowerisbroadlydistributedwilltendtodevelopefcientinformalrules.Ostrom()foundthatmanycommunitiesmanagetodeveloprulestosuccessfullyavertthetragedyofthecommonsinthemanagementofcommon-poolresources,suchassheries,forests,andcommonpasture.Othercommunities,however,donot,andOstromfoundthatsuccessfulrulesweremorelikelytoemergeingroupswithsmallnumbersofdeci-sionmakers,longtimehorizons,andmemberswithsimilarinterests.Ascommunitiesbecomelarger,therefore,bothEllickson’sandOstrom’sstudiessuggestthatinformalcommunityenforcementislesslikelytobeabletosupportefcientoutcomes.Forexample,astheonlinecommunityoftradersoneBaygrewinthelate1990s,the“trust”sustainedbyamultilateralreputationmechanismbasedonuserfeedbackhadtobegraduallysupplementedbyformalrulesdevelopedbyeBaytodiscouragecheating,resolvedisputes,andpreventillegaltrades(Baron20013.2OrganizationsOrganizationsareakintoarticialcommunitiesofindividualsbroughttogetherforaspecicpurpose–suchasproduction,politicalactivity,religiousworship,recreation,andsoon.Whilesomeorganizationsmaybeginasinformalgroupswhosememberslaterdecidetodevelopaformalgovernancestructure,othersarecreateddenovoby“entrepreneurs”withagoalinmind.Assuch,organizationsarebothcohesiveentitieswhichimpactandinteractwiththebroaderworldaroundthem,andgovernancestructureswhichdevelopformalrulestogoverntheinterac-tionsamongtheirmembersandbetweenmembersandoutsiders.Withintheinsti-tutions-as-rulesframework,differentauthorshavefocusedoneachofthesetwoaspectsoforganizations.Someauthors,notablyDouglassNorth,havetreatedorganizationsprimarilyasuniedentitiesthatinteractwiththebroadereconomicandpoliticalsystemwithinwhichtheyareembedded,andinparticular,mayactas“players”ofthepoliticalgame,attemptingtoalterbroaderinstitutionalrulesforthebenetoftheirmem-bers.ThisaspectoforganizationswillbediscussedinSect.(“Politics”).Theotheraspectoforganizations–theirinternalgovernance–isstudiedineconomicsprimarilyintheguiseofthetheoryoftherm.Asiswellknown,themoderntheoryofthermoriginateswithCoase’s(insightthatorganizationsandmarketsarealternativemodesoforganizingtransac-tions,andtheclaimthatthescopeofactivitycarriedoutwithinorganizationswillthereforebedeterminedsoastominimize“transactionscosts”.Toexplainthestruc-tureofanorganization,therefore,weneedtoexplainitsfunction:whatcontractualproblemitefcientlysolves.Butwhywouldefcientorganizationsemerge?Onepossibilityisthatthestructureoforganizationsisaproductofrationaldesign.Iftheorganization’screatorshaveacorrectunderstandingoftheeffectsofdifferentorganizationalforms,thenitmaybereasonabletoassumethattheywilldesignef-cientorganizations.Institutions:RulesorEquilibria?19 However,analternativeexplanationfortheemergenceofefcientorganizationsisthatevolutionarypressureforcesrmstoselectefcientorganizationalformsbydrivingless-efcientorganizationsoutofbusiness.Alchian()wasanearlyproponentofthisview,anditalsoimplicitlyunderliesWilliamson’s“TransactionsCostEconomics”,whichassumesthatorganizations(governancestructures)willdevelopsoastoachieveanoptimal(efcient)matchwiththetransactionstheygovern.Theevolutionaryapproachhastheadvantage,notedbybothAlchianandWilliamson,thatitenablesustoassumethatefcientinstitutionswilldevelopevenifthepeopledesigningthemareboundedlyrational.Ifaparameterchange,suchasachangeintechnology,rendersexistinginstitutionsinefcient,thenovertime,byaccidentordesign,somermswilldevelopmoreefcientsetsofrules(“gover-nancestructures”),andthroughcompetitivepressure,thesenewinstitutionswillgraduallyspread,sothattheinstitutionsgoverningtherelevanttransactionwillevolvetowardoptimality.Thus,theusefulnessoftherational-choiceframeworkdoesnotrestonanassumptionofrationality.Thevalidityofthisapproach,however,restsontheimplicitassumptionthattherearedeeperunderlyinginstitutionsthatleadtotheselectionofoptimal(ef-cient)institutions.Theissueofwhatexactlytheseunderlyinginstitutionsareisfrequentlyleftunexplored,andthustheanalysiscanofferonlyapartialexplanationfortheobservedcongurationofrules.Nevertheless,forthepurposeforwhichitwasdeveloped,namelyexaminingthegovernancestructuresofrmsoperatingincompetitivemarketswithinamoderneconomy,thisapproachworkswellandisan“empiricalsuccessstory”(Williamson,p.607).Theassumptionthatorganizationsareorganizedefciently(whetherthroughevolutionordesign)alsounderpinsmuchofthemoderntheoryoftherm,inclu-dingtheliteratureonprincipal-agentproblemswithintherm,whichstudieshowmanagementcandesignoptimalincentivesystemstomotivateworkers;theproperty-rightsapproachfollowingHart(),whichpostulatesthattheboundaryoftherm(ownershipofassets)isdeterminedinsuchawayastominimizetheinefciencieswhichresultfromtheinabilitytowritecompletecontracts;andthetheoryofmechanismdesign.Informalrulesandnorms,suchasa“corporateculture”,mayalsodevelopwithinorganizations,includingrms.Theinternalgovernanceoforganizationstypicallyinvolvesacombinationofbothformalandinformal“rules”.Forexample,oneapp-roachtoovercomingtheprincipal-agentproblembetweenmanagementandwork-erswithinarmisthroughoptimalwageandbonusstructuresbasedoncontractibleoutputmeasures.However,analternativewaytomotivateworkereffort,giventherepeatednatureoftherelationship,isviathethreatofringaworkercaughtshirking(BowlesandGintis).Whiletheformalcontract,accordingtowhich NelsonandWinter()builtanevolutionarytheoryofthermbasedontheevolutionofroutines–sequencesofactionwhichcoordinatetheactivitiesofmanyindividuals–ratherthanrules.Routinesevolveassuccessfulrmsexpandandtheirroutinesareimitated–perhapsimperfectly–byothers,creatingatendencytowardstheadoptionofefcientroutines(althoughpossiblywithconsiderableinertia).20A.GreifandC.Kingston theworkerispaidawageforshowinguptowork,regardlessofhereffort–isenforceableinthecourts,theworker’seffortlevelisnotcontractible,andsotheemploymentrelationshipisgovernedbybothformalandinformalrules:higheffortisenforcedinformallythroughthreatofnon-renewaloftheformalcontract.3.3Politics,InformalRules,andInstitutionalChangeThestate,ofcourse,isthemostimportantsourceofformalrules,includinglaws,constitutionalrulesanddecreespassedbyrepresentativebodies,votedonbycitizens,orproclaimedbykings.Standardneoclassicaleconomicsassumestheexistenceofawell-functioning“state”,andstateactivitiessuchastaxation,regulation,andtheprovisionofpublicgoodsaretreatedaswell-functioningpolicyinstrumentsinthehandsofabenevolentpolicymaker.Whilethistreatmentofthestateisusefulforsomepurposes,itiswoefullyinadequateforothers.Itmakesimprobableassumptionsaboutthestate’sabilitytoobtainandprocesstheinformationneededtoarriveatanoptimalconclusion(Hayek1945),anditobscuresthefactthatpolicydecisionsaregenerallytheresultofbargainingandnegotiationamongorganizationsandindivi-dualswithdivergentinterests,andthatimplementingthesedecisionsinvolvesmoti-vatingandcoordinatingtheorgansofthestate,suchasregulatoryagencies,courts,andthepolice.Akeyfunctionofthestate,takenasagiveninneoclassicaleconomics,istoprovidesecurityofpropertyrightsandcontractenforcement.Intheabsenceofastate(anarchy),individualsmustinvestresourcesintheprivateproductionofsecuritybyacquiringacapacityforviolence(Skaperdas2006).Thewell-knownHobbesianjusti-cationforthecreationofthestateisthatthepresenceofahigherauthorityenablespeopletoreplacethecostlyandinefcientspontaneousorderofanarchywithasetofrulesdesignedtoimproveoverallwelfare.InYoramBarzel’sHobbesiantheoryoftheoriginsofthestate(Barzelindividualsbegininastateofnaturewithoutinstitutions,andtheynditintheirintereststocreateastate,asamonopolistofviolence,toprovideorder.However,theywishtoefcientlylimitthestate’sscopeofactivity.Thisraisesthequestionofwhythestate(whichBarzeltreatsasasingleactor)wouldobeythe“rules”thatitssubjectscreateforit,ratherthanusingitscapacityforviolencetoexpropriatethoseunderitsruleorexpandthescopeofitsactivitybeyondthatwhichisoptimal.Barzelnotesthisdanger,andpostulatesbeforethepeoplecreateastate,theywillalsocreatecollective-actionmechanismsthatconstrainthestate’sactionsbyenablingthemtooverthrowthestateifitbecomespredatory.However,inkeepingwiththeinstitu-tions-as-rulesapproach,Barzeltreatstheenforcementofthesecollectiveactionmechanismsasexogenous.Asaresult,theenforcementproblem(keepingthestate “Althoughthe“social”arrangementsusedtoenforcedecisionsbycollective-actionmechanismsseemtobeofutmostimportance,thereislittlethatI,asaneconomist,cansayaboutmostofthem.Isimplyassumethatsucharrangementsexistandareputintouse”(Barzel,p.119).Institutions:RulesorEquilibria?21 honest)ismerelypushedbackonelevel;ultimately,theenforcementoftheformalrulesistakenasexogenous.Theproblemofempoweringthestatetocreateorderwhileconstrainingitfrompredationisoffundamentalimportance.Djankovetal.()postulatethatthe“institutionaldesign”ofthestateinvolvesafundamentaltradeoffbetween“disor-der”and“dictatorship”:creatingamorepowerfulstatehelpstoreducedisorderandtherisksofprivateexpropriation,butatthecostofincreasingthecostsofdictator-ship,corruptionandexpropriationbythestate.Eachsocietyhasasetoffeasiblecombinationsofdictatorshipanddisorder(an“institutionalpossibilitiesfrontier”),whichdependsonavarietyofsocietalcharacteristicsincludingtechnology,culture,education,socialcapital,ethnicheterogeneity,history,factorendowmentsandthephysicalenvironment.InDjankovetal.’sbasicmodel,societieschooseanoptimalpoliticalsystem(thatis,onewhichminimizesthesumofthecostsduetoprivateandpublicexpropriation)subjecttotheconstraintofitsinstitutionalpossibilityHowever,thereareavarietyofpotentialimpedimentstotheselectionofefcientpoliticalrules.Djankovetal.arguethatcountrieswhichareformercoloniesmighthaveinefcientrulesiftherulesweretransplantedorimposedbytheirformalcolonialmastersratherthanarisingindigenously.LaPortaetal.()ndthatcountries’legaloriginsaffecteconomicoutcomes.Thecivillawsystem,theyarguefavorsagreaterdegreeofstatecontrolandregulation,whereasthecommonlawsys-temreliesmoreonmarket-supportingregulationandprecedent-settingprivatelitigation.Manyauthorsemphasizethatdistributionalconictcanleadtotheselectionofinefcientrules.Forexample,Libecap(1989)exploresthedevelopmentofthe“propertyrights”rulesthatgoverntheuseofavarietyofresourcessuchassheries,mineralrights(mining),andtheuseofpublicland.Differentrulesentaildifferentdistributionalconsequences,andindividualsandgroupsthereforeengageinbargain-ing,lobbying,andpoliticalactiontotrytoaltertherulesfortheirownbenet.AsinOstrom’sschema,thisrule-changingactivity(“contracting”)isitselfagamegovernedbyahigherlevelofpoliticalrules,andthesehigher-levelrules,togetherwiththeacti-vitiesandperceptionsoftheactorsthereforeshapethedirectionofinstitutionalchangeofthelower-level(propertyrights)rules.Acemoglu()andAcemogluandRobinson(2006)emphasizetheimpor-tanceofcommitmentproblemsasanimpedimenttotheselectionofefcientrules.Politicalincumbentsmightbewillingtomakeconcessionstodisenfranchisedgroupsinordertoavertacostlyorviolentrevolution,butiftheycannotcrediblycommitthemselvestohonortheircommitmentstoreformafterthemomentofcrisisispassed,thenwhenevergroupshavetheopportunity,theywillseizepowerandcraftrulestobenetthemselveswithoutregardfortheothergroups. See,however,Hadeld(),whocastsdoubtontheimportanceofthecivil-law/common-lawdistinction,andprovidesaricherandmorerenedalternativesetofkeyparametersfortheclassicationoflegalregimes.22A.GreifandC.Kingston Evenifasocietydoesinitiallyselectruleswhichare“efcient”inastaticsense,theserulesmayultimatelyturnouttobesuboptimalinadynamicsense.Forexample,EngermanandSokoloff()arguethatthesoilandclimateinEurope’sSouthAmericanandCaribbeancoloniesweresuitablefortheproductionofcashcrops,suchassugar,thatcouldbeefcientlyproducedonlargeslaveplantations,resultinginhighlyunequaldistributionsofwealth,income,andhumancapital,whichinturnenabledtheelitestoestablishlegalandpoliticalinstitutionsthatpromotedtheirinterests.IntheNorthAmericancolonies,incontrast,theinitialfactorendowmentsweremorefavorabletotheproductionofcropsandlivestockthatcouldbeefcientlyproducedinsmallfamilyfarms.Thisledtothedevelopmentofmoreegalitariananddemocraticpoliticalinstitutions,higherlevelsofpublicgoodsprovision(suchasprimaryschooling),andgreaterlevelsofsocialmobility.Acemogluetal.()tellarelatedstory,butwiththefocusonthediseaseenvironmentandindigenouspopulationdensityratherthansoilandclimateasthekeyexogenousvariablesexplainingtheinitialdevelopmentofstateinstitutions.InplaceswhereEuropeansfoundsettlementdifcult,theycreated“extractivestates”aimedattransferringresourcestothemothercountry.InareasmoreconducivetoEuropeansettlement,theyfounditmoreprotabletobuildinstitutionsaimedatpro-tectingprivatepropertyandencouraginginvestment.Theseinstitutionspersistedevenafterindependence,andledtoa“reversaloffortune”inthenineteenthcentury,becauseregionsthathadpreviouslybeenpoorinheritedinstitutionsthatlaterenabledthesocietiestoindustrialize.Theseargumentsgivehistoryaroleinexplainingthescopeandfunctioningofthestate.Institutionsdevelopedasanefcientresponsetocircumstancesinonetimeperiodmaypersisteveniftheylaterbecomeinefcient.Butwhydoinstitu-tionspersist?Again,thebasicanswerwithintheinstitutions-as-rulesapproachisduetoNorth(),whodevelopedatheoryofinstitutionalchangethatcombinesdeliberatechangesinformalruleswithevolutionarychangeininformalrules.InNorth’stheory,giventhecurrentstructureofformalandinformalrules,entrepre-neursformorganizationstotakeadvantageofperceivedopportunities.Overtime,astheyacquireskillsandknowledge,theymaynditworthwhiletoattempttochangethestructureofformalrules.Whenchangesinformalrulesoccur,thentheinformalruleswhich“hadgraduallyevolvedasextensionsofpreviousformalrules”(p.91)adjustinresponse,andtheendresult“tendstobearestructuringoftheoverallconstraints–inbothdirections–toproduceanewequilibriumthatisfarlessrevolutionary”(North,p.91).Thus,Northarguesthatbecauseofthepersistenceoforganizationsandinformalrules,overallinstitutionalchangeis“overwhelminglyincremental”(Northp.89),andthatinstitutionalchangeisapath-dependentprocess:“theconsequenceofsmalleventsandchancecircumstancescandeterminesolutionsthat,oncetheyprevail,leadonetoaparticularpath”(North1990,p.94).Currentinstitutionsprovideincentivestocreateparticularkindsoforganizationsandtoinvestinparticularkindsofskillsandknowledge.Theyalsoaffectthedistributionofwealthandpoli-ticalpower,thepreferencesoftheactors,andthestockofphysicalandhumancapital.AlloftheseendogenousparameterchangesinturnaffectthecostsandbenetsofInstitutions:RulesorEquilibria?23 alternativeinstitutions,people’sperceptionsofnewpossibilities,andtheirabilitytobringaboutorstieinstitutionalchange.Inalltheseways,pastinstitutionscaninuencethedirectionofinstitutionalchange(Libecap;Pierson;NorthBuildingonNorth’swork,agrowingrecentliteratureconsidersprocessesofinstitutionalchangethatexploretheinteractionbetweenformalandinformalrules.InRoland()’stheory,informalrules(“slow-movinginstitutions”)arecon-stantlyevolving,andifthesechangesbecomeincompatiblewithexistingformalrules,thenpressureforchangebuildsup,leadingtoperiodicabruptandsubstantialchangesinformalrules(“fast-movinginstitutions”).BrousseauandRaynaud(2008buildatheoryinwhichnewrulesbeginasinformal,localandexibleorders,whichcompeteforvoluntaryadherents.Successfulrulesspread,andastheyspread,theybecomeincreasinglyglobalandmandatoryand“harden”intorigidformalrules.Aldashevetal.()showthatchangesinformalrulescanalteroutsideoptionsandthereforebargainingpowerwithininformalrelationships,andtherebyshiftcustomaryinformalrulesinthedirectionoftheformallaw,evenifitisneverexplicitlyused.Onedifcultywhicharisesinthinkingaboutinstitutionalchangeinthisway–asaninteractionbetween“formal”and“informal”rules–isthatthenatureofthe“informalrules”isoftenleftrathervague,andhowtheyinteractwithformalrules–forexample,whichruleisfollowedwhenthetwokindsofrulesconict–remainsunclear.Asnotedabove,theinstitutions-as-rulesapproachtreatsthequestionofhowrulesareenforced,andthereforewhytheyarefollowed(ornotfollowed),asaseparateissuefromtheircontent.Thusbydenition,ifbehaviordoesnotconformtoformalrules,bydefaultitisattributedto–andassumedtobegovernedby–unobservedinformalrules.Yet,sinceinformalrulesaregenerallyimplicit,itishardtoobservewhattheseinformalrulesare,whetherinfacttheyareindeedbeingfollo-wed(andifso,why),andwhatkindsofbehaviortheyareaffecting,andinwhatway.Attributingunexplainedbehaviortoinformalrulesthereforeamountstoaleapoffaiththatinvokesamysteriousandscienticallyuntestableexplanationfortheobservedbehavior.Theproblemiscompoundedbythefactthattheterm“informalrules”hasbeenusedtodescribeseveralquitedistinctphenomena.Someauthorstreatinformalrulesasinternalized“ethical”codesofconductwhicharedirectlyreectedinplayers’preferences(e.g.,Ostrom).Forothers,informalrulesareruleswhicharenotwrittendown,orwhicharenotenforcedbythestate.Stillothersidentifyinformalrulesasself-enforcingcodesofconduct,sharedcultural“focalpoints”,oras“socialnorms”enforcedwithinacommunityusingamultilateralreputationmechanism–orasallofthesethings,astheoccasiondemands.Forsome(e.g.Williamsoninformalinstitutionschangeonlyoveraperiodofcenturiesormillennia,sotheymaysafelybetakenasexogenousandxed,whileothers,suchasRoland,holdthatgradualchangesininformalrulesareoftenanimportantpartofthestoryofinsti-tutionalchange.Ultimately,therefore,theinstitutions-as-rulesapproachislimitedinitsabilitytoexplaininstitutionalchangebecauseakeyelementinhibitingandshapingthe24A.GreifandC.Kingston directionofinstitutionalchange,informalrules,originatesoutsidetheanalyticalframework.Forexample,Ostrom(,p.138)notesthat“manywrittenstatementshavetheformofaruledonotaffectbehavior.Suchstatementsareconsideredrules-in-formratherthanrules-in-use.”Yet,becauseshetreatstheenforcementofrulesseparatelyfromtheircontent,anyexplanationofwhatmakessomerules“rules-in-use”whileothersrulesareignoredisoutsideherframework.Shenotesthat“insettingswhereaheavyinvestmentisnotmadeinmonitoringtheongoingactionsofparticipantsconsiderabledifferencebetweenpredicatedandactualbehaviorcanoccur,”(p.21),butachievingthismonitoringandenforcementistreatedsimplyasacost;theincentivesofthemonitorsarenotexamined.4Self-EnforcingInstitutions,or“Institutions-as-Equilibria”:ConceptualIssuesThecoreideaintheinstitutions-as-equilibriaapproachisthatitisultimatelythebehaviorandtheexpectedbehaviorofothersratherthanprescriptiverulesofbehaviorthatinducepeopletobehave(ornottobehave)inaparticularway.Theaggregatedexpectedbehaviorofalltheindividualsinsociety,whichisbeyondanyoneindivi-dual’scontrol,constitutesandcreatesastructurethatinuenceseachindividual’sbehavior.Asocialsituationis‘institutionalized’whenthisstructuremotivateseachindividualtofollowaregularityofbehaviorinthatsocialsituationandtoactinamannercontributingtotheperpetuationofthatstructure.Thefocusonregularitiesofbehaviorandthemotivationtofollowthemrespondstotheobservationthatthesefactors,ratherthanrules,arethedirectcauseofdistinctwelfare-relatedoutcomes.Thecorruptionplaguingmanypoliticalsystemsintheworldisnotcausedbyanabsenceofrulesprescribingpreventivemeasures.Itisduetoparticularregularitiesofbehavior.Focusingonmotivationhasthekeyadvantageofavoidingtheconceptualdif-cultiesthatcomewithtreatinginstitutionsasrules.Forexample,thelegalspeedlimitonhighwaysinMassachusettsis65mph,butthislimitiswidelyignored.Thisisnottosaythatthereareno“rules”,however.Policeofcersdosometimespullovercarstravelingat85mph,buttheyneverpullovercarstravelingat68mph.Whataccounts Thisideabuildsonthe‘conventions’literature(e.g.,Sugden).SeealsoSchotter(Calvert(),Aoki(),Dixit(),KingstonandCaballero()andGreif(Aregularityofsocialbehaviordoesnotimplyuniformityofbehaviorasitisacharacteristicofaggregatesofindividualsandnotofeachindividual.Furthermore,socialbehaviorisusuallyconditionalonsocialrolesanddoesnotnecessarilyimplythesamebehaviorbyindividualswiththesamerole.Thebehavioralregularityof‘malesproposetoafemaleandonlywhentheycansupportafamily,’forexample,capturesgenderrolesandimpliesthatsomemaleswillnevermarryandtheagesofthosewhodo,willvary.Similarly,regularityofbehaviorisnotnecessarilyfrequentbehavior.TheprocessofimpeachingaUSpresidentisregularizedalthoughrarelyInstitutions:RulesorEquilibria?25 forthedifferencebetweenthebehaviorspeciedbythe“formalrule”andthebeha-vioractuallyobserved?Fromtheinstitutions-as-rulesperspective,thestandardanswerwouldbethatthepoliceandmotoristsmustbefollowingan“informalrule”–forexample,thatthetruespeedlimitis75mph.Butthisinvokesanexogenousandad-hocexplanationforpreciselywhatwewouldmostliketoexplain.Focusingonmotivationcomplicatestheanalysis,however.Onereasonisthatregularitiesofbehaviorareoftencausedbytheneteffectofmultiple,andpossiblyconicting,motivatingfactors.Thefearoflegalsanctionsmightmotivateateenagedrivertoslowdown,butsocialpressurefromhispeersmighthavetheoppositeeffect.Theevolvinginstitutions-as-equilibriaapproachhasnotyetconvergedonanagreeddenitionofinstitutions.Ontheonehand,Calvert(),forexample,literallyequatesinstitutionswithgametheoreticequilibria.“Thereis,strictlyspea-king,noseparateanimalthatwecanidentifyasaninstitution.Thereisonlyrationalbehavior,conditionedonexpectationsaboutthebehaviorandreactionsofothers“Institution”isjustanamewegivetocertainpartsofcertainkindsofequilibria”(pp.22–23).Thepremiseofthisdenition,however,istoorestrictive.Gametheoryprovideslittleguidanceforidentifyinginstitutionsorstudyingtheirdynamics.Greif,Chaps.2and5)denesaninstitutionasasystemof‘institutionalelements,’particularlybeliefs,norms,andexpectationsthatgeneratearegularityofbehaviorinasocialsituation.Theseinstitutionalelementsareexogenoustoeachdecision-makerwhosebehaviortheyinuence,butendogenoustothesystemasawhole.Thesocial‘rules’whichemergecorrespondtobehaviorwhichisendogenouslymotivated–constrained,enabled,andguided–byself-enforcingbeliefs,normsandexpectations.Inaddition,foraninstitutiontobeperpetuated,itsconstituentele-mentsmustbe(1)conrmed(notrefutedoreroded)byobservedoutcomes(2)rein-forcedbythoseoutcomes(inthesensethatitsabilitytobeself-enforcingdoesnotdeclineovertime)and(3)inter-temporallyregeneratedbybeingtransmittedtonewcomers.4.1Self-EnforcingExpectationsandMotivationAnempirically-orientedanalysisrelyingontheinstitutions-as-equilibriaapproachfocusesprimarilyonmotivationprovidedbyself-enforcingexpectations(behav-ioralbeliefs).Suchananalysisusuallybeginsbyidentifyingthe‘essential’physi-cal,technologicalandsocialattributesnecessaryforthesituationtobeofinterest.Inthecaseofregularitiesofbehavioramongdrivers,forexample,essentialattri-butesincludethattheremanydriverswhohavepropertyrights(oruserrights)incars,canbenetfromdrivingcomparedtoalternativemodesoftransportation,canobserveothercars,andprefertoavoidaccidents.Withoutanyofthesefeatures,consideringdrivingbehaviorismeaningless.Similarly,theanalysiswouldbetoogeneralwithoutmorenarrowlydelineatingtheregularitiesofbehaviorweareinterestedin:isitthedirectionoftrafc,priority-rulesatintersections,speedingorpassing?26A.GreifandC.Kingston Byfocusingonasituation’sessentialattributes,weinitiallysetasidethosepotentiallyrelevantsocialconstructsthatweinitiallywishtotreatasexogenoustotheanalysis.Inthecaseofdriving,thesemightincludesuchconstructsasdrivers’licenses,socializationtodrivecarefully,oraHighwayPatrolAgencywiththecapa-citytoimposelegalsanctions.Initiallyignoringsuchpotentiallyimportantcon-structsisnotastatementabouttheirirrelevancebutameanstoanalyticallyexaminewhethertheyarerelevant,whytheyarerelevant,andtowhateffect.Thenextstepintheanalysisistofocusonthesetofself-enforcingexpectationsandtheimpliedbehaviorthatcanprevailinthissituation,bymodelingthesituationasagame(specifyingthesetofplayers,theirpossibleactions,theorderofmoves,information,andpayoffs)andndingequilibria.By“self-enforcingexpectations”wemeanthatifthedecision-makerssharetheexpectationthatotherswillgenerallyfollowtheequilibriumbehavior,theneachofthemwillbemotivatedtofollowitaswell(theNashcriterion).Fromeachdecision-maker’sperspective,theothers’expectedbehaviorconstitutesthestructuremotivatinghertoconformtothebeha-viorexpectedofher.Butbyconforming,shecontributestomotivatingotherstoconformtoo.Thus,thestructureisself-perpetuating,andalthoughitisbeyondthecontrolofeachdecisionmaker,itisendogenoustoallofthemtakentogether.Notethatthe“self-enforcing”requirementincludesexpectationsabouthowotherswillbehaveinsituationsthatwouldnottranspireinequilibrium.Forexample,ifaplayerdoesnotstealbecauseofafearofpunishment,the‘off-the-path’expectationofpunishmentmustbecredible(thisthesub-gameperfectionrequirement).Havingfoundequilibriaintheminimalgame,wecannextexaminehowvarioussocialconstructscanchangethesetofself-enforcingexpectationsbychangingtheexpectedresponsesbyotherplayerstoparticularactions.Whentheseexpectationsarecredible,thecostsandbenetsassociatedwithactionsintheminimalgamearechanged,andthesetofpotentiallyself-enforcingbehaviorsmaybeenlarged.Forexample,thecreationofa“group”cancreaterestrictionsonentrytothesituation(whotheparticipantsare)andchangethepatternofrelations(e.g.,repeatedinter-actionsamongthesameindividuals).Otherkindsofsocialconstructsmightaltertheinformationstructure,orintroduceanewactorwiththeabilitytopunishorrewardplayers(e.g.,ajudge).Theintroductionofnewsocialconstructscanchangepeople’sexpectations(andthereforeincentivesandbehavior)inmanyways.Sanctionscanbecoercive(suchasviolenceorimprisonment),social(suchasostracism),oreconomic.Guiltandthefearofexpectedpunishmentintheafter-lifeareothermeanstolinkpastactionstofuturerewards.Theinstitutions-as-equilibriaapproachfocusesonhowsuchexpec-tationsareformed,why,andtowhateffect.Notethatthisinvolvesmuchmorethanjusttheintroductionofnew“rules”.Inordertoshiftpeople’sexpectations,cogni-tivecategories(e.g.,“honesty”,“cheating”)needtobecoordinateduponsothatallplayerssharecoordinatedexpectationsaboutpunishment.Ifthedesiredbehavioristobeself-enforcinginthemodied,extendedgamethatincludesthenewinterac-tions,thenthepunishmentshouldbesufcientlycostlytomakedeterrenceeffective.Thosewhoaretoretaliatemusthavetheinformationaboutwhoandwhentopunish,whichpotentiallyincludesmotivatingthosewhoknowaboutthetransgressiontoInstitutions:RulesorEquilibria?27 informothers.Peoplemustalsobemotivatedtopunish,astheexpectationthatpunishmentwillbeinictedhastobecredible.Furthermore,theyhavetohavethephysicalcapacitytopunishandthosewhoaretobepunishedshouldnotbeabletoevadepunishment.4.2RulesandOrganizationsintheInstitutions-as-EquilibriaIntheinstitutions-as-rulesapproach,rulesareinstitutionsandinstitutionsarerules.Rulesprescribebehavior.Intheinstitutions-as-equilibriaapproach,theroleof“rules”,likethatofothersocialconstructs,istocoordinatebehavior.Becausetherearemultiplepotentiallyself-enforcingexpectationsinagivensituation,coordinationmechanisms,includingrules,playanessentialroleingeneratingregularitiesofbehaviorandsocialorder.Rulesfulllthiscoordinatingrolebyspecifyingpatternsofexpectedbehavior,andalsobydeningthecognitivecategories–signs,symbols,andconcepts–onwhichpeopleconditiontheirbehavior.Actionshavetobegivenmeaningsbecause,forexample,‘cheating’isnotnaturallydened,butitmustbedenedbeforeitcanbediscouraged.Aroadsigninstructingadrivertoyieldatapedestriancrossinghasmeaning,andmotivatesbehavior,onlybecauseitisacom-ponentofasystem(“rulesoftheroad”)thatmotivatesbehaviorbasedonroadThebehaviorthatpeoplecanbemotivatedtofollowdependsonthesecogni-tivecategoriesandontherules’abilitytocoordinateexpectationsbasedonthesecategories.Focusingonmotivationexposesthelimitsontherealitiesthathumanscanuserulestoconstruct.Inorderfora“rule”tomatter,thebehaviormustbeself-enforcinganditmustbeconditionedonobservableaspectsofthesituation.Ifdriverscannotobserveapedestrian’sage,theycannotconditiontheirbehavioronit.Anditmustbesufcientlycostlytocircumventthecategories.Forexample,arulewhichconditionsbehaviorongendermaynotbeself-enforcingifmalescaneasilypretendtobefemalesandviceversa.Ofcourse,thebehavioralexpectationsandcognitivecategorieswhichpeopleactuallyusetocoordinatetheirbehaviormaybequitedifferentfromthosespeciedby‘formalrules’.Nevertheless,weobservethatexplicit“rules”areoftenforma-lizedanddisseminatedinacentralizedmanner.Fromtheinstitutions-as-equilibriaperspective,thecreationofsuchformal“rules”canbeinterpretedasanattemptachieveacoordinatedshiftofmanypeople’sexpectations,whileconvincingtheagentsthattheseexpectationsareindeedwidelyshared.Thismechanismcanalsobeused,ofcourse,toservetheinterestsofthepoliticallypowerful–thosewiththepowertochangeformalrules.Butifthenewrulesdonotspecifyaself-enforcingpatternofbehavior,theymaynothavetheirdesiredeffect.Organizations,too,aresocialconstructsthatchangethesetofself-enforcingexpectationsamongtheagentsintheoriginalinteraction.Formalorganizations,suchasparliamentsandrms,andinformalorganizationssuchascommunitiesand28A.GreifandC.Kingston businessnetworks,haveadualrolebothasinstitutionsthatgoverntheirmembers’behavior,andasinstitutionalelementswithinthebroaderinstitutionsofsociety.Withinthegroup,anorganizationcanchangetherelevantrulesofthegame,suchasinformation,actions,andpayoffssets,andcanthereforeincreasethecredibilityandseverityofsanctions,specifyrules,andcreatesharedknowledge.Organizationsmayalsoplayaroleinattemptingtoshapethepreferencesofcommunitymembers,particularlychildren,throughaprocessofsocialization.Anorganizationanditsmembersalsointeract,individuallyandasagroup,withtheoutsideworld,andthebeliefs,normsandexpectationsthatgoverntheinternalinteractionsbetweenthemembersoftheorganizationwilloftendifferfromthosegoverningtheirinteractionswithoutsiders.Apoliceforce,forexample,hasinternalstructuresandrulestogovernitsmembers’behavior,butitalsoactsasanorgani-zationforenforcingotherrulessetbythegovernmentofthesocietyofwhichitisapart.Therelianceonorganizationsisfundamentallyduetothefactthatorganiza-tionshavecapabilitiesthataremorethanthesumoftheindividualcapabilitiesoftheirmembers,duetotheirabilitytocoordinatetheirmembers’activities,econo-miesofscaleandscopeintheireffortstochangetherulesofthegame,andduetotheorganizations’longertimehorizonandmemory.5InstitutionsasEquilibria:Applications5.1MarketsandNetworksTheabilitytoengageinvoluntaryexchangeencouragesproduction,specialization,andinnovation,andisakeyprerequisiteforeconomicefciency.However,inallbutthesimplestmarketexchanges,enforcementproblemsariseduetothe“funda-mentalproblemofexchange”(Greif).Forexample,inlabormarketandcreditmarkettransactions,thereisanunavoidableseparationbetweenthequidandthequo,andatleastonepartythereforemayhaveanopportunityto“cheat”theother:aborrowermaychoosenottorepayaloan,oranemployeemaychoosenottoworkhard.Marketscanfunctiononlywhenthisfundamentalproblemofexchangeisovercome.Neoclassicalmicroeconomicstendstoeitherassumeawayenforcementpro-blems,ortotakethepresenceofwell-functioningmarket-supportinginstitutionsasgiven.Theinstitutions-as-rulesapproachisanimprovement,inthatitconsidershowmarketexchangemaybesupportedbyinstitutions(rules)thatpunishdefec-tion,suchasalegalsystem,orinformalcodesofconduct.Forsomepurposesthisisadequate.However,theserulesthemselvesultimatelyrequireenforcement.Foraconvincingaccountoftheinstitutionalfoundationsofmarkets,therefore,weneedtoconsidertheenforcementnotjustofthemarkettransactionsthemselves,butoftheruleswhichgovernthosetransactions.Sayingthatthereisanormagainstcheating,forexample,isinsufcient.ItiscriticaltostudyhowthenormissustainedaspartofInstitutions:RulesorEquilibria?29 aself-enforcingequilibriumoutcomeofagameinwhichtheenforcementofthenormresultsfromthebehavior,andexpectedbehavior,oftheplayers.Thesameistrueifenforcementiscarriedoutbyformalorinformalorganizations,suchasacourt,acreditbureau,oracommunity.Insituationswhereformalinstitutionsdonotfunctionwell,economicagentsmayrelyheavilyoninformalenforcement.McMillanandWoodruff(),forexample,foundthatinVietnam,rmsscrutinizeprospectivetradingpartnerscare-fully,andrelyoninformal“privateorder”institutions,includingbothbilateralenforcement(ceasingtodobusinesswitharmthatcheats)andmultilateralcomm-unityenforcement(sharinginformationaboutcheaters).However,thesemechan-ismsareapoorsubstituteforwell-functioningformalenforcementmechanisms.McMillanandWoodruffdonotinvestigatehowthebehaviorwhichsustainstheseprivateorderinstitutionsismadeself-enforcing:whydopeopleshareinformationandpunishcheaters,giventhatdoingsoiscostly?Kandori()usesagame-theoreticmodeltoshowthatamongacommunityofplayerswhoarerandomlymatchedintopairseachperiodtoplayaprisoner’sdilemma,amultilateralcommunityenforcementmechanismcansupportcoopera-tioniftheplayerscanobservealabelwhichindicates(roughlyspeaking)whethertheircurrenttradingpartneris“acheat”,andwhichishonestlyupdatedthroughsomeexogenousprocess.Severalpapersstudyhowthiskindofreputationalinfor-mationmightbesharedwithinacommunity.Gazzale()showsthatplayersmayhaveanincentivetogossipbecauseareputationforgossipingcandetertheirfuturetradingpartnersfromcheating.Greif()showshowinfor-mationtransmittedincorrespondenceamongacommercialandsocialnetworkofmedievaltraders(theMaghribitraders)supportedareputationmechanismthatsuccessfullydissuadedcheating.Merchantswhocheatedcouldexpectthattheiractionswouldbewidelyreportedwithinthenetwork.Sincemerchantswhowereostracizedfromthenetworkforcheatinghadnofurtherreputationtolose,theywouldbeexpectedto(rationally)cheatinanyfuturetransactions;andtherefore,eachmerchantinthenetworkwasmotivatedtopunishcheatersbytheexpectationthatotherswouldalsodoso,sothepunishmentwasself-enforcing.Notethatthisapproachdirectsattentionawayfromthecontentofthe“rules”aboutcheatingtothenetworksandinformationowsthatenabletheexpectationthatotherplayerswillpunishcheatstobesustainedaspartofaself-enforcingequi-librium.Whilethepresenceoftheseinformation-sharingstructures(networksandcommunities)themselvescanbetakenasexogenousintheshortrun,iftheinformalpunishmentmechanismssustainedbythecommunityaretosurviveasinstitutions,themaintenanceofsuchnetworksmustalsoultimatelybemadeendogenoustotheanalysis.Inmanycases,suchastheMaghribitradersstudiedbyGreif,thestructureofthenetworkisinlargepartanoutcomeofanhistoricalprocess.However,evennetworksthatarebequeathedbyhistoryneedtobemaintained.Onewaytostudytheoriginsandstabilityofsuchnetworksistoconsidergroupmembers’incentivestoretaintheirafliationexpost,bysubmittingtopunishmentratherthanattemptingtoevadeit(Greif).Anotherapproachistoconsiderthenetworksthemselvesasanoutcomeofapriorgameinwhichplayers(individuals,rms,orcountries,for30A.GreifandC.Kingston example)deliberatelyformlinkswithotherplayers(friendships,supplylinks,ormilitaryalliances,forexample)(Fafchamps;Jackson).Ingeneral,itisnotnecessarilythecasethatthenetworksthatemergethroughsuchaprocesswillbe“efcient”fromthepointofviewofsocietyoverall,becauseindividualschoos-ingtobuildlinksdonottakeintoaccounttheexternaleffectsofthoselinksonotherplayers.Greif()emphasizesthatwhenthegroupisanoutcomeofahistoricalprocess,theremaynotbeamechanismtocoordinateinclusionofnewmembers.Inefcientsizeisthelikelyoutcome.Evenifformalenforcementmechanismsareavailable,theymaynotbeemployedinequilibrium.Kranton()studiesamodelinwhichagentschoosebetweentradingwithinaninformalnetworkorinananonymousmarketinwhichagencyproblemsareabsent–forexample,becausethereisawell-functioningformalsystemtogovernmarketexchange.Thevalueofmarketinteractiondependsonthefractionofplayerswhochoosetobuyandsellwithinthemarket.Therefore,twoequilibriaarise:ifeveryonemakesuseoftheirinformalrelationshipstoobtaingoods,thenthemarketisthin,thesearchcostsofndingatradingpartnerinthemarketarehigh,andeachindividualhasanincentivetousehernetworkratherthanthemarkettoobtaingoods.However,ifinsteadmanypeoplechoosethemarket,thenndingatradingpartnerinthemarketbecomesrelativelyeasier,andtheinformalrelationshipsbreakdownasplayers’outsideoptionimproves.Therefore,eitherpatternofbehavior–amarketinwhichpeoplecarryoutephemeral,anony-moustransactions,orapatternoftradingwithinlong-lived,“trusting”informalrelationships–canemergeasaninstitutioncorrespondingtoanequilibriumofthegame.Ifthesocietybeginsinoneequilibrium,intheabsenceofexogenousshocks,orsomecoordinatingmechanismtoengineeracoordinatedshiftinbehavior,itmayremainstuckinthatequilibriumevenifthealternativeequilibriumwouldbemoreefcient.Thus,startingpointsmatter,andhistoryplaysaroleinequilibriumselection.Kranton’sanalysistakesthemarketasgiven,butthepresenceoflegalcontractenforcementmayitselfbetreatedasendogenous.Greif(19942006)hasarguedthatinmedievalEurope,theGenoesesocietycharacterizedbyindividualisticculturalbeliefsandinterest-basedcommunitiesexperiencedahigherdemandforlegalcon-tractenforcementthanthecollectivistculturalbeliefsandkin-basedcommunityoftheMaghribitraders.Asaresult,theGenoesedevelopedformalinstitutionsinclud-ingcodiedcontractlaws,double-entrybook-keeping,familyrms,billsoflading,andotherantecedentsofmodernbusinesspractices.5.2OrganizationsAsthesizeofacommunitygrows,relyingsolelyoninformalgovernancemaytendtobecomeproblematicasplayers’abilitytoobserveeachother’sactions,andtoshareinformationabouttransgressions,diminishes.Asaresult,thereisatendencytoevolvefrominformaltomoreformalmodesofgovernance.Institutions:RulesorEquilibria?31 Forexample,atthemedievalChampagneFairs,largenumbersofmerchantsfromalloverEuropecongregatedtotrade.Merchantsfromdifferentlocalitiesenteredintocontracts,includingcontractsforfuturedelivery,thatrequiredenfor-cementovertime(Verlinden;Milgrometal.).Therewasnostatetoenforcethesecontracts,andthelargenumberofmerchantsaswellastheirgeo-graphicdispersionmadeaninformalreputationmechanisminfeasible.Greif(arguesthatimpersonalexchangewassupportedbya“communityresponsibilitysys-tem”.Traderswerenotatomizedindividuals,butbelongedtopre-existingcommu-nitieswithdistinctidentitiesandstronginternalgovernancemechanisms.Althoughparticulartradersfromeachcommunitymayhavedealtwithmerchantsfromano-thercommunityonlyinfrequently,eachcommunitycontainedmanymerchants,sotherewasanongoingtradingrelationshipbetweenthecommunities,takenasawhole.Merchantsfromdifferentcommunitieswereabletotrusteachother,eveninone-shottransactions,byleveragingtheinter-community“trust”whichsustainedtheseinteractions.Ifamemberofonecommunitycheatedsomeonefromanothercommunity,thecommunityasawholewaspunishedforthetransgression,andthecommunitycouldthenuseitsowninternalenforcementinstitutionstopunishtheindividualwhohadcheated.Thissystemwasself-enforcing.Tradershadanincentivetolearnaboutthecommunityidentitiesoftheirtradingpartners,andtoestablishtheirownidentitiessothattheycouldbetrusted.Thecommunitieshadanincentivetoprotecttherightsofforeigntraders,andtopunishtheirmembersforcheatingoutsiders,soastosafe-guardthevaluableinter-communitytrade.Communitiesalsodevelopedformalinstitutionstosupplementtheinformalreputationmechanismandcoordinateexpec-tations.Forexample,eachcommunityestablishedorganizationsthatenabledmem-bersofothercommunitiestoverifytheidentityofitsmembers.Ultimately,thegrowthoftradethatthisinstitutionenabledcreatedtheimpetusforitseventualreplacementbymoreformalpublic-order(state-based)institutionswhichcoulddirectlypunishtradersby,forexample,jailingthemorseizingtheirproperty.However,publicorderneverentirelyreplacesprivateorder:marketsinmoderneconomiescontainamixofprivate-orderandpublic-orderinstitutions,andtrans-actionsmayrelyonboth(Greif).Forexample,alendermayobtainacreditreportonaprospectiveborrowerfromaprivatecredit-ratingrmthatlacksanyenforcementpowerbeyondtheabilitytoshareinformation,butitmayalsorelyonstateenforcementtoseizetheborrower’scollateralifshefailstorepay.Asnotedintheprevioussection,withintheinstitutions-as-rulestradition,itiswidelyheldthattheboundarybetweenrmsandmarketsasalternativemodesoforganizingtransactionswillbedeterminedsoastoachieveefciency(minimizetransactioncosts).Fromtheperspectiveoftheinstitutions-as-equilibriaapproach,organizationsappearascomponentsofbroaderequilibria,interactingwithotherinstitutionalelements,andconstrainedbythepast.Inparticular,iftherearemanypossibleequilibria,thentheremaybedifferentcongurationsoforganizationsasso-ciatedwitheachoftheseequilibria,andthestructureoforganizations,includingtheboundariesbetweenrmsandmarkets,cannotthereforebededucedfromaknowl-edgeofthecharacteristicsofthetransactionalone.32A.GreifandC.Kingston Forexample,duringtheeighteenthcenturytheinstitutionsgoverningmarineinsurancetransactionsdevelopedindifferentwaysindifferentcountries.InBritain,acoffeehouse(Lloyd’sofLondon)graduallydevelopedintoasophisticatedmarketplaceforunderwritingbyprivateindividualswho“under-wrote”theamo-untstheywerewillingtoinsureonpoliciespresentedtothembymerchantsorbrokers.Incontrast,inFrance,HollandandtheUS,privateunderwritingdisap-pearedandwasreplacedbyjoint-stockcorporations.Theunderlyingmarineinsur-ancetransactionswereplaguedbyseriousinformationasymmetriesandagencyproblems,includingthepotentialforvariouskindsoffraud.Eachinstitutionalform–privateorcorporateunderwriting–hadadvantagesanddisadvantagesindealingwiththeseagencyproblems.Kingston(2007)arguesthattheindustrywascharacterizedbymultipleequilibria.InBritain,anequilibriumbasedonprivateunderwritingbecameinsti-tutionalizedovertimethroughthedevelopmentofspecializedinstitutions,inparticularLloyd’scoffee-house,whichbecameahubforinformationaboutshipsandtheircrews,politicalandeconomicdevelopments,andthemanyotherfactorsaffectingtheriskofavoyage,andalsoforinformationaboutthereputationsofmarketparticipants.ThisultimatelymeantthatinBritain,thecorporationssuffereda“lemons”problembecauseoftheirinferioraccesstoinformationaboutvesselsandotherdevelopmentsaffectingtheriskofavoyage,enablingtheprivateunder-writerstodominatethemarket.Incontrast,intheAmericancolonies,althoughprivateunderwritinghadbeendevelopingrapidly,itneverreachedthelevelofcomplexityofLloyd’s.Instead,inthelate1790stherewasashiftbetweenequili-briaasprivateunderwritingwasextinguishedbycompetitionfromjoint-stockcorp-orations(Kingston).Kingston()showshowthetimingofaseriesofhistoricalevents,involvingbothexogenousshocks(suchaswar)andendoge-nousparameterchangesandlearningprocesses,drovetheprocessofinstitutionalchange(equilibriumselection),leadingultimatelytoapath-dependentbifurcationofinstitutionalstructurebetweenBritainandtherestoftheworld.Eachequilibrium,onceestablished,provedstable.Thus,althoughthefundamentalpurposeofthetransaction–sharingrisk–wasthesame,bytheendoftheNapoleonicwars,themannerinwhichthetransactionwasaccomplishedwasverydifferentindifferentcountries.The“governancestructures”thatemergedweretheoutcomeofahistor-icalprocesswithmultiplestableend-points,ratherthanbeingdesignedto“mini-mizetransactioncosts”.5.3PoliticsAsnotedintheprevioussection,afundamentalrationalefortheexistenceofthestateisthatitcanuseitscapacityforcoerciontoprovideorderandsecurity.Manycontemporarysocietiesfacethechallengeofbuildingstatesthateffectivelypro-motepoliticalstability,curtailpoliticalviolence,andfostereconomicprosperity.Thishasprovennoeasytask,despitethefactthatcopyingformalrules,includingInstitutions:RulesorEquilibria?33 constitutionsandlaws,isrelativelystraightforward.Whydon’tcountrieswithinefcientorineffectivepoliticalstructuressimplycopytheinstitutionalstructureofmoresuccessfulones?Andwhydocountriesthat“transplant”formalrulesfrequentlyndthatthisfailstoreproducethedesiredpatternsofbehavior?Withintheinstitutions-as-rulesframework,theexplanationofferedisthatbeha-viorisconstrainednotjustbyformalrules,butalsobythe“informalrules”presentinasociety.Butaswesawearlier,thisleavesunansweredthequestionofwheretheseinformalrulescomefrom,andhowthe“rules”,includingtherulesgoverningthebehaviorofthestateitself,areultimatelyenforced.Thekeypuzzleishowtoconstructastatethatisstrongenoughtoprovideorderandprotecttherightsofitscitizens,butinwhichpoliticalpower-holdersareneverthelessmotivatednottousethispowertoabusethoserights:inotherwords,allactors,including“rulers”,mustobey“therules”.Thus,astable,well-functioningpoliticalsystemshouldbeviewedasadesirableequilibriumoutcomeratherthanasasetofrules.Batesetal.()studyamodelinwhichplayerscanchoosetoallocatetheireffortamongthreegoals:production,leisure,andarmingthemselvestoengageinviolence.Acapacityforviolenceenablesplayersbothtodefendtheirownproduc-tiveoutputandto“raid”theoutputofothers.Inanarchy(asituationwithno“state”),therearetwokindsofequilibria:oneinwhichthereislittleviolence,butalsolittleproduction,sothatmosteffortgoesintoleisure;andanotherequilibriuminwhichthereisproduction,butalsoalotofviolence,aspeoplewhoproducemustalsobewillingtodefendtheiroutput.AsintheHobbesianvision,thecreationofastateasaspecialistinviolencecanimproveefciencybyenablingplayerstoproducewith-outfearofbeingraidedbyothers.Inreturnfortaxingtheoutputoftheplayers,thestateundertakestopunishraiding.Batesetal.(),however,gobeyondHobbesinprobingtheincentivesofthestateitself;theyshowthatthestatecanbecons-trainedfrompredationbytheshadowofthefuture,sinceafailuretoprotectthepropertyrightsofthecitizenscanleadtoreversiontoa“warlordequilibrium”inwhichnotaxesarepaid,andtheplayersarmthemselvesnotjustagainstraidingbyotherplayers,butagainststatepredationaswell.Thus,thebalanceofcoercivepowerscansustainthestateasanequilibrium.Weingast()showshowtheruleoflawemergedasanequilibriumoutcomeofagamebetweenarulerandhissubjectsinseventeenth-centuryEngland.Theking(James)initiallysupportedtheinterestsofthelandedToriesattheexpenseofthemercantileWhigs,wholackedthepowertooverthrowthisrulingcoalition. Ofcourse,real-worldprocessesofstate-buildingdonotstartfromthe“cleanslate”envisionedbyHobbes.Bates()arguesthathistorically,monarchicalstatesemergedoutofcompetitionamongfeudallineagesasrural,agrariansocietiesbasedonkinshipnetworksbecameincreasinglyurbanizedandindustrial.Olson()providesanalternativeparablefortheoriginsofthestate,arguingthatthestateemergedasthosewiththegreatestcapacityforviolencefounditprivatelymoreprotabletousethiscapacitytoprovideorderinexchangefortaxrevenue,ratherthansimplytolivebyplunder.34A.GreifandC.Kingston However,afterthekingbegantoinfringeontherightsoftheTories,theWhigsandToriescombinedtooverthrowhim,andinstalledasuccessor(William).Atthesametime,theycreatedanewconstitutionwiththeaimofpreventingthekingfromfuturepredation.Thenewconstitution,Weingastargues,wasfundamentallyacoor-dinationdevicethatlaidouttheconditionsthatwouldtriggeracoordinatedreac-tionbythecitizensagainstthekinginfuture.Thereby,itenabledashiftfromanequilibriuminwhichthekingwasabletotransgresstherightsoftheWhigswithimpunitytooneinwhichtheWhigsandToriesundertooktojointlyresistanytransgressionbythekingagainsteitheroftheirrights.Toachievethis,thenewagreementneededtobeself-enforcing.Bothgroupsofcitizenshadanincentivetoabidebytheagreement,asafailuretodosowouldenablethekingtoabusetherightsofbothgroupsinthefuture.Thekingwasmotivatedtorespectpropertyrightsbythecredibleexpectationthatbothgroupswouldreactinconcerttoaninfringementoftherightsofeithergroup.Thus,asinBatesetal.’smodel,whiletheplayersmayarticulate“rules”togoverntheirbehavior,itistheultimatelythethreatofabreakdownofcooperationinaninnitely-repeatedgamethatenablesanon-predatorystatetobesustainedasanequilibriumoutcome.Greif(2006,Chap.8)studiestheprocessofstate-buildinginmedievalGenoa.Genoa’scommercialexpansionhadbeenhinderedbythethreatofconictbetweentworivalfeudalclans,whichledeachclantowastesubstantialresourcesdefendingitselffromtheother.Toachievegainsfromcooperation,thewarringclansagreedtoinviteanon-Genoeseruler/administrator,thepodesta,torulethecity.Theheldthebalanceofpowerbetweentheclans,butwasnotmilitarilystrongenoughtoimposehiswillonthemandbecomeadictator.Toavoidthedangerofthealigninghimselfwitheitherclan,heandhisfamilywereforbiddenfrominvolve-mentinGenoesesocietyorpolitics.Indeed,thepodesta’spositiondependedonensuringthatneitherclanbecamedominant,ortheywouldhavehadnofurtherneedofhim.Thus,thepodesteriasystemwasasetofself-enforcinginstitutionsthatpromotedinter-clancooperationandreducedthethreatofconict.Theprocessofinstitutionalchangewasshapedbytheinitialconditions,includingthesetoforga-nizations(inthiscase,clans),andthefeudalrules,beliefs,andnormsinheritedfromthepast.ScartasciniandTommasi()studyamodelofpolicymakinginwhichindividualscaneitherpursuetheirinterestsviatheformalpoliticalprocessorthroughviolence,protests,bribery,andsoon.Theyshowthattheremaybemultipleequilibria:oneequilibriuminwhichallplayerschooseformalchannels,andanotherinwhichsomeplayersusetheformalprocess,butmanyplayers“gotothestreets”.Moreover,thestabilityoftheseequilibriaisreinforcedbyactors’investmentsovertime.Theirmodelcanaccountforthedifferencesinobservedpoliticalbehaviorbetweencountrieswithsimilar“formalrules”(suchastheU.S.andArgentina)butthesebehaviorsareexplainedasequilibriumoutcomesratherthanbyinvokingdifferencesinunobservable“informalrules”.Institutions:RulesorEquilibria?35 6Institutions-as-Equilibria:TheFrontier6.1Beliefs,Culture,andInstitutionalTrajectoriesIntheinstitutions-as-rulesperspective,beliefsinuencebehaviorthroughtheirimpactonthechoiceofrules.InNorth(2005)’sframework,economicagentshave“mentalmodels”whichreecttheirunderstandingoftheworldandwhichtheyusetoevaluatethedesirabilityofparticularrulechanges.Overtime,astheylearnabouttheworld,theyrevisetheirmentalmodels,andthismayaltertheirperceptionsaboutthenetgainsfromalternativepossiblerules,orleadthemtoperceivenewpossiblerules,leadingthemtotrytochangetherules.Thus,“thekeytounderstandingtheprocessofchangeistheintentionalityoftheplayersenactinginstitutionalchangeandtheircomprehensionoftheissues”(North2005,p.3).Theinstitution-as-equilibriaperspectiveemphasizesadditionalcausalrelationsbetweenbeliefsandoutcomes.Beliefsmotivatepeople’sbehaviorbyinuencingtheperceivedcostsandbenetsofvariousactions,includingexpectationsaboutothers’behavior.Beliefsmatterbecauseindividualshavepotentiallylimitedanddifferentinformation,knowledge,andcognitiveunderstandingabouttheenviron-mentandthestrategiesofotherplayers.Rulesprovideonemeansforpeopletocoordinatetheirbeliefs.Consider,forexample,theseeminglyunnecessarylawspecifyingthedirectionoftrafc(driveontheleft,ordriveontheright).Suchatrafclawprovidesnewdrivers(orthosevisitingfromabroad)withtheknowledgerequiredtomakeaninformeddecisionbasedonaminimalunderstandingofthesystem.Furthermore,becausesuchrulesspecifyself-enforcingbehavior,agentsaremotivatedtoacquireknowledgeoftherulesandfollowthem.Theanalysisoftheprocessesthroughwhichrulesaggregateknowledgeandinformationisinitsinfancy.AnotablecontributionisAoki()whoproposesthatasanexistingequilibriumbreaksdown,cognitivelylimitedagentsperceivethattheirformerstrategiesarenolongoptimal,withoutnecessarilyunderstandingwhy,andbegintoexperimentwithnewstrategies.Astheirbehaviorandexpectationschange,institutionalchange–amovementtoanewequilibrium–occurs.Eventu-ally,agents’strategiesandbeliefsystemsarebroughtbackintoalignmentwitheachotherasmutuallyconsistentcomponentsofanewinstitutionalequilibrium.Greif()proposesasocial,ratherthanindividualisticprocessoflearningandconvergence.Agentsrespondtotheexpectedbehaviorofothersasarticulatedinaknownruleofbehavior(eitherformalorinformal).Thetrafc-lawspecifyingaspeedlimit,forexample,constitutesasocialrule,knowntothedriversandtowhicheachofthemresponds.Theirresponsesleadthemtochooseaspeedhigherthanthelegalmaximum.Butthisisnotexplainedbyassertingthatthereisahidden“informalrule”specifyingtheobservedbehavior.Rather,asplayersobservebeha-viorandoutcomesovertime,theinstitutionalized‘rulesoftheroad’whichdevelopreectthedispersedbeliefsandinformationofindividualsrespondingtoa‘structure’whichissimultaneouslycreatedbytheiraggregatedresponsestothestructureitself.36A.GreifandC.Kingston Theinstitutions-as-rulesapproachmakesacleardistinctionbetweenformalrules,whicharecreatedinparticularbythestate,andculture,whichconsistsofinformalrulesformulatedbysociety.Incontrast,thedistinctionbetween‘rules’andcultureintheinstitutions-as-equilibriaapproachisoneofkindandnotoneofessence.Bothrulesandcultureinuencebehaviorbygivingrisetosharedbeliefs,norms,andexpectationsthatgenerateregularitiesofbehavior.Thisfacilitatesstudyingtheinter-relationsbetweeninstitutionsandculture.Clearly,neitherculturenorinstitutionsareimmutable.Therearemanyhistoricalexamplesofrapidculturalchange.Yet,totheextentthatoneassociatesculturewithinstitutionalelementsthatprevailedpriortostateformationoremergeindepen-dentlyofit,distinctculturescanleadsocietiesalongdistincttrajectoriesofinstitu-tionaldevelopment.Theperpetuationandimplicationsofbothinstitutionsandculturedependonthecontext,unintendedconsequences,andhistoricalcontingen-ciessuchasthesequenceofvariousexogenousevents,leadership,andtheoutcomesofmilitaryconicts.Culturalchangeandculture’simpactarenotdeterministic,butaspecicculturecanrendersomeinstitutionaltrajectoriesmorelikelythanothers.BenabouandTirole()giveanexampleofhowmultipleequilibriacanresultfromthetwo-wayinteractionbetweensharedbeliefsandpublicpolicies.Theyarguethatinsocietieswheremanypeopleholdabeliefina“justworld”–thebeliefthateconomicsuccessishighlydependentoneffort–thesepeoplewillfavorlowlevelsofredistributionandlowtaxrates.Thesepoliciesincreasetherewardtoeconomiceffort,givingpeopleanincentivetoadopt(andteachtheirchildren)the“justworld”ideology.If,instead,peoplebelievethatluckplaysmoreofaroleindeterminingindividualsuccess,theymayfavorhigherlevelsofredistribution,whichdampenstheincentivesforhigheffort,conrmingthebasesfortheirbeliefs.Notethattheseideologicalbeliefsaremorethanjustareectionofdifferentinstitutionalstructures.Theyareafundamentalpartofeachequilibrium.Greif()showedhowdistinctculturalbeliefsledtodistinctdevelop-mentsofcontractenforcementinstitutionsamongeleventhcenturyJewishmerch-antsoperatingintheMuslimworldandtheLatin-ChristianGenoese.Collectivismamongtheformerfosteredrelianceonenforcementbasedonamultilateralreputa-tionmechanismwhileindividualismamongthelatterfosteredenforcementbasedonbilateralreputationandthelaw.Thelatter’srelianceonthelaw,inturn,wasfacilitatedbythefactthatitwasaman-madelawandnotadivinelaw.ThisfundamentaldistinctioninlegalconceptionswasnotinstitutedbystatesorrulersbutreectedthedistincthistoricalprocessesthroughwhichChristianityandIslamemerged.6.2MoralNormsandEndogenousPreferencesAsecondfrontierissueintheinstitutions-as-equilibriaapproachistheinter-relationsamonginstitutionsandpreferences.Althoughsomeaspectsofindividualpreferences,suchasthosedirectlyrelatedtosurvival,areprimordialandselsh,Institutions:RulesorEquilibria?37 otheraspectsofpreferencesareshapedbysocietybecausehumanshaveother-regardingpreferencesandseekmoraljusticationfortheirbehavior.Thecrucialelementhereisinternalizedmoralnormsorvaluesthatindividualsarepsychologi-callymotivatedtofollow.Aninternalizednormagainststealing,forexample,placesawedgebetweenthenetutilityvalueofvedollarsearnedandvedollarsstolen.Suchmoralnormsbasedonintrinsicmotivationaredifferentfrom‘socialnorms’whichrelyonextrinsicmotivationprovidedbythethreatofnon-legalpunishments.Moralnormsinuencebehaviordirectlythroughtheirimpactonpreferences,andindirectlybyinuencingtheexpectedbehaviorofthosewhoareperceivedtohaveinternalizedsuchnorms.Peoplearebornwiththecapacityandthepropensitytointernalizenorms,andabsorbnormsthroughsocializationbyrolemodels,parents,peersandorganizations(suchasschoolsandchurches).Institutionscaninuencenormsthroughtheirimpactonthesesocializingagents.Tabellini()providesawonderfulanalysisofhownormsofgeneralizedorlimitedmoralitycanevolveinthesamesituationdepend-ingontheincentivesinstitutionsprovidetoparents.Specically,aparentfacesatrade-offbetweensocializingherchildtohavehernormsorsocializingthechildtohavethenormswhichwouldbeoptimalforthechildinthefuture.Institutionsinuencethistrade-off.AkerlofandKranton()discussthevalueofintrinsicmotivationwithinorganizations.Theyusethemilitaryasacompellingexampleofanorganiza-tionwhosemembersareprimarilymotivatedbynon-monetaryincentives(suchashonor).Suchatheoryofmotivationhasimportantimplicationsfororganiza-tionaldesign.Forexample,employingasupervisortomonitoraworker’seffortmayenablethermtomotivatetheworkerusinghigh-poweredmonetaryincen-tives,butKrantonandAkerlofarguethatthereisalsoahiddencost:hiringthesupervisormayalsoreducetheemployee’ssenseofidenticationwiththermanditsgoals,therebyerodingtherm’s“motivationalcapital”(worker’sloyalty).AkerlofandKranton’sworkisrelatedtoanemergingliteratureon“endogenouspreferences”,muchofwhichusesevolutionaryargumentstoinvestigatetheroleinstitutionsplayinmoldingnotjustpeople’sbehavior,butalsotheirgoals(Bowles1998).Thesetheoriesemphasizethatwhileinstitutions,beingman-made,arecreatedthroughhumanaction(whetherintentionalornot),institutionsalsoplayaroleinreconstitutingthegoalsandperceptionsoftheindividualstheygovern.Theinte-grationoftheseconsiderationsintheinstitutions-as-equilibriaperspectiveisinitsinitialstages.Ifsuccessful,itwillimproveourabilitytostudynormsasoneelementinalargersysteminwhichpeoplearemoral,yetmaterialistic,andmotivationisprovidedbyendogenousbeliefs,norms,andexpectations.6.3Origin,DynamicsandComplementaritiesTheinstitutions-as-rulesapproach,aswehaveseen,studiesinstitutionsas(exoge-nous)constraints(rules)leadingto(endogenous)behavior,whileenforcementoftherulesistreatedasaseparateissue.Institutionaldynamicsisfundamentallyabout38A.GreifandC.Kingston changingrules,andtheanalyticalfocusisonchangingformalrules.Incontrast,theinstitutions-as-equilibriaapproachfocusesonthebehavioralmanifestationsofendogenousmotivation;how(endogenous)behaviorgenerates(endogenous)insti-tutionsthatperpetuatethisbehavior.Institutionaldynamicsisthereforefundamen-tallyaboutchangesinmotivationandregularitiesofbehavior,andtheanalyticalfocusisonchangesinbeliefs,norms,andexpectations(GreifandLaitinTwocausesofinstitutionalchangeareparticularlyimportant.Therstisanintentionalattempttobringaboutchangebythosewhorealize(orhope)thattheycanbenetfromit.Thiskindofinstitutionalchangecanresultfromtheperceptionofnewinstitutionalpossibilities,perhapsbroughtaboutbylearningorbynewinteractionswithoutsiders.Becausetheexistinginstitutionsareequilibria,how-ever,theygenerallycannotbechangedunilaterallybyasingleactor.Bringingaboutachangemaythereforeinvolveovercomingcollectiveactionproblemsaswellasovercomingtheoppositionofthosewhostandtolosefromthechange.Suchcollectiveactionmayoccurthroughpersuasionorthroughtheuseofneworexistingorganizations,or,lesscommonly,throughtheriseofacharismaticleader.Thesecondmaincauseofinstitutionalchangeis“institutionaldisequilibrium”whichresultswhenaninstitutionceasestobeself-enforcing.Thiscanoccureitherduetoexogenousshocksorduetoendogenouschangesin“quasi-parameters”(GreifandLaitin):variableswhichchangegraduallyovertimeasaresultoftheoperationoftheinstitutionitself,andmayultimatelypassacriticalthresholdsothattheinstitutionceasestobeself-enforcing.Whetherthesubsequentinstitutionalchangeisgradualorabrupt,evolutionaryorintentionaldepends,inparticular,onwhethertheactorsarecognitivelyawareoftheprocessleadingtochange,whoisawareofit,andhowtheycaninstitutionallyrespond.Thedetailsoftheresultingnewinstitutions,iftheyleadtotheintendedout-comes,arepartiallydictatedbythefunctiontheyhavetoserve.Yet,therearemanyinstitutionsthatcanachievethesameobjectives.Fromtheinstitutions-as-equilibriaperspective,pastinstitutionalelementsinuencethedetailsofsubsequentinsti-tutionsbecauseinstitutionalizedbeliefs,norms,andexpectationsareembodiedinpeople’sbeliefssystems,preferencesandmemories,whileexistingorganizationshaveenduringphysicalcapacities,routines,andotherresources.Althoughitmaybetechnologicallypossibletocreatenewbeliefs,norms,expectationsandorgani-zations,doingsoisusuallycostly,timeconsuming,andrequiresventuringintothecognitivelyunknown(GreifThereisthereforeafundamentalasymmetrybetweeninstitutionalelementsinheritedfromthepastandtechnologicallypossiblealternatives.Pastinstitutionalelementsaretherawmaterialonwhichnewinstitutionsarebased.Sugden(arguesthatpeoplewishingtocoordinatetheirstrategieswillgenerallyadoptruleswhichareanalogoustoruleswithwhichtheyarealreadyfamiliar,forexample,the“rstcome,rstserved”rulewhichisusedtoassignpropertyrightsinmanysituations(andtherebyavoidpotentiallycostlyconict).Greif()arguedthatorganizationsinheritedfromthepastandculturalbeliefsdetermineselectionamongalternativeinstitutions.Patternsoforganizationalmembershipdeterminewhoseidentityisknowntowhom,andwhereinformationows,whileculturalbeliefsInstitutions:RulesorEquilibria?39 coordinateexpectations.Campbell()arguesthatactorsoftencreatenewinstitutionsthroughaprocessof“bricolage”:recombiningelementsintheirinstitu-tionalrepertoiretodealwithnewsituations.Newinstitutionsoftenresembleolder,familiarinstitutionsbecausetheycontainelementsinheritedfromorinspiredbypastinstitutions.Greif()delineateshowexactlypastinstitutionalelementsinuencesubsequentinstitutionsthroughtheirenvironmental,coordination,andinclusioneffects.Thus,theinstitutions-as-equilibriaapproachconceptualizesinstitutionaldynam-icsasanaccumulativehistoricalprocessofinter-relatedinstitutionalelements.Pastinstitutionalelementsareincorporatedintonewinstitutionsthatemergewithinthecontextof–andhencearecomplementaryto–existinginstitutions.Theresultsareinstitutionalcomplexes,whichareasetofinstitutionsthatgovernvariousinterac-tions,havecommoninstitutionalelements,andarecomplementarytoeachother.Society’sinstitutionshavetobestudiedfromaholistic,systemicperspective(Aoki7ConcludingNotesRecentscholarshiphasdemonstratedthepoweroftherationalchoiceframeworkforadvancingourunderstandingofinstitutionsandinstitutionalchange.Andasourunderstandingimproves,theconceptualframeworksemployedtostudyinstitutionscontinuetoevolve,enablingustodevelopricherandmorecompleteanswersevenasweprobedeeperandmorecomplexquestionsaboutthenatureofinstitutionsandprocessesofinstitutionalemergenceandchange.Fromtheinstitutions-as-rulesperspective,institutionsarerulesthatareeitheroptimalresponsestotheinstitutionalenvironmentoraredeterminedbytheinterestsofthepoliticalactorswhomaketherules.Thisapproachhasbeenfruitfullyapp-liedtoexplorehowthe“rulesofthegame”areformedindiversesettings.However,becauseenforcementoftherulesistreatedasexogenous,theinstitutions-as-rulesapproachworksbestinsituationswherethereare,infact,well-functioningandtransparentenforcementinstitutionswhichcanbetakenasgiven,andinwhichtherules(whetherformalorinformal)areeasytoobserve,sothattheymaybeexpectedtotranslatemoreorlessdirectlyintoeffectsonhumanbehavior.Iftheseconditionsarenotmet,thenultimately,toexplainhowtherulesareenforced(ornot),andwhytheyarefollowed(ornot),theinstitutions-as-rulesframeworkmustbesupplementedorreplacedbyatheoryinwhichenforcementistreatedasendogenous,andtheincentivesofallplayerstofollowtherulesareexplainedratherthanassumed.Theinstitutions-as-equilibriaapproachfocusesonmotivationprovidedbybeliefs,norms,andexpectationsthatbothshapeindividualbehaviorwhilstsimultaneouslythemselvesbeingaproductofthestrategicinter-playbetweenagents(individualsororganizations).Thus,boththecontentoftherules(behavior)andtheirenforcement(people’smotivationforfollowingthem)canbestudiedwithinauniedframework.Thekeytoinstitutionalchange,from40A.GreifandC.Kingston thisperspective,isnotjustchangingrules,butchangingplayersmotivationsandpatternsofbehaviorinaself-enforcingway.Thesetwoapproachescanbeseenascomplementarypartsoftheanalyststoolkit.Theinstitutions-as-rulesapproachseemsappropriateforstudyingthedevelopmentofinstitutionswithinanestablishedstructurethatcanenforcetherules,forexample,inastabledemocracywithinwhichbasicmarket-supportinginstitutionsarealreadywell-established.Theinstitutions-as-equilibriaapproachmightbebettersuitedtostudyingtheinstitutionalfoundationsofmarketsanddemocraticpoliticalstructures,andothersituationsinwhichenforcementofthe“rules”mustbeconsideredasanendogenousoutcomeratherthantakenasgiven.AcemogluD(2003)WhynotapoliticalCoasetheorem?Socialconict,commitmentandpolitics.JCompEcon31:620–652AcemogluD,RobinsonJ(2006)Economicoriginsofdictatorshipanddemocracy.CambridgeUniversityPress,CambridgeAcemogluD,JohnsonS,RobinsonJ(2001)Thecolonialoriginsofcomparativedevelopment:anempiricalinvestigation.AmEconRev91:1369–1401AcemogluD,JohnsonS,RobinsonJ(2002)Reversaloffortune:geographyandinstitutionsinthemakingofthemodernworldincomedistribution.QuartJEcon117:1231–1294AkerlofG,KrantonR(2005)Identityandtheeconomicsoforganizations.JEconPerspect19(1):AlchianA(1950)Uncertainty,evolutionandeconomictheory.JPolitEcon58:211–221AldashevG,ChaaraI,PlatteauJ-P,WahhajZ(2007).Customintheshadowoftheformallaw:aneconomicanalysis.Workingpaper,UniversityofNamurAokiM(2001)Towardsacomparativeinstitutionalanalysis.MITPress,Cambridge,MAAokiM(2007)Endogenizinginstitutionsandinstitutionalchanges.JInstEcon3(1):1–31BaronD(2001)PrivateorderingontheInternet:theeBaycommunityoftraders,StanfordGSBresearchpaperno.1709BarzelY(2002)Atheoryofthestate:economicrights,legalrights,andthescopeofthestate.CambridgeUniversityPress,CambridgeBatesR(2001)Prosperityandviolence:thepoliticaleconomyofdevelopment.Norton,NewYorkBatesR,GreifA,SinghS(2002)Organizingviolence.JConRes46(5):599–628BenabouR,TiroleJ(2006)Beliefinajustworldandredistributivepolitics.QuartJEcon699–746BowlesS(1998)Endogenouspreferences:theculturalconsequencesofmarketsandotherecono-micinstitutions.JEconLit36:75–111BowlesS,GintisH(1993)Therevengeofhomoeconomicus:contestedexchangeandtherevivalofpoliticaleconomy.JEconPerspect7(1):83–102BrousseauE,RaynaudE(2008)Climbingthehierarchicalladdersofrules:thedynamicsofinstitutionalframeworks.Workingpaper,UniversityofParisXCalvertR(1995)Rationalactors,equilibrium,andsocialinstitutions.In:KnightJ,SenedI(eds)Explainingsocialinstitutions.UniversityofMichiganPress,AnnArbor,MICampbellJ(2004)Institutionalchangeandglobalization.PrincetonUniversityPress,Princeton,NJCoaseR(1937)Thenatureoftherm.Economica4:386–405DemsetzH(1967)Towardatheoryofpropertyrights.AmEconRev57(2):347–359DixitA(2004)Lawlessnessandeconomics:alternativemodesofgovernance.PrincetonUniversityPress,Princeton,NJInstitutions:RulesorEquilibria?41 DjankovS,GlaeserE,LaPortaR,Lopez-de-SilanesF,ShleiferA(2003)Thenewcomparativeeconomics.JCompEcon31:595–619EasterlyW,LevineR(2003)Tropics,germsandcrops:howendowmentsinuenceeconomicdevelopment.JMonetaryEcon50:3–39EllicksonR(1991)Orderwithoutlaw.HarvardUniversityPress,Cambridge,MAEngermanS,SokoloffK(2002)Factorendowments,inequality,andpathsofdevelopmentamongnewworldeconomies.Economia3:41–109FafchampsM(2004)MarketinstitutionsinSub-SaharanAfrica.MITPress,Cambridge,MAFearonJ(2007)Fightingratherthanbargaining.Workingpaper,StanfordUniversityGazzaleR(2005)Givinggossipstheirdue:informationprovisioningameswithprivatemonitor-ing.Workingpaper,WilliamsCollegeGreifA(1989)Reputationandcoalitionsinmedievaltrade:evidenceontheMaghribitraders.JEconHist49(4):857–882GreifA(1993)Contractenforceabilityandeconomicinstitutionsinearlytrade:theMaghribitraders’coalition.AmEconRev83(3):525–548GreifA(1994)Culturalbeliefsandtheorganizationofsociety:ahistoricalandtheoreticalreectiononcollectivistandindividualistsocieties.JPolitEcon102(5):912–950GreifA(1998)Historicalandcomparativeinstitutionalanalysis.AmEconRev88(2):80–84GreifA(2000)Thefundamentalproblemofexchange:aresearchagendainhistoricalinstitutionalanalysis.EurRevEconHist4(3):251–284GreifA(2006)Institutionsandthepathtothemoderneconomy:lessonsfrommedievaltrade.CambridgeUniversityPress,CambridgeGreifA,LaitinD(2004)Atheoryofendogenousinstitutionalchange.AmPolitSciRev98(4)HadeldG(2008)Theleversofinstitutionaldesign:institutionaldeterminantsofthequalityoflaw.JCompEcon36(1):43–73HartO(1995)Firms,contractsandnancialstructure.OxfordUniversityPress,OxfordHayekF(1945)Theuseofknowledgeinsociety.AmEconRev35(4):519–530HayekF(1973)Law,legislationandliberty,vol1:Rulesandorder.UniversityofChicagoPress,JacksonM(2006)Theeconomicsofsocialnetworks.In:BlundellR,NeweyW,PerssonT(eds)VolumeIofadvancesineconomicsandeconometrics,theoryandapplications:ninthworldcongressoftheeconometricsociety.CambridgeUniversityPress,CambridgeKandoriM(1992)Socialnormsandcommunityenforcement.RevEconStud59:63–80KeeferP,KnackS(1997)Doessocialcapitalhaveaneconomicpayoff?Across-countryinvesti-gation.QuartJEcon112(4):1251–1288KingstonC(2007)MarineinsuranceinBritainandAmerica,1720–1844:acomparativeinstitu-tionalanalysis.JEconHist67(2):379–409KingstonC(2008)Adverseselectionandinstitutionalchangeineighteenthcenturymarineinsurance.Workingpaper,AmherstCollegeKingstonC(2011)MarineinsuranceinPhiladelphiaduringthequasi-warwithFrance.JEconHistKingstonC,CaballeroG(2009)Comparingtheoriesofinstitutionalchange.JInstEcon5(2):KrantonR(1996)Reciprocalexchange:aself-sustainingsystem.AmEconRev86(4):830–851LaPortaR,Lopez-de-SilanesF,ShleiferA(2008)Theeconomicconsequencesoflegalorigins.JEconLit46(2):285–332LibecapG(1989)Contractingforpropertyrights.CambridgeUniversityPress,CambridgeMcMillanJ,WoodruffC(1999)DisputepreventionwithoutcourtsinVietnam.JLawEconOrg15MilgromP,NorthD,WeingastB(1990)Theroleofinstitutionsintherevivaloftrade:thelawmerchant,privatejudges,andthechampagnefairs.EconPolit1:1–23NelsonR,WinterS(1982)Anevolutionarytheoryofeconomicchange.HarvardUniversityPress,Cambridge,MA42A.GreifandC.Kingston NorthD(1990)Institutions,institutionalchangeandeconomicperformance.CambridgeUniversityPress,CambridgeNorthD(2005)Understandingtheprocessofeconomicchange.PrincetonUniversityPress,Princeton,NJOlsonM(1982)Theriseanddeclineofnations.YaleUniversityPress,NewHaven,CTOlsonM(1993)Dictatorship,democracy,anddevelopment.AmPolitSciRev87(3):567–576OstromE(1990)Governingthecommons:theevolutionofinstitutionsforcollectiveaction.CambridgeUniversityPress,CambridgeOstromE(2005)Understandinginstitutionaldiversity.PrincetonUniversityPress,Princeton,NJPiersonP(2000)Increasingreturns,pathdependence,andthestudyofpolitics.AmPolitSciRevQianY(2003)HowreformworkedinChina.In:RodrikD(ed)Insearchofprosperity:analyticnarrativesoneconomicgrowth.PrincetonUniversityPress,Princeton,NJRolandG(2000)Transitionandeconomics:politics,marketsandrms.MITPress,Cambridge,MARolandG(2004)Understandinginstitutionalchange:fast-movingandslow-movinginstitutions.StudCompIntDev38(4):109–131ScartasciniC,TommasiM(2009).Themakingofeconomicpolicy:institutionalizedornot?IDBworkingpaperno.108SchotterA(1981)Theeconomictheoryofsocialinstitutions.CambridgeUniversityPress,SkaperdasS(2006)Anarchy.In:WeingastB,WittmanD(eds)Oxfordhandbookofpoliticaleconomy.OxfordUniversityPress,OxfordSugdenR(1989)Spontaneousorder.JEconPerspect3(4):85–97TabelliniG(2008)Thescopeofcooperation:valuesandincentives.QuartJEcon123(3):905–950VerlindenC(1979)Marketsandfairs.In:PostanM,RickE,MilteyM(eds)TheCambridgeeconomichistoryofEurope,vol3.CambridgeUniversityPress,Cambridge,pp119–153WeingastB(1997)Thepoliticalfoundationsofdemocracyandtheruleoflaw.AmPolitSciRevWilliamsonO(1979)Transactioncosteconomics:thegovernanceofcontractualrelations.JLawEcon22:233–261WilliamsonO(2000)Thenewinstitutionaleconomics:takingstock,lookingahead.JEconLitInstitutions:RulesorEquilibria?43