Hart-and-Holmstrom(1987)-The-Theory-of-Contracts

--CHAPTER3ThetheoryofcontractsOliverHartandBengtHolmstromlntroductionThepastdecadehaswitnessedagrowinginterestincontracttheoriesofvariouskinds.Thisdevelopmentispartlyareactiontoourratherthor-oughunderstandingofthestandardtheoryofperfectcompetitionundercompletemarkets,butmoreimportantlytotheresultingrealizationthatthisparadigmisinsufficienttoaccommodateanumberofimportanteco-nomicphenomena.Studyinginmoredetailtheprocessofcontracting-particularlyitshazardsandimperfections-isanaturalwaytoenrichandamendtheidealizedcompetitivemodelinanattempttofittheevidencebetter.Atpresentitisthemajoralternativetomodelsofimperfectcom-petition;wewillcommentonitscomparativeadvantagebelow.Inonesense,contractsprovidethefoundationforalargepartofeco-nomicanalysis.Anytrade-asaquidproquo-mustbemediatedbysomeformofcontract,whetheritbeexplicitorimplicit.Inthecaseofspottrades,however,wherethetwosidesofthetransactionoccuralmostsimultaneously,thecontractualelementisusuallydownplayed,presum-ablybecauseitisregardedastrivial(althoughthisneednotbethecase;seeSection3).Inrecentyears,economistshavebecomemuchmoreinter-estedinlong-termrelationshipswhereaconsiderableamountoftimemayelapsebetweenthequidandthequo.Inthesecircumstances,acontractbecomesanessentialpartofthetradingrelationship.Ofcourse,long-termcontractsarenotnewineconomics.ContingentcommoditytradesoftheArrow-Debreutypeareexamplesparexcellenceofsuchcontracts.WhatdoesseemnewistheanalysisofcontractswrittenWewo~lldlike10thankJonathanFeinstein,PaulJoskow,JohnMoore,SherwinRosen,JeanTirole,andAndyWeissforcommentsonanearlierdraft.FinancialsupportfromNSFandtheSloanFoundationisgratefullyacknowedged.72OliverHartandBengtHolmstrombyandcoveringasmallnumberofpeople.Thatis,therehasbeenamoveawayfromtheimpersonalArrow-Debreumarketsettingwherepeoplemaketradeswiththemarket,toasituationwherefirmAandfirmB,orfirmCandunionD,writealong-termcontract.Thisdepartureisnotwith-outeconomicsignificance.Williamson(1985),inparticular,hasstressedtheimportanceofsituationswhereasmallnumberofpartiesmakeinvest-mentswhicharctosomeextentrelationship-specific;thatis,oncemade,theyhaveamuchhighervalueinsidetherelationshipthanoutside.Giventhislock-ineffect,eachpartywillhavesomemonopolypowerexpost,althoughtheremaybeplentyofcompetitionexantebeforeinvestmentsaresunk.Sincethepartiescannotrelyonthemarketoncetheirrelation-shipisunderway,theobviouswayforthemtoregulate(anddividethegainsfrom)tradeisviaalong-lermcontract.Untiltheadventofcontracttheory,economistsdidnothavethetoolstoanalyzeexantecompetitive,expostnoncompetitiverelationshipsofthistypeviaformalmodels.Researchoncontractshasprogressedalongseveraldifferentlines,eachwithitsownparticularinterests.Itmaybeusefultobeginbymention-ingsomeofthesedirectionsbeforeoutliningthesubjectswewillconcen-trateoninthischapter.Onestrandoftheliteraturehasfocusedontheinternalorganizationofthefirm,viewingthefirmitselfasaresponsetofailuresinthepricesys-tem.Questionsofinterestincludestructuringincentivesformembersofthetirm,allocatingdecisionauthority,andchoosingdecisionrulestobeimplementedbysuitablerewardstructures.Ofcourse,theobjectiveispartlytogaininsightintoorganizationtheoryassuch.Butperhapsmoreimportantly,oneisinterestedinknowingwhetherorganizationtheorymattersintheaggregate-thatis,towhatextenttheconductoffirmswillbedifferentfromtheassumedprofit-maximizingbehavior;and,ifitdif-fers,whatramificationsfollowformarketoutcomesandoverallalloca-tionsintheeconomy.Anotherprominentlineofresearchhasexploredtheworkingsofthelabormarket.Aplausiblehypothesisisthatcontingentclaimsforlaborservicesarelimitedforreasonsofopportunism.Thisinvitesinnovationofothertypesofcontractsthatcanbeusedassubstitutes.Theresearchhascenteredonthestructureofoptimalbilaterallaborcontracts(undervariousassumptionsaboutenforcementopportunities),onthepropertiescontractualequilibriawillhave,andinparticularonwhethertheseequi-libriawillcxhibitthecommonlyclaimedinefficienciesassociatedwithreal-worldadjustmentsinemployment.lnspiredbythepossibilitythatlong-termcontractsmayembodypriceandwagesluggishness,arelatedbodyofworkhasexploredtheirmacro-Theoryofcontracts73economicimplications[see,e.g.,Fischer(1977)andTaylor(1980)l.Un-likemostcontractanalysisthisliteraturehastakentheformofcontractsasgiven,typicallywithnominalwageandpricerigidities.Thisisnotassatisfactoryasworkingfromfirstprinciples,butithasmadepolicyanal-ysisquitetractable.Financialmarketsofferanotherarenaofsubstantialpotentialforcon-tracttheoreticstudiesthatisbeginningtoberecognized.Theimportanceoflimitedcontractingfortheemergenceoffinancialservicesandinstitu-tionshasbeensuggestedbyD.Diamond(1984),GaleandHellwig(1985),andTownsend(1980).Thislineofresearchalsooffersprospectsforacarefulmodelingoftheroleofmoneyandtheconductofmonetarypol-icy[seeTownsend,Chapter11thisvolume,andD.Diamond(1985)l.Asthefieldisprogressing,itbecomeshardertoplacemodelsinspe-cificcategories.Initially,modelsoforganizationaldesignignoredmarketforces,oratleasttreatedtheminaveryprimitivefashion.Incontrast,thetheoryoflaborcontractsstart