行政补偿制度的外文文献

NBERWORKINGPAPERSERIESEXECUTIVECOMPENSATIONASANAGENCYPROBLEMLucianAryeBebchukJesseM.FriedWorkingPaper9813(Bebchuk)andtheBoaltHallFundandU.C.BerkeleyCommitteeonResearch(Fried).TheviewsexpressedhereinarethoseoftheauthorsandnotnecessarilythoseoftheNationalBureauofEconomicResearch©2003byLucianAryeBebchukandJesseM.Fried.Allrightsreserved.Shortsectionsoftextnottoexceedtwoparagraphs,maybequotedwithoutexplicitpermissionprovidedthatfullcreditincluding©notice,isgiventothesource.ExecutiveCompensationasanAgencyProblemLucianAryeBebchukandJesseM.FriedNBERWorkingPaperNo.9813July2003JELNo.D23,G32,G34,G38,J33,J44,K22,M14ABSTRACTThispaperprovidesanoverviewofthemaintheoreticalelementsandempiricalunderpinningsofa“managerialpower”approachtoexecutivecompensation.Underthisapproach,thedesignofexecutivecompensationisviewednotonlyasaninstrumentforaddressingtheagencyproblembetweenmanagersandshareholdersbutalsoaspartoftheagencyproblemitself.Boardsofpubliclytradedcompanieswithdispersedownership,weargue,cannotbeexpectedtobargainatarm’slengthwithmanagers.Asaresult,managerswieldsubstantialinfluenceovertheirownpayarrangements,andtheyhaveaninterestinreducingthesaliencyoftheamountoftheirpayandtheextenttowhichthatpayisde-coupledfrommanagers’performance.Weshowthatthemanagerialpowerapproachcanexplainmanyfeaturesoftheexecutivecompensationlandscape,includingonesthatmanyresearchershavelongviewedaspuzzling.Amongotherthings,wediscussoptionplandesign,stealthcompensation,executiveloans,paymentstodepartingexecutives,retirementbenefits,theuseofcompensationconsultants,andtheobservedrelationshipbetweenCEOpowerandpay.Wealsoexplainhowmanagerialinfluencemightleadtosubstantiallyinefficientarrangementsthatproduceweakorevenperverseincentives.LucianAryeBebchukJesseM.FriedJohnM.OlinCenterBoaltHallSchoolofLawforLaw,Economics,andBusinessUniversityofCaliforniaatBerkeleyHarvardLawSchoolBerkeley,[e-mail],[e-mail]haslongattractedagreatdealofattentionfromfinancialeconomists.Indeed,theincreaseinacademicpapersonthesubjectofCEOcompensationduringthe1990’sseemstohaveoutpacedeventheremarkableincreaseinCEOpayitselfduringthisperiod(Murphy(1999)).Muchresearchhasfocusedonhowexecutivecompensationschemescanhelpalleviatetheagencyprobleminpubliclytradedcompanies.Toadequatelyunderstandthelandscapeofexecutivecompensation,however,itisnecessarytorecognizethatthedesignofcompensationschemesisalsopartlyaproductofthissameagencyproblem.I.ALTERNATIVEAPPROACHESTOEXECUTIVECOMPENSATIONOurfocusinthispaperisonpubliclytradedcompanieswithoutacontrollingshareholder.Whenownershipandmanagementareseparatedinthisway,managersmighthavesubstantialpower.Thisrecognitiongoesback,ofcourse,toBerleandMeans(1933)whoobservedthat“[D]irectors,whileinoffice,havealmostcompletediscretioninmanagement”(p.139).SinceJensenandMeckling(1976),theproblemofmanagerialpoweranddiscretionhasbeenanalyzedinmodernfinanceasan“agencyproblem.”Managersmayusetheirdiscretiontobenefittheirprivateinterestsinavarietyofways(ShleiferandVishny(1997)).Forexample,managersmayengageinempirebuilding(Jensen,(1974),Williamson(1964)).Theymay,asJensen(1986)suggests,failtodistributeexcesscashwhenthefirmdoesnothaveprofitableinvestmentopportunities.Managersalsomayentrenchthemselvesintheirpositions,makingitdifficulttooustthemwhentheyperformpoorly(ShleiferandVishny(1989)).Anydiscussionofexecutivecompensationmustproceedagainstthebackgroundofthefundamentalagencyproblemafflictingmanagementdecision-making.Therearetwodifferentviews,however,onhowtheagencyproblemandexecutivecompensationmightbelinked.Amongfinancialeconomists,thedominantapproachtothestudyofexecutivecompensationviewsthesepayarrangementsasa(partial)remedytotheagencyproblem.Underthisapproach,whichwelabel“theoptimalcontractingapproach,”boardsareassumedtodesigncompensationschemestoprovidemanagerswithefficientincentivestomaximizeshareholdervalue.Financialeconomistshavedonesubstantialworkwithinthisoptimalcontractingmodelinanefforttounderstandexecutivecompensationpractices.RecentsurveysofthisworkincludeMurphy(1999)andCore,Guay,andLarcker(2001).Toresearchersworkingwithintheoptimalcontractingmodel,themainflawwithexistingpracticesseemstobethat,duetopoliticallimitationsonhowgenerously1executivescanbetreated,compensationschemesarenotsufficientlyhigh-powered(JensenandMurphy(1990)).Anotherapproachtostudyingexecutivecompensationfocusesonadifferentlinkbetweentheagencyproblemandexecutivecompensation.Underthisapproach,whichwelabelthe“managerialpowerapproach,”executivecompensationisviewednotonlyasapotentialinstrumentforaddressingagencyproblems–butalsoaspartoftheagencyproblemitself.Asanumberofresearchershaverecognized,somefeatur