BoardsofDirectorsasanEndogenouslyDeterminedInstitution:ASurveyoftheEconomicLiteratureBenjaminE.HermalinUniversityofCaliforniaatBerkeleyandMichaelS.WeisbachUniversityofIllinoisJune15,2000AbstractThispapersurveystheeconomicliteratureonboardsofdirectors.Althoughalegalrequirementformanyorganizations,boardsarealsoanendogenouslydeterminedgovernancemechanismforaddressingagencyproblemsinherenttomanyorganizations.Formaltheoryonboardsofdirectorshasbeenquitelimitedtothispoint.Mostempiricalworkonboardshasbeenaimedatansweringoneofthreequestions:1)Howdoboardcharacteristicssuchascompositionorsizerelatedtoprofitability?2)Howdoboardcharacteristicsaffecttheobservableactionsoftheboard?3)Whatfactorsaffectthemakeupofboardsandhowtheyevolveovertime?Theprimaryfindingsfromtheempiricalliteratureonboardsare:Boardcompositionisnotrelatedtocorporateperformance,whileboardsizeisnegativelyrelatedtocorporateperformance.Bothboardcompositionandsizearecorrelatedwiththequalityoftheboard’sdecisionsregardingCEOreplacement,acquisitions,poisonpills,andexecutivecompensation.Finally,boardsappeartoevolveovertimeasafunctionofthebargainingpoweroftheCEOrelativetotheexistingdirectors.Firmperformance,CEOturnover,andchangesinownershipstructureappeartobeimportantfactorsaffectingchangestoboards.TheNSF(GrantSBR-9616675)andtheWillisH.BoothChairinBanking&Financeprovidedfinancialsupport.WethankKevinHallockandAnilShivdasaniforhelpfulcommentsonanearlierdraft,andHamidMehranforencouragingustowritethispaper.Theauthorscanbereachedathermalin@haas.berkeley.eduandweisbach@uiuc.edu.11.IntroductionMostlargeandevenmanysmallorganizationsaregovernedbyaboardofdirectors.Havingaboardofdirectorsisoneofthelegalrequirementsforincorporation,andmanynon-incorporatedentitieshaveagoverningboardofsomesort(e.g.,auniversity’sboardofregents).Whydoboardsexist?Whatdotheydo?Cantheybe“improved”?Thesequestionsgetattheheartofgovernance,andtoacertainextent,management.Assuch,theyhavebecomeasourceofmuchresearch.Thispapersurveystheresearchonboardsofdirectorsintheeconomicsandfinanceliteratures.Boardsofdirectorsareaneconomicinstitutionthathelpssolvetheagencyproblemsinherentinmanaginganyorganization.Althoughtheysatisfynumerousregulatoryrequirements,theyexistprimarilybecauseoftheconflictofintereststheyhelptoaddress.Yet,formaleconomictheoryontheboardhasbeenquitelimited.Forexample,thecharacteristicsofagencyproblemsleadingtoboardsastheequilibriumsolutionhasnotyetbeenspecified.Norhavetheconditionsunderwhichregulationsofboardswillleadtoimprovements.Despitetheabsenceofformaltheory,wehaveastrongsenseoftheunderlyingtensionssurroundingboards.ThemajorconflictofinterestwithintheboardroomisbetweentheCEOandthedirectors.TheCEOhasincentivesto“capture”theboard,soastoensurethathecankeephisjobandincreasehisflowofrents.DirectorshaveatleastsomeincentivestomonitortheCEOandtoreplacehimifhisperformanceispoor.Wepositthattheboardevolvesovertimedependingthenatureofthebargainingpositionofeachsideinthisconflict.Tosomeextent,thevacuumintheoryhasbeenfilledbyempiricalworkonboards.The“cost,”ofthisapproach,however,isthatlittleoftheempiricalworkonboardshasbeenmotivatedbyformaltheory.Rather,ithassoughttoansweroneofthreequestions:21)Howdoboardcharacteristicssuchascompositionorsizeaffectprofitability?2)Howdoboardcharacteristicsaffecttheobservableactionsoftheboard?3)Whatfactorsaffectthemakeupofboardsandhowtheyevolveovertime?Akeyissueinthisempiricalworkishowtoproxyfortheboard’sdegreeofindependencefromtheCEO.Muchofthisworkstartsfromtheoftenimplicitassumptionthatobservableboardcharacteristics,suchassizeorcomposition,arerelatedthislevelofindependence.Anumberofempiricalresultshavebeendocumentedfairlyconsistently.First,boardcomposition,asmeasuredbytheinsider/outsiderratio,1isnotcorrelatedwithfirmperformance.However,boardsizeisnegativelyrelatedtoafirm’sfinancialperformance.Second,boardactionsdoappeartoberelatedtoboardcharacteristics.FirmswithhigherfractionsofoutsiderdirectorsandsmallerboardstendtomakebetterdecisionsconcerningCEOreplacement,acquisitions,poisonpillsandexecutivecompensation.Finally,boardsappeartoevolveovertimedependingontherelativebargainingpositionoftheCEOrelativetotheexistingdirectors.Firmperformance,CEOturnover,andchangesinownershipstructureappeartobeimportantfactorsaffectingchangestoboards.Twoimportantissuescomplicateempiricalworkonboardsofdirectors,aswellasmostotherempiricalworkongovernance.First,endogeneityisanimportantconsiderationthatisoftendifficulttoaddressempiricallyinthesestudies.Firmperformanceisbotharesultoftheactionsofpreviousdirectorsand,itself,afactorthatpotentiallyinfluencesthechoiceofsubsequentdirectors.Studiesofboardsoftenneglectthisissueand,so,obtainresultsthatare1Mostdirectorscanbeclassifiedasinsidedirectorsoroutsidedirectors.Insidedirectorsareemployeesorformeremployeesofthefirm.TheygenerallyarenotthoughttobeindependentoftheCEO,sincethesuccessoftheircareersisoftentiedtotheCEO’s.Outsidedirectorsarenotemployeesofthefirmandusuallydonothaveanybusinesstiestothefirmasidefromtheirdirectorship.Outsidedirectorsaretypically