Individual and Corporate Social Responsibility

doi:10.1111/j.1468-0335.2009.00843.x

IndividualandCorporateSocialResponsibility

NABOUwandJEANTIROLEzByROLANDBE

wPrincetonUniversityzToulouseSchoolofEconomics

Finalversionreceived6November2009.

Society’sdemandsforindividualandcorporatesocialresponsibilityasalternativeresponsestomarketand

distributivefailuresarebecomingincreasinglyprominent.Wedrawonrecentdevelopmentsinthepsychologyandeconomicsofprosocialbehaviourtoshedlightonthistrendandtheunderlyingmixofmotivations.Wethenlinkindividualconcernstocorporatesocialresponsibility,contrastingthreepossibleunderstandingsoftheterm:firms’adoptionofamorelong-termperspective,thedelegatedexerciseofprosocialbehaviouronbehalfofstakeholders,andinsider-initiatedcorporatephilanthropy.Wediscussthebenefits,costsandlimitsofsociallyresponsiblebehaviourasameanstofurthersocietalgoals.

INTRODUCTION

Economists’viewofhowsocietyshouldbeorganizedhastraditionallyrestedontwopillars.Theinvisiblehandofthemarket,describedinAdamSmith,harnessesconsumers’andcorporations’pursuitofself-interesttothepursuitofefficiency.Thestatecorrectsmarketfailureswheneverexternalitiesstandinthewayofefficiency,andredistributesincomeandwealth,astheincomeandwealthdistributiongeneratedbymarketshasnoreasontofitsociety’smoralstandards.Inindustrializeddemocracies,muchofthepoliticalspectrumhasconvergedonthisdivisionoflabour,albeitwithsomesharpdivergencesofopinionastotherelativerolesofthemarketandthestate.1Textbookeconomicshasthuslongembracedtheshareholder-valueapproach,whichpositsthatfirmsshouldbecontrolledbyprofit-maximizingshareholderswhileotherstakeholdersareprotectedbycontractsandregulation.Stakeholders’insulationfrommanagerialdecisionsoperatesthroughfixednominalclaims(wagesandseverancepay,fixeddebtrepaymentcombinedwithpriorityandcollateral,etc.)andexitoptionsassociatedwithgeneraltraining,flexiblelabourmarketsandshort-termdebtmaturities.Thestateissupposedtostepinonlywhensatisfactorycontractingsolutionswouldinvolvehightransactioncostsorrequiremoresymmetryintheinformationheldbytheparties;casesinpointincludeenvironmentaltaxation,anti-trust,prudentialregulation,andtheregulationofnetworkindustries.Thestatealsocontrolsredistributivetaxation.Inanutshell,followingPigou(1920),thestate,andnotcitizensorfirms,isinchargeofcorrectingmarketfailuresandincomeorwealthinequality.

Yetsociety’sandlaw-makers’demandsforindividualandcorporatesocialresponsibilityasanalternativeresponsetomarketandredistributivefailureshaverecentlybecomemoreprominent.2Certainly,callsforpeopletocontributetimeandmoneytogoodcauseshaveexistedthroughouthistoryandinallsocieties,fromAntiquity3to,say,eighteenthcenturyQuakers’andMennonites’refusaltoinvestinweaponsandslavery.Butthemovementisgainingmomentum,especiallywiththeempowermentofcivilsociety(non-governmentalorganizationsorNGOs)andtheequitable-trade/responsible-investmentmovement.Avarietyoffactorsprobablycombinetoaccountforthistrend:(i)socialresponsibilityislikelytobeanormalgood;(ii)informationaboutcompanies’practicesthroughouttheworldhasbecomemuchmoreaccessibleandquicktotravel;(iii)thescopeofenvironmentalandsocialexternalitiesexertedbymultinationalsinlessdeveloped,morelaxlyregulatedcountriesislikelyto

rTheLondonSchoolofEconomicsandPoliticalScience2010

2

ECONOMICA

[JANUARY

haveexpandedinpacewithglobalization;(iv)thelong-runcostofatmosphericpollution(e.g.globalwarming),oratleastthepublic’sawarenessofit,hasrisensignificantly.

Respondingtosuchdemands,businessleaders,governmentsandacademicsarenowalsoemphasizingthenotionofcorporatesocialresponsibility(CSR).AstandarddefinitionofCSRisthatitisaboutsacrificingprofitsinthesocialinterest.Fortheretobeasacrifice,thefirmmustgobeyonditslegalandcontractualobligations,onavoluntarybasis.CSRtherebyembracesawiderangeofbehaviours,suchasbeingemployee-friendly,environment-friendly,mindfulofethics,respectfulofcommunitieswherethefirm’splantsarelocated,andeveninvestor-friendly.Sometimes,thecallfordutyextendsbeyondthecorporation’simmediaterealmandincludessupportingthearts,universitiesandothergoodcauses.

Whydocitizensandcorporationsempowerthemselvesandsubstituteforelectedgovernment?Afirstandclearlyrelevantmotivationisthatgovernmentmayitselffail.Governmentfailureshavemultipleorigins:

󰀂Capturebylobbiesandotherinterestgroups.GovernmentsunderinfluencemayfailtocorrectexternalitiesasPigovianprincipleswouldrecommend,orbendtowealthyconstituents’oppositiontoredistributivepolicies.

󰀂Territorialityofjurisdiction.Forinstance,onecannotruleagainstchildlabourinadistant,sovereigncountry,andanoutrightimportbanmaybeinfeasibleduetointernationaltradeagreementsorotherpolicyconstraints.Consumerboycottsandinvestoractivismthenbecometheoutletthroughwhichcitizenscanexpresstheiroppositiontothesepractices.

󰀂Acombinationofinefficiency,hightransactioncosts,poorinformationandhighdeliverycosts.Thestatethushasacomparativedisadvantageinpolicingminornuisancessuchasalackofrespectforemployeesorconspicuousconsumptionbyexecutives,orindirectingresourcestoverylocalneeds.‘Appropriate’behavioursinsuchcontextsareinsteadenforcedthroughthepressureofsocialnormsandpopulardemandsthatfirmsbesociallyresponsible.4Asecondimportantmotivationisthateconomicagentsmaywanttopromotevaluesthatarenotsharedbylaw-makers.Becausepreferencesareheterogeneous,itisinevitablethatsomeconsumers’,investors’orworkers’valueswillnotbefullyreflectedinpolicy.They,ororganizedgroupsactingontheirbehalf,willthenbecomeactivists.5Despiteitsgrowingimportance,littleisknownabouttheeconomicsofindividualandcorporatesocialresponsibility.Thispaperdrawsonrecentdevelopmentsinpsychologyandbehaviouraleconomicstoshedsomelightonthenewtrend,itsfutureanditslimits.Itconsidersthebenefitsandcostsofsociallyresponsiblebehaviour(SRB),andaskswhetheritisaviablemodelfortheachievementsofsocialgoals.Thisworkisstillatanearlystageandthepapercanthusprimarilyhelpustoorganizeourthinkingontheseissues.Subjecttothiscaveat,thetake-homemessageisthatSRBholdsrealpromise,providedthatweunderstanditsmotivationsandlimitations.

Thepaperisorganizedasfollows.SectionIdiscussesindividualsocialresponsibility,andSectionIIdiscussescorporatesocialresponsibility.SectionIIIconcludeswithasummaryandsomeopenquestions.

I.INDIVIDUALSOCIALRESPONSIBILITY

Manypeoplegivetocharities,investinsociallyresponsiblefunds,consumegreenproducts,supplytheirblood,orgivetheirtimeandsometimeseventheirlivesforgood

rTheLondonSchoolofEconomicsandPoliticalScience2010

2010]

INDIVIDUALANDCORPORATESOCIALRESPONSIBILITY

3

causes.Suchprosocialbehavioursobeyacomplexmixofinterdependentmotivations.First,theyaredrivenbygenuine,intrinsicaltruism:tovaryingdegrees,weallaspiretodogoodandhelp.Second,materialincentivesmaycomeintoplay:wearemorelikelytogivetocharitiesifcontributionsaretax-deductible.Third,wearealsodrivenbysocial-andself-esteemconcerns.Ourconductdefineswhatkindofpersonweare,intheeyesofothersand,nolessimportantly,inourowneyes.Themainfocusofthissectionwillbetoidentifythebrightanddarksidesofsuchimageconcerns,howtheseinteractwithintrinsicmotivationandextrinsicincentives,andtheimplicationsforpublicpolicy.Someevidenceonimageconcerns

Letusmakethecase,ifneeded,thatprosocialbehaviourscannotbeunderstoodifoneignoresimageconcerns,startingwithsomeexamplesinwhichsocialimageplaysakeyrole.

Anonymousdonationsarewidelyconsideredtobemostadmirable,andyettheytypicallyrepresentatmost1%ofthetotalnumberofdonations(GlazerandKonrad1996).Theimpliedconclusionthatbuyingsocialprestigeispartoftheincentivetoengageinprosocialbehaviourisconfirmedbyseveralrecentexperiments.

AblooddonationexperimentinItalyawardedbronze,silverandgold‘medals’forhowoftenpeopledonatedblood(LaceteraandMacis2008).Suchrankingshadasignificanteffect,butonlywhentheresultswerepublishedinnewspapers.Anotherinterestingfindingofthestudy,whichconcurswithsimilaronesforphilanthropyintheartsandeducationinseveralcountries,isthatpeopletendto‘bunch’rightabovethecut-offsforeachcategory(e.g.BuraschiandCornelli2003).Intheabsenceofimageconcerns,contributionswouldbemuchmoreevenlyspread.

AsimilarconclusionemergesfromtheintroductionofmailballotsinsomecantonsinSwitzerland(Funk2008).Theabilitytovotebymailreducesthecostofvoting,soaccordingtostandardpricetheory,itshouldincreaseturnout.Itactuallydidnotincreasevotingonaverage,andevenreducedparticipationinsmallcommunities.Thissupportstheideathatdemonstratingone’sgoodcitizenshiptoneighboursandfriendsisanimportantpartofthemotivationforvoting,especiallyinsmalltownsandvillageswherethisbehaviourismoreobservablethaninlargecities.Bymakingno-showslessvisible(creatingapotentialexcuse),mailballotsthentendtodepressturnoutinsuchcommunities.Asimilareffectwasrecentlydemonstratedinafieldexperimentondoor-to-doorcharitablesolicitationsbyDellaVignaetal.(2009).Whenhouseholdswereinformedinadvance(byaflyerontheirdoorknob)oftheexacttimebywhenthecollecterwouldcome,thefractionopeningthedoordeclinedby10–25%,andwhentheflyerallowedcheckinga‘Donotdisturb’box,givingwas30%lower.

AnotherinterestingexperimentfocusesonvolunteerfiremeninVermont(CarpenterandMyers2007).Theauthorsfirstmeasuredsubjects’altruismbytheirwillingnesstoshareinanexperimentaldictatorgame,thenrelatedthismeasuretotimespentinvariousfire-volunteeringactivities.Altruismwasfoundtomakeadifferenceforparticipationintraining,whichhasalowpublicvisibility,butnotforcallresponse,wherevisibilityismuchhigher.Conversely,havingonone’scar‘vanity’licenceplatesofatypetypicallyassociatedwithrescueactivities(theMaltesecross)wasassociatedwiththereversepatternoftimeallocation.

Self-imageconcernsarealsoimportantmotivators:weactprosociallyinparttoreassureourselvesthatwearegoodpeople.6Whilethisideaisfamiliarandintuitive,itismostcleanlydemonstratedinlaboratoryexperiments,whereanonymity(eventotheexperimenter,underthedouble-blindprocedure)ensuresthatsocialimageconcernsareirrelevant.

rTheLondonSchoolofEconomicsandPoliticalScience2010

4

ECONOMICA

[JANUARY

Inthestandarddictatorgameexperiment,forinstance,manysubjectsarewillingtomakeatleastasmallsacrificeintheirownpayofftobenefitothers.Thisisprimafacieevidenceoftheexistenceofaltruistic,other-regardingpreferences.Therealpictureismorecomplex,however,asacleverexperimentbyDanaetal.(2007)demonstrates.Astheseauthorsshow,whengiventheopportunitynottoknowwhethertheiractionsactuallyhurtothers,manypeopletakeadvantageofthis‘moralwriggleroom’tomakeselfishchoices.Theirmainexperimentcontraststwovariantsofadictator-likegame,withanonymousplayers.Inafirst,‘known’condition,subjectschoosebetweenaselfishoptionAthatdelivers6tothemand1tosomeoneelse,andafairoptionBdelivering5toeach.Inconformitywithmanyprevioussuchexperiments,aboutthree-quartersofsubjectschooseB.Inasecond,‘uncertain’condition,subjectscanagainchooseeither6(optionA)or5(optionB)forthemselves,buttheydonotknowwhattheotherplayerwillreceiveineachcase.Theyaretoldonlythattherearetwoversionsofthegame,drawnatrandomwithequalprobabilities,thattheycouldbeplaying.Inversion1,thepayoffstoAandBareasabove,namely(6,1)and(5,5),creatingthesametrade-offbetweenmaterialgainandgenerosity.Inversion2,thereisnotrade-off:Aleadsto(6,5)andBtoonly(5,1).Subjectsarealsogiventheopportunitytofindout,immediatelyandatnocost,whichofthetwoversionsofthegametheyhaverandomlydrawnandwillbeplaying.Accordingtostandarddecisiontheory,trulygenerousindividualsshouldstrictlyprefertolearnthestateofnature,asthiswouldallowthemtochooseBinstate1andAinstate2.YethalfofthesubjectschoosenottoknowandproceedtoselectoptionA,presumablyseizingthe(false)excusethattheymaynotbehurtingtheother.Thisbehaviour,akintocrossingthestreettoavoidpassingnearabeggar,showshowsubtlewearewhenweplaygameswithourselves.Weunderstand(probablynotevenconsciously)thatourmemoryisimperfectandthatcreatinga‘cloudofsmoke’aroundwhetherweareactuallyselfishcanthusprovideaself-excuseforpursuingourowninterests.

Otherexperiments,sometimesinvolvingevenmoretransparentself-deception,leadtosimilarresultsandconclusions.Danaetal.(2006)thusshowthatmanypeoplewhowouldvoluntarilyshare$10withananonymousotherinadictatorgameprefertojusttake$9forthemselvesandnotfacethatchoice.Inarelatedvein,Hammanetal.(2009)showthatmanysubjectswhowouldotherwisebehavegenerouslyinadictatorgamewill,giventheopportunity,delegatethesharingdecisiontoathirdpartywhohasacquiredareputationforbeingbiasedinfavourofdelegatingprincipals.Putdifferently,economicagentsareeagertodelegatethe‘dirtywork’thattheywouldnotwanttodothemselves;somehow,notdirectlychoosingtheselfishactionseemstoexoneratethemfromwhatwouldbethelogicaldamagetoself-esteem.

Alltheseexperimentspointtotheideathatself-signalsplayanimportantrole.Accordingly,onewouldexpectthecosttoself-esteemincurredfromselfishactionstobemagnifiedwhenthesebecomemoresalientormemorable.Indeed,Mazaretal.(2008)findthatsubjectswhocancheat(formoney)onataskwithoutanyriskofdetection,cheatlesswhentheyarefirstmadetoreadtheTenCommandmentsortheiruniversity’shonourcode.Suchremindersofmoralpreceptsshould,accordingtothestandardeconomicmodel,beirrelevant.Infact,bymakingtransgressionofthesepreceptsmoresalient,theyreducetheambiguityonwhichself-deceptionrelies,inducingmorehonestbehaviour.Thebrightsideofimageconcerns

Fromapolicyperspective,theimportanceofimageconcernsoffersanopportunity,astheycancheaplybeleveredtoincreaseprosocialbehaviour.Therearenumerous

rTheLondonSchoolofEconomicsandPoliticalScience2010

2010]

INDIVIDUALANDCORPORATESOCIALRESPONSIBILITY

5

PhotographsbyStephanieKuykendalfortheNewYorkTimes

‘AttheShakespeareTheatreCompany’snewhomeinWashington,prettymucheverything,includingstaircasesandthecoatroom,hasbeennamedaftersomebody’(Isherwood2007).

FIGURE1.Thegraffitiofthephilanthropicclass.

examplesofsuchstrategies,suchasthefollowing:

󰀂Onthegoodbehaviourside(tryingtoleveragehonour-seeking),contributorstopublicgoodsreceivepublicpraise,medals,T-shirts,‘bestof’awards,andsoforth.Publicityisamplygivenforgooddeeds,orevenboughtoutright(somethingthatwewillcomebackto),asexemplifiedbyFigure1.

󰀂Onthebadbehaviourside(tryingtoleveragestigma-avoidance),variousmodernversionsofthepilloryaregainingpopularity,suchasinternet‘wallsofshame’fordrunk-drivingarrestsornon-paymentofchildsupport,televisedarrests,andsoforth.7Thisdiscussionpointstoapotentialroleofsociallyresponsibleinvestment,greenproductsandother‘goodcitizenship’modesofconsumptionandsaving.Whenself-viewsareinvolved,weareverygoodatdeception,inattentionandrationalization.Themainvirtueofsociallyresponsibleproductsandtheirwidespreaddiffusionisthattheyprovidefrequentremindersofthingsthatweprefernottothinkabout,suchaspoverty,injustice,ortheenvironmentalimpactofourconsumptions.8Importanceofimageconcerns:thedarkerside

Therearefourpotentialelementsonthe‘darkside’ofexploitingsocial-andself-imagemotivestospurprosocialbehaviourFallofwhichbecomequicklyapparentwhenreflectingonthenatureandcontentofFigure1.

rTheLondonSchoolofEconomicsandPoliticalScience2010

6

ECONOMICA

[JANUARY

(a)Theefficacyofpublicizingpeople’sgoodandbaddeedsis,inasense,self-limiting.As

publicityisscaledup,peoplediscountthemeaningofprosocialacts,attributingtheirmotivationmoretoimage-seekingandlesstoaltruism.Thisformof‘over-justification

naboueffect’impliesapartialcrowdingoutoftheincentiveprovidedbypublicity(seeBe

andTirole2006aforaformalanalysis).Theimplicationisthatthemore‘advertised’sociallyresponsibleinvestments(SRIs)are,themoretheywillbediscounted.

(b)Anothercostrelatestoindividuals’choiceofsignals.Givingisheavilydistorted

towardthemorevisibleormemorabletargets:Americans,forinstance,donatesubstantialamountstoHarvard,Yale,Princetonandotherwell-knownalmamaters,butfarlesstoprimaryandsecondaryschools.Similarly,givingtoconcerthallsandmuseumshasmuchgreater‘glamour’andnetworkingvaluethangivingtothepoor.9Asimilarpointappliestogreeninvestmentbyhouseholds:asArielyetal.(2009)note,buyingahybridcarorinstallingsolarpanelsbuysmoresocialprestigethaninsulatingone’shouseorbuyinganenergy-efficientfurnace.Whilethelatterinvestmentsareinfactmoreecologicallyvirtuous,theyarealsoinvisibletoothers,andafterawhileeventooneself.

(c)Thequestforsocialprestigeortheenhancementofself-imageis,initself,azero-sum

gameFintheparlanceofsociologists,apositionalgood.Thebuyerofahybridcarfeelsandlooksbetter,butmakeshisneighbours(bothbuyersandnon-buyersofhybridcars)feelandlookworseFa‘reputationstealing’externality.Inthelimit,wheneveryonebehavesinasociallyresponsibleway,noonegetscreditforit.

Figure2depictswhathappenswhenparticipationinaprosocialactivityincreases.Individualswithintrinsicmotivation(altruism)abovesomethresholdreceivethehonourattachedtoparticipation,whilethosebelowitabstainandsufferstigma.Whenparticipationincreases,thehonourinvolveddecreases(prosocialbehaviourbecomesmorecommon)andthestigmafornotparticipatingincreases(onlythe‘verybadapples’donotparticipate).

Evidencethatthequestforsocialprestigeisazero-sumgame(oratleastisperceivedtobe)isprovidedbyMonin(2007)andMoninetal.(2008).Inaseriesofexperiments,theseauthorsdocumentafrequentbacklashagainst‘moralrebels’suchasvegetariansorsubjectswhorefusetoplayalonginanethicallyquestionablescenarioFpeoplewhoinprincipleshouldrestoreourfaithinhumannature,butinpracticemakeusfeelmorallyinadequate.Outsidethelaboratory,moreover,suchadversereactionsto‘do-gooders’aresociallycostly:theywillbeostracized,threatenedorharmed,creatingawelfareloss.

(d)Ourmoralstandardsarechallengedalongmultipledimensionsandinrepeated

instances.Goodbehaviourinonecontextmay‘justify’moremediocrebehaviourinanother,andpeoplewhohaverecently‘donegood’inonedimensionmayfeelimmunizedagainstnegative(socialorself)inferences,andthuslateronactlessmorallyconstrained.MoninandMiller(2001)documentthis‘moralcredentializing’inexperimentsongenderorethnicstereotyping:subjectswhowerepreviouslygiven

AbstainStigmaParticipatecut -off v*Honourv:altruism

nabouandTirole2006a.)FIGURE2.Honourandstigma.(Source:Be

rTheLondonSchoolofEconomicsandPoliticalScience2010

2010]

INDIVIDUALANDCORPORATESOCIALRESPONSIBILITY

7

anopportunitytomakestatements‘demonstrating’alackofbiasagainstwomenorminoritiesaresubsequentlymorelikelytomakea(hypothetical)decisionorexpressanopinionthatconformstonegativestereotypesaboutthegroupinquestion.Materialincentives,imageconcerns,andpolicy

Economists’standard‘internalization’recommendationwithrespecttoexternalitiesdatesbackatleasttoPigou(1920):taxesonpollutionorothersociallyharmfulbehavioursshouldequaltheirnetsocialcost;subsidiesforprosocialbehavioursshouldbesetattheirnetsocialbenefit.Amorepsychologicallyinformedapproachtopubliceconomicsdoesnotinvalidatebutqualifiesthisbasicprinciple,whichunderliessomuchofpublicpolicy.Indeed,thereisasecondexternalityinvolved,arisingfromthe‘reputationstealing’effectdiscussedabove.ThesociallyoptimalincentiverateshouldthereforebelowerthantheusualPigovianlevel:itmustsubtractfromthestandardexternalitycosttheimagevalueimplicitly‘bought’bytheindividualorfirm.

AmendingPigouThissimplepolicyprescriptionraisesseveralissues.First,imagevaluesmaynotalwaysbeeasytomeasure,althoughthereareimplicitmarketsforitemssuchasanamedbuilding,aplaqueorauniversitychair.Second,andrelatedly,onewouldwanttoknowwhenthecorrectioncanbesignificantandwhenitislikelytobenegligible.Inthisrespect,wehavealreadyseenthatimageconcernsareinfluencedbyvisibility:efficientfurnacesmayneedtobesubsidizedmorethanhybridcars.

Perhapsmoresubtly,imageconcernsdependendogenouslyontheprevalenceofthebehaviourinquestion.Providedthatthedistributionofaltruisminthepopulationis

nabouandTirole2010)thatforagivenunimodal(asinFigure2),onecanshow(seeBe

socialcostoftheexternality,theoptimalsubsidyvariesnon-monotonicallywiththeproportionofpeoplechoosingthesociallyresponsibleoption:seeFigure3,whichdepictsthiscorrectiontoPigoviantaxation.Intuitively,‘admirable’actsFthosethatveryfewindividualsinsocietyperform(orwouldperformwithoutconsiderableincentiviza-tion)Fcarrysubstantialsocialprestige,andshouldthereforecommandonlymodestmaterialincentives.Attheotherextreme,‘respectable’actsFthosethat‘anydecentperson’wouldengagein(absentexceptionalconstraintsorinducements)Faresuchthatfailingtoperformthementailsenormousstigma;sohereagain,materialincentives

nabouandTirole2010).FIGURE3.CorrectingPigou’srule.(Source:Be

rTheLondonSchoolofEconomicsandPoliticalScience2010

8

ECONOMICA

[JANUARY

shouldbelow.‘Modalacts’,whichareneitheradmirablenormerelyrespectable,arethoseforwhichPigovianpreceptsapplybestandextrinsicincentivesshouldbestrongest.Crowdingoutandtheover-justificationeffectTheobservationthattheimagevalueofdifferentbehavioursisendogenous(determinedtogetherwiththeequilibriumbehaviourofthewholepopulation)hasanotherimportantimplication:thelevelofmaterialincentiveswillaffectthisattributionproblem,andthereforethe(self-)reputationalreturn

nabouandTirole(2006a).toprosocialacts.ThisisillustratedinFigure4,drawnfromBe

󰀂toapublicgood(say,Thisfiguredepictsapopulation’saveragecontributiona

donatingblood)orconformitytosomesociallydesirablebehaviour,asthematerialrewardyfordoingso(or,conversely,thepenaltyforfailingto)increases.10Intheabsenceofimageconcerns,aggregatebehaviourconfirmstherepresentationofHomoEconomicusinEcon101:ahigherrewardelicitsincreasedsupply.Asimagebecomemoreimportant,anunusualphenomenonoccurs:oversomeinterval,thesupplyresponsetoincentivesflattensout,andeventuallybecomesdownward-sloping.Thisviolationofbasicpricetheoryresultsfromwhatpsychologistscalltheover-justificationeffect,andeconomistscallasignal-extractionproblem.Whenthereisnoorlittlereward,aprosocialactisinterpretedasgenuinealtruism.Asmaterialincentivesbecomemoresubstantial,the‘meaning’oftheactchanges:itbecomesmoredifficulttoknowtowhatextentitismotivatedbyaltruismorbygreed(sincepeoplealsodifferalongbothdimensions).Thesignallingvalueofprosocialbehaviourthusweakens,offsettingorevenreversingthedirecteffectofhigherincentive.Ofcourse,whenrewardsbecomesufficientlylarge,supplyeventuallyrecoversitsstandardupward-slopingnature,butsuchlevelsofpaymentscouldbeverycostly.

Figure4andthemoregeneralanalysisunderlyingithaveothertestableimplicationsbesidesthepossibilityofcrowdingout.First,prosocialcontributionsincreasewhentheirvisibility,andthereforetheirimpact,onsocialesteem,increases.Second,thepowerofincentivesisweaker(orevennegative)whenimageconcernsarelarge.Arielyetal.(2009)provideempiricalvalidationsofthesepredictions.Intheirexperiments,subjectsperformedtaskstoearnmoneyforcharitablecausesundertwoconditions.Ina‘private’condition,onlytheindividualobserveshiseffortandamountearned(onlyhisself-imagecanbeatstake).Inthe‘public’condition,allsubjects,andpossiblysomeotherpeers,

aggregate supply

increasing image concern:people start wondering

whether done by altruism orfor the money

5037.525no image concern (perhapsindividual is not observed):homo economicus12.5incentive: y

–2.5

0

2.5

5

7.5

10

12.5

15

nabouandTirole2006a.)FIGURE4.Theoverjustificationeffect.(Source:Be

rTheLondonSchoolofEconomicsandPoliticalScience2010

2010]

INDIVIDUALANDCORPORATESOCIALRESPONSIBILITY

9

observeeachperson’schoices(sobothself-andsocial-imageareinvolved).Thefirstobservation,whichcorroboratesmanyearlierfindings(e.g.Freeman1997),isthatcontributionsarelargerwhentheyarepublic.Thesecondobservationismorestriking:inconformitywiththetheoreticalprediction,theimpactofmaterialincentivesweakensconsiderably(orevenreverses)whentheyarepublic.Moreover,thischangeinslopeoccursonlywhenthetaskisperformedforagoodcause:itisabsentwhenitisknownthatsubjectsaresimplyearningmoneyforthemselves,orforanorganizationthatmostobserversdisapproveof.Theseresultsareconsistentwiththeideathatsubjectsinthepublicconditionseemworriedthatthepresenceofrewardsmaycausetheirprosocialbehaviourtobeattributedtogreedmorethantoaltruism.Movingbeyondthelaboratory,suchfindingsprovideafurtherrationaleforsubsidizingefficientfurnacesmorethanhybridcars.

Aswenotedearlier,asecondformoftheover-justificationeffectarisesfromthefactthatpeoplealsodifferintheintensityoftheirimageconcerns.Someareliterallyobsessedwithsocialesteem,whileotherscareaboutitmuchless.Whengooddeedsareveryvisible(bytheirnatureorbypolicydesign),theywillthenbe‘discounted’asbeinglikelytoreflectastrongdesireto‘buy’admiration.Thisdiscount,inturn,willdampenthesignallingincentivetoactprosocially,andthisevenwhenimageseekingisnotfrownedonperseFitjustreducestheattributionofgoodbehaviourtogenuinealtruism.Thiseffectneverfullycrowdsouttheimpactofanincreaseinthepublicityofactions(medals,citations,wallofshame,etc.),butcansubstantiallydampenit.11Such‘concernsaboutnotlookingimage-concerned’resonate,astheyarepresentinoureverydaylives.Mostpeoplehesitatetoboastabouttheirgoodactionsandwouldmuchratherhavethirdpartiesadvertisetheseactsforthem.Thisfollowsnaturallyfromthelogicofinference:someonewhobragsabouthisgooddeedssignalsthatheisimage-concernedandcreatesasuspicionabouttheextentofhistruealtruism.

II.CORPORATESOCIALRESPONSIBILITY

Corporatesocialresponsibility(CSR)issomewhatofa‘catch-all’phraseforanarrayofdifferentconcepts.AnanalysisofCSRmustthereforeclarifyitsexactmeaning,andinparticularthepresumedimpactofCSRonthecostofcapital.SomeCSRadvocatesarguethatthereisabusinesscaseforgoodcorporatebehaviour,whileothersdiscussitintermsofsacrificingsomeprofitinthequestforthesocialgood.Thissectionis,accordingly,dividedintotwoparts.ThefirstdiscussesthreealternativevisionsofCSR.TheseconddiscussesthechallengesandopportunitiesfacedbytheCSRmovement.ThreeviewsonCSR

Vision1:‘Win–win’(‘doingwellbydoinggood’)AccordingtosomeadvocatesofCSR,beingagoodcorporatecitizencanalsomakeafirmmoreprofitable.Sincefirmspresumablyhavenointerestinsimultaneouslyreducingprofitsandharmingsociety(damagingtheenvironment,hurtingworkersorconsumers),forsuchaclaimtobemorethanmanagementconsultants’promiseofa‘freelunch’,itrequireselaboration,aswellassupportingevidence.Therearetwowaysinwhichitcouldbereasonablyinterpreted,correspondingrespectivelytovision1hereandvision2below.

Thefirstoneinvolvestheexistenceoflimitstogovernanceandmanagers’temporalhorizons.Asalargeliteratureinfinancehasemphasized,firmsoftensufferfromashort-termbias.Thismaybeduetopoorlystructuredmanagerialincentives,butsuchbiases

rTheLondonSchoolofEconomicsandPoliticalScience2010

10

ECONOMICA

[JANUARY

canalsoresultfromwell-designedschemes.First,monetaryincentivesoftenputmoreweightonshort-termthanonlong-termperformance.Althoughtherecentcrisishasbroughtmorewidespreadrecognitionofthehazardsthatshort-term-orientedcompensationschemescreateforcorporationsandsociety,somedependenceoncurrentorrecentfirmperformanceisinevitable.Second,decisionsbyboardsandshareholdersaboutwhethertokeepcurrentmanagement,changeitoralterthescopeofitsactivitiesarealsonecessarilybasedinpartonrecentobservation(evenifsomeofthelong-termimpactofmanagers’actionsmayfilterthroughlong-termindicatorssuchasthestockprice12).Thuscareerconcernsalsogeneratesomeshort-termbiases.

Inpractice,short-termismoftenimpliesbothanintertemporallossofprofitandanexternalityonstakeholders.Thatis,managerstakedecisionsthatincreaseshort-termprofit,butreduceshareholdervalueandhurtworkersorotherconstituencies.Forexample,afirmmayrenegeonanimplicitcontractwithitslabourorsupplierssoastoreducecosts,therebydamaginggoodwillFmakingitmoredifficulttoattractmotivatedworkersinthefuture,ortoinducesupplierstomakerelationship-specificinvestments(e.g.KruegerandMas2004).Alternatively,afirmcouldeconomizeonsafetyorpollutioncontrol;thisalsoincreasesshort-runprofits,butcreatescontingentliabilitiesdowntheroadFriskoffuturelawsuits,consumerboycottsandenvironmentalclean-upcosts.Insuchcases,the‘win–win’argumentmakesclearsense.

Theupshotisthatinthisfirstvision,CSRisabouttakingalong-termperspectivetomaximizing(intertemporal)profits.Thissuggeststhatsociallyresponsibleinvestorsshouldpositionthemselvesaslong-terminvestorswhomonitormanagementandexertvoicetocorrectshort-termism.

Inthesamevein,butwithmuchmoreambiguouswelfareconsequences,‘strategicCSR’(touseatermcoinedbyBaron2001)consistsintakingasociallyresponsiblestanceinordertostrengthenone’smarketpositionandtherebyincreaselong-termprofits.Forinstance,CSRcouldbeameansofplacatingregulatorsandpublicopiniontoavoidstrictsupervisioninthefuture,ortoattempttoraiserivals’costsbyencouragingenvironmental,labourorsafetyregulationsthatwillparticularlyhandicapcompetitors.

Vision2:Delegatedphilanthropy(thefirmasachannelfortheexpressionofcitizenvalues)ForthereasonsdiscussedinSectionI,somestakeholders(investors,customers,employees)areoftenwillingtosacrificemoney(yield,purchasingpowerandwage,respectively)soastofurthersocialgoals.Putdifferently,stakeholdershavesomedemandforcorporationstoengageinphilanthropyontheirbehalf.ThecorrespondingCSRprofitsacrificeisthenpassedthroughtostakeholdersattheirdemand.

Notethatoneneedstoexplainwhypeoplewouldwantcorporationstodogoodontheirbehalf,ratherthandoingitontheirownorthroughcharitableorganizations,churches,etc.Informationandtransactioncostsareclearlyimportanthere.Intheory,consumerscouldsendmoneytodirectlysupplementtheincomeofworkersinthecoffeeplantationssupplyingStarbucks.Buttheywouldhavetobeinformedabouttheoccurrenceofindividualtradesandcontracts,andtheirfinancialtransferswouldinvolveenormoustransactioncosts.Somehow,philanthropymustthusbedelegated.Itcouldperhapsbeentrustedtosomecharitableorganization,buttransactioncostsarestilllikelytobemuchlowerifdelegationgoesthroughthecorporation,whichalreadyisinvolvedinafinancialrelationshipwiththeworkers.

Anotherargumentforaskingcorporationstobehaveprosociallyisthatthedesiredactionsareoftennotabouttransferringincometoless-favouredpopulations,butabout

rTheLondonSchoolofEconomicsandPoliticalScience2010

2010]

INDIVIDUALANDCORPORATESOCIALRESPONSIBILITY

11

refrainingfromspecificbehaviours,suchaspollutingtheenvironment;herethereisnosubstituteforaskingthefirmtobehavewellwhenthestatedoesnotimposeconstrainingregulations.Arelatedcaseiswhenafirmdrawsonitstechnicalexpertiseorexploitscomplementaritiestodelivergoodsandservicestothoseinneedmoreefficientlythangovernmentsorotherphilanthropic‘intermediaries’could.Examplesincludeagiantsupermarketchainorganizingreliefconvoystoazonehitbyahurricane,oralargewater-treatmentutilitysettingupaprogrammeofdiggingwaterwellsforpoor,remotevillagesinadevelopingcountry.

Manyexamplescometomind:Starbucksincreasesitsdemandbybuyingfair-tradecoffeeandtea.Otherfirmsadvertiseheavilythattheirclothingismadefromorganiccotton,oristheproductoffairtrade.Somecorporationsprovideincentivesforemployeeengagementincommunityservice,therebyboostingpublicrelationswiththelocalcommunitiesandattractingmotivatedemployees(BesleyandGhatak2005;BrekkeandNyborg2008).Theviewthatcorporationsengageinsociallyresponsiblebehaviour(SRB)onbehalfofstakeholdersisalsosupportedbytheobservationthat‘sinstocks’(tobacco,alcohol,casinos)exhibithigherreturns(HongandKacperczyk2009).13Ontheotherhand,andaswewilllaterdiscuss,otherempiricalresearchfailstodemonstratealinkbetweenCSRbehaviourandalackofprofitability.

TheideathatfirmsexercisingCSRarerespondingtoconsumerandinvestordemand,thatthey‘dogood’ontheirbehalf,isconsistentwiththegreaterprevalenceofsuchpracticesamongfirmsthatarelarge,areprofitable,producefinalgoodsandarescrutinizedbyNGOs.VisibilitywithrespecttostakeholdersdemandingSRBthusincentivizesfirmstoengageinsuchbehaviour.14Aswithindividuals,theimageconcernsofcorporationsalsohavetheirdarkerside,takingheretheformof‘greenwash’:disseminatingamisleadingpictureofenvironmentalfriendlinessorotherSRB,oronethatisaccurateinsomedimensionsbutservestoobscurelesssavouryones.ThusEnronwasahugecorporategiver,particularlytotheHoustonarea,andoneofthemostimpressive‘glossybrochures’documentingthemultiplefacetsofafirm’sCSRbenevolenceistheoneissuedin2007bytheAmericanInternationalGroup.15Vision2ofCSRdoesnotraiseanyspecificcorporategovernanceissue:managementcaterstodemandandmaximizesprofit.Aswiththelong-termperspective,profitmaximizationandCRSareconsistent.

Vision3:Insider-initiatedcorporatephilanthropyInthisinterpretationofCSR,corporateprosocialbehaviouris(atleastinpart)notmotivatedbystakeholders’demandsorwillingnesstosacrificemoneyforagoodcause,butratherreflectsmanagement’sortheboardmembers’owndesirestoengageinphilanthropy.Forinstance,corporationsoftengivetocharitiesontheboardsofwhichtheirexecutivesorownboardmemberssit,ortoinstitutions(concerthalls,operahouses,museums)andcauses(e.g.politicalthinktanks)whichtheirtopmanagementfavours.16Profitisthentypicallynotmaximized.

Thistypeofphilanthropyhascomeunderattackfromboththerightandleftsidesofthepoliticalspectrum.Inawell-knownpiece,MiltonFriedman(1970)wroteinessencethatcorporationsshouldnotdocharitywithothers’money.Rather,managersanddirectorsshouldemploytheirownwealthtothispurpose.Ontheothersideoftheaisle,RobertReichhasarguedthatthereisnowaytoensurethatprivatemoneywillgotothe‘right’causesandthatfirmsshouldnotsubstituteforthestate(meaning,presumably,thatelectionsprovidethelegitimacytodefinewhatis‘right’).Inpractice,thestaterestrictsthesetofpotentialrecipientsofcorporategenerositybydecidingwhich

rTheLondonSchoolofEconomicsandPoliticalScience2010

12

ECONOMICA

[JANUARY

institutionsareeligiblefortax-deductiblecontributions.So,whileithasveryimperfectcontrolovertheallocation,itatleastkeepscontrolovertheidentityofrecipients.

Unlikethecitizen-delegationview,theviewofcorporatephilanthropyasmanage-ment-initiatedraisessubstantialcorporategovernanceissues.Indeed,ifinvestorssimplydemandthehighestpossiblereturn(implyingnotoleranceforSRBiftheydonotbenefitfromit),managersandtheboardmustbesomewhatentrenchedinordertobeabletopractisecorporatephilanthropyonalargescale(CespaandCestone2007).Thenotionofthemissionofmanagementbeingbroaderthanjustmaximizingshareholdervaluenecessarilyinvolvessomecosttocorporations,makingitmoredifficultforthemtoraisefundsfrominvestors.17Itmayalsoweakenmanagerialaccountabilitybycreatingmultipleobjectivesandperformancecriteria;attheextreme,toomanymissionsamounttonomissionatall(Dewatripontetal.1999).

Inpractice,thedividinglinebetweenthedifferentnotionsofCSRFlong-termperspective(vision1),delegatedphilanthropy(vision2)andinsider-initiatedcorporatephilanthropy(vision3)Fmaybeelusive.Consider,forinstance,theincreasinglypopularpracticesofnominatingsomeonewithagoodexternalreputationasaCorporateSustainabilityOfficer,orofturningtoreputableNGOsforadviceandhelp.DofirmshireaCorporateSustainabilityOfficertoserveasanadvocateagainstshort-termismFakindofenvironmentalriskmanager,ineffectFwithinthefirm?Oristhisexecutivemeanttobeavoiceforcostlyprosocialbehaviourdemandedbystakeholders?Similarly,considertheintroductionofmonetaryincentivesbasedonenvironmentalperformance.Aretheymeant,say,toencouragedivisionalmanagerstoinstallcarbon-freeequipmentthatiscurrentlyunprofitableduetotheworld’slaxattitudetowardclimatechange,butwillreducefuturecostswhencarbondioxideistaxedatamorereasonablelevel?Orisagenuinedesiretodogoodabetterinterpretationofgreenincentivesthanthelong-termperspectiveview?Inthelattercase,arethese‘greenincentives’theobjectofintensecommunicationtoshareholders(suggestingvision2)oraconfidentialpolicy(suggestingvision3)?

Insum,weseethat,aswithindividualconsumersandinvestors,corporate‘sociallyresponsiblebehaviours’oftencarrymuchambiguityastotheirexactmotivation.

Dothedatahelpustotellthesetheoriesapart?Empiricalstudiesoftenrelatecorporateprofitabilitywithsociallyresponsiblebehaviour.18Thereseemstobe,overall,nooraslightlypositivecorrelationbetweensociallyresponsiblebehaviourandcorporatereturns;seeOrlitzkyetal.(2003),MargolisandElfenbein(2007)andMargolisetal.(2007)formeta-analyses,Heal(2005)andReinhardtetal.(2008)forfurtherdiscussions.Theinterpretationofempiricalanalysesis,however,subjecttothreedifficulties.

ThefirstdifficultyconcernswhichCSRtheoryisbeingtested.Weobservedthatvisions1and2bothpredictapositivecorrelationbetweenCSRandprofits,19whilevision3predictsthereverse.Inpractice,CSRislikelytoinvolveamixofallthreeacrossthecorporatesample,soitisunclearwhatspecificchannelisbeingtested.

Theseconddifficultypertainstotheempiricalstrategy.Onthecausal-inferencefront,SRBandprofitabilityareclearlybothendogenousvariables,raisingtheusualissueofwhataretheexogenousdrivers(equivalently,whatwouldbeappropriateinstruments)underlyingtheobservedcorrelationpatterns.Ifmanagersdifferexogenouslyintheirindividualtimehorizons,forinstance,thenapositivecorrelationistobeexpected(fromvision1).Iftheydifferintheprivatevaluethattheyderivefrombeingassociatedwithafirmknowntoactwelltowarditsemployees,consumersortheenvironment,then(fromvision3)onemightexpectanegativerelationship.Therecouldalsobereversecausation,

rTheLondonSchoolofEconomicsandPoliticalScience2010

2010]

INDIVIDUALANDCORPORATESOCIALRESPONSIBILITY

13

forinstanceifthemostprofitablefirmsarethebestabletoaffordthecostofCSRor,assuggestedbyMargolisandElfenbein(2007)andMargolisetal.(2007),havethemostincentivetoengageinit.20Anempiricalstrategymustalsochoosebetweenoperatingperformanceandstockmarketreturnstomeasure‘corporatereturns’.Bothinvolvedifferentproblems.Foroperatingperformance(e.g.returnonassets),theempiricalchallengeisthefollowing:ifonebelievesthatCSRincreasesthemeanprofitbylimitingraredisasters(vision1),thenoneneedsverylargesamplestoseethatsignificantlyinthedata.Moststudieslookinsteadatstockreturns.ThisraisestheissueofwhetherCSRcompanieshavedifferentsystematicriskexposures(eitherduetotheirresilienceinperiodsofcrisisorbecausetheyfaceaspecificCSRriskfactor),inwhichcasetheycommanddifferentriskpremia,andthereforehavedifferentexpectedreturns.AsAndries(2008)notes:‘Newlawsandregulationsaretypicallyintroducedwhenthingsgowrong...Sociallyresponsibleinvestorswouldsimplybetheoneswithhigherriskaversiontoadeteriorationofthe‘‘stateoftheworld’’.’

AnotherissueiswhetherfinancialmarketsarestilllearningaboutCSR.Somepeoplehaveinmind(implicitlyorexplicitly)thatthestockmarkethasbeenundervaluingthe‘trueimportance’ofCSR,sothatvirtuouscompanieshavepositiveabnormalreturns(theykeepsurprisingthemarketpositively,incontradictiontomarketefficiency).Anintermediatestoryisa‘slowrepricing’,wherebyenvironmentalandsocialfactorsaregraduallybecomingrecognizedasrelevantpricefactorsforvaluingacompany;virtuousfirmsexperiencehighreturnsduringthisrecognitionperiod,butshouldexperienceloweronesoncetherepricingiscompleted.

Thethirdunknowninvolvesextrapolation.Sociallyresponsibleinvestmenthas,sofar,mostlybeenamarginalphenomenon,andthereisnoreasonwhyprofitabilityshouldremainthesameasitsprevalenceincreases.Akeyissuehereiswhichtypesofsociallyresponsibleactivitiesinvolvedecreasingor,conversely,increasingreturns.Forinstance,asmallnumberofenvironmentallyconcernedinvestorscanalwaysinvestintheexistingcleancompanies;itisthereforelikelythatCSRhashadlittleimpactonstockpricesinthepast.AsSRIpopularityincreases,otherfirmsmayhavetostartincurringhigherabatementcostsinordertoattractfunding,depressingyields.21ConsistentwiththisideaofanincreasingCSRdiscount,Andries(2008)finds,inthecross-sectionofS&P500firmsandaftercontrollingforindustryandfirmcharacteristics:(i)nosignificantrelationshipbetweenassetreturnsandCSRoverthe1991–2006period;(ii)anemergingnegativerelationshipforsomecriteriarelatedtonegativescreening(tobacco,alcohol,gambling,nuclearpower)overthemorerecent2000–06period.Challengesfordelegatedphilanthropy

Thissubsectionfocusesonwhatmaybethefastest-growingformofSRB(atleastintermsofvisibility),delegatedphilanthropy,anddiscussessomeofthechallengesahead.Delegatedphilanthropy,aswenoted,isaresponsetoawidespreaddemandbystakeholdersthatfirmstheyinteractwithbegoodcorporatecitizens.Whileitdoesnotcallintoquestiontraditionalformsofgovernance,itdoesfaceseriouspracticalchallenges,whichwillneedtobeconfrontedheadonifthisapproachtoCSRistorealizeitsfullpotential.

FreeridingSociallyresponsibleinvestment,thefair-trademovement,andthedesiretobeemployedbyasociallyresponsibleinstitutionallinvolveaprivateprovisionofapublicgood.Thetemptationtofreerideissubstantial.Wemaybewillingtopay5%morefor

rTheLondonSchoolofEconomicsandPoliticalScience2010

14

ECONOMICA

[JANUARY

ourelectricityifitiscarbon-free,tosacrificeabitontheyieldonoursavings,ortobepaidalittlelessinordertofeelgoodaboutouremployer.Buttherearelimitstosuchsacrifices.Whilepeoplealmostunanimouslydeclarethemselveswillingtoincurcoststoimprovetheenvironmentorpromotedevelopment,attitudeschangewhenthingsbecomemoreconcrete.Accordingtorecentpolls,65%ofFrenchpeopleareopposedtothecurrent,modestcarbontax,whichtheyandallthecandidateshadembracedduringthepresidentialcampaigntwoyearsagoasadesirablenationalcontributiontothereductionofglobalwarming.ItwouldbeinterestingtoknowhowfreeridingwillevolveifSRIbecomesverypopularandifthis,throughdecreasingreturns,significantlyreducestheyieldonsociallyresponsiblefunds.

InformationInordertochooseacompanytoinvestin,buyfromorworkfor,investors,consumersandworkersneedinformationastowhetheritreallybehavesprosocially.Thisraisesthreechallenges:

󰀂Datacollectionisitselfapublicgood.Itisthereforeimportantthatspecializedratingagenciessupplytherequiredinformationtothepublic.Ofcourse,theseagenciesmayfaceinadequateincentives,aswasdemonstratedintherecentfinancialcrisis,andtherearegroundstobevigilant.Buttheproblemofreportingisnotspecifictothiscontext,andwehavetomeettheneedfordelegation.

󰀂Differentdimensionsofgoodcorporatecitizenshipneedtobeaggregated.Firmsdowellinsomedimensionsandpoorlyonothers,sooneofthechallengesfacingratingagenciesistofindamethodologyforaddingthemupintoasyntheticindex(orasmallnumberthereof).HowdoesoneassesstheclosureofaplantthatemitsalotofCO2butprovidesjobstoalocalcommunity?Doesoneviewnuclearpowerintermsofitsdefinitelyfavourableimpactagainstglobalwarming,orofitslong-lastinghazardouswaste?Canamultinationaloffsetsomelocalenvironmentaldamagebyfinancingaschool,clinicorwaste-treatmentfacilityinthecommunity?

󰀂Shouldcorporatesocialperformancebeassessedinabsoluteorrelativeterms?Forexample,anoilcompanymaypollutealot,butmakesubstantialeffortstoreduceitspollutionandbe‘bestinclass’.Relatedly,LandierandNair(2008)havesuggestedincludingrelativeperformancewithintheindustryasacriterionfordelineatingsociallyresponsibleportfolios.DefiningwhatissociallyresponsibleCSRinheritsthestrengthsandweaknessesofthedemocraticprocess.Asiswell-known,politiciansoftentendtofollowpublicopinioninordertopandertotheelectorate.22Goodpublicpolicyisthereforeoftenconditionedbyaproperunderstanding,oratleastalackofprejudicebyvotersintherelevantmatter.Acaseinpointisenvironmentalpolicy.Europeanpoliticiansmayimposeacostofover1000eurostoeconomizeonetonofcarbonbysubsidizingsolarpanels(mostcostestimatesrangefrom600to1400euros),whentheycouldhaveeconomizedover50tonswiththesamemoneybybuyingoutthecarbonrights.(Atthetimeofwriting,thepriceofatonintheEU’sEmissionsTradingSystemis15euros.)Butduetopoorinformationoranimperfectunderstandingoftheeconomicsofthematter,oragainthebiastowardmorevisible,‘symbolic’interventionsintroducedbyvoters’identityconcerns,thereisnoelectoralsanction,butactuallyanincreaseinpopularity,attachedtosuchmoves.

Similarly,thechallengeforaproperassessmentofwhichfirmsaresociallyresponsibleistodisincentivizeratingagenciesandCEOsfrompanderingtothebiasesoftheirconstituencies(investors,consumersandworkers).Thesameproblemarisesin

rTheLondonSchoolofEconomicsandPoliticalScience2010

2010]

INDIVIDUALANDCORPORATESOCIALRESPONSIBILITY

15

manycontexts.NGOsresellingconfiscatedivorywillbefrownedonbythepublicforbehaving‘immorally’,evenwhenthismovelowersthepriceofivoryandtherebydiscouragespoachers.ApollutermaystillbefrownedonafterataxonthepollutantisimposedatitsPigovianlevel(thepolluteristhensubjectto‘doubletaxation’).Regardlessoftheirownattitudestowardgeneticallymodifiedorganisms,mostexpertswouldadmitthatpopularattitudesonthesubjectareoftendrivenbyratherunscientificargumentsandbasedonlittleinformation.Finally,framingisboundtodistortassessments:thechoiceofpresentingafirm’srationaleforlocatingaplantinalow-labour-costcountryas‘helpingapoorcountrytodevelop’or‘minimizinglabourcost’isobviouslynotneutral.Whatkindofactivism?Amuchdebatedquestionforgreenorethicalfundsisthenatureoftheiraction.Shouldtheyinterveneingovernanceormorepassivelyvotewiththeirfeet?Inthelattercase,shouldthefundengageinadialoguewiththefirmbeforeexcludingitfromitsportfolio?

Typically,greenorethicalfundsareinformed(bylobbies,somemultilateralorganizations,etc.)ofan‘incident’,i.e.afirm’spotentialmisbehaviour.Thefundthen‘monitors’andentersa‘dialogue’or‘engagement’withthefirmtopersuadeittoalteritsbehaviour,threateningaboycottifitdoesnotcomply.SomefundsandNGOsprefera‘cleanhands’approachthatdoesnotinvolveanysuchdialogue,forfearofbeingseenassoftontheinfringerorofbeingledtoacompromisethatdoesnotsatisfythesociallyresponsibleinvestors.

Anotherdebaterelatestothechoicebetweenpositiveandnegativescreening.Negativescreeningreferstotheexclusionofunrepentantfirms;positivescreeningconsistsininvestingonlyin‘best-in-class’corporations.

Athirddebateconcernswhetherthenamesofcompaniesfacingengagement,andnotonlythosethatendupexcluded,shouldbepublished.Theremaybeagoodargumentfornotdisclosingthenamesoffirmsinvolvedinnegotiations,soastogivethemmoreofanincentivetoaltertheirbehaviour.Ontheotherhand,thismayincreasetheriskofcollusionorcaptureoftheNGOorfundinvolved.

III.CONCLUDINGTHOUGHTS

Whiletheinvisiblehandofthemarketandthemorevisibleoneofthestatehavebeentheobjectsofmuchresearch,westillknowlittleaboutthedecentralizedcorrectionofexternalitiesandinequality.

Thispaperhasarguedthattherearethreepossibleunderstandingsofcorporatesocialresponsibility:theadoptionofamorelong-termperspective,thedelegatedexerciseofphilanthropyonbehalfofstakeholders,andinsider-initiatedcorporatephilanthropy.Thelattertwounderstandingsbuildonindividualsocialresponsibility,whichledustoreviewindividualmotivationsforprosocialbehaviour.Wesawthatprosocialbehaviourbyinvestors,consumersandworkersisdrivenbyacomplexsetofmotives:intrinsicaltruism,materialincentives(definedbylawandtaxes)andsocial-orself-esteemconcerns.Thesethreemotivesaremutuallyinterdependent,andbothpolicy-makersandsocialactivistsmusthaveagoodunderstandingoftheseinteractionsinordertoproperlyharnesspeople’sdesiretobehaveprosocially.

Wethussawthattheimportanceofsocial-andself-signallingconcernsprovidesapowerfulandcheapleverforcertaintypesofinterventions(uptoapoint)butmakesotherscounterproductive,andthatthepursuitofsocial-andself-esteemperseisazero-sumgamedistortingactionstowardsthemorevisible.

rTheLondonSchoolofEconomicsandPoliticalScience2010

16

ECONOMICA

[JANUARY

Whilewehavefocusedafairamountofattentiononimageconcerns,other‘behavioural’aspectsshouldalsobeaddedtothepictureinordertoaugmentourunderstandingofgivingpatterns.Aswithotherchoicessuchasretirementsavings,defaultoptions(e.g.havingtooptinorout)willberelevant,ifonlybecausetheyconveyinformationastowhatisexpectedbysociety.Shouldcarbonoffsetsbeincludedinaeroplaneticketswithanopt-outoption?Shouldafixedpercentageofsalariesbeallocatedtocharitiesbydefault,orsomefractionofaUS401Kcontributiondirectedtosociallyresponsiblefunds,againwithanopt-outoption?Anotherinterestingquestioniswhetherprosocialbehaviouris,likeexerciseandgoodhygiene,habitforming.Presumably,abeliefinsuchvirtuoushabitformationunderlies‘compulsoryvolunteer-ing’asaprerequisiteformostcollegeadmissionsintheUSA.

Researchshouldobviouslytackleanotherfundamentalquestion:who,amongthestate,stakeholdersandfirms,isbestplacedtoaddressmarketfailuresandinequality?Aretheiractionscomplementaryorsubstitutes?Couldpublicprivatepartnershipsmaintainpublicinvolvementwhileharnessingprivatesectorresources,efficiencyandvalidation?

InthispaperwepurposelyofferedasomewhatcautiousviewofSRIandCSR,butourchoicetoworkonthistopicshouldbeinterpretedasasignofoptimism.Caringabouttheenvironment,thewelfareofpeopleinpoorcountries,andothergoodcausesisanormalgood.Thericheroursocieties,thehigherthedemandforSRB.Butoneneedstoleveraltruismandsignallingconcernsintherightway,andthisrequiresagoodunderstandingofthepsychologyofgivingandhowitinteractswithmarketsandothergeneralequilibriumforces.Economists,guidedbythehandsofPigouandSmith,havesofarpaidinsufficientattentiontothistopic.Wehopethatthispaper,despiteitsmodestscope,willencouragemoreresearchonthistopic.

ACKNOWLEDGMENTS

ThispaperformedthebasisfortheCoaselecture,deliveredbyJeanTiroleattheLondonSchoolofEconomicson19February2009.WearegratefultoFrancescoCaselliandAugustinLandierfor

nabougratefullyacknowledgessupportfromtheCanadianInstituteforvaluablecomments.Be

Advancedresearch.JeanTirolegratefullyacknowledgesthefundingofthechair‘SustainableFinanceandResponsibleInvestment’bytheAssociationFrancaisedeGestion(AFG)atIDEI.

NOTES

nabouandTirole(2006b)andBenabou(2008a).1.Onthislastpoint,seeBe

2.Forinstance,accordingtodatafromtheSocialInvestmentForumReport(2007),citedinAndries(2008),thevalueofassetsundermanagementintheUSAthatfallintotheSRIcategorygrewatannualratesofabout12%over1995–2005and18%over2005–07;bytheendof2007,theseSRIassetsaccountedfor11%oftotalassetsunderprofessionalmanagement.Othersignsofthesametrendincludetheproliferationoffair-tradeproducts,carbonoffsetsandnewlycreatedpositionssuchasCorporateSustainabilityOfficerinmanylargecompanies.3.See,forexample,Avlonas(2009)onancientGreece.

4.SeeShavell(2002)andBesleyandGhatak(2007).Forexample,thelatterauthorsemphasizethatforCSRtobesociallybeneficial,corporationsmustbemoreefficientthanthegovernmentatproducingpublicgoods.

5.InhisreviewofwhateconomicshastocontributetoourunderstandingofCSR,Kitzmueller(2008)notesthatifethicalconsumersareaminority(majority),thegovernmentmayunder(over)regulate.

nabouandTirole(2007).6.SeeBe

7.Ofcoursetherearealsodownsidestosuchshamingpunishments.Weleavetheseforfutureresearch.8.Thoseremindersoperatenotonlyattheindividual’slevel,butalsoatgrouplevel.Withinfirmsortheirboards,inparticular,sociallyresponsibleinvestmentadvocatescanthusactassafeguardsagainstover-optimistic‘groupthink’,systematicallyaskingmanagementtopresentworst-caseordisasterscenarios,appointingdevil’sadvocates,protectingwhistleblowers,andsoon.Formoreongroupthinkandits

nabou(2008b).prevention,seeBe

rTheLondonSchoolofEconomicsandPoliticalScience2010

2010]

INDIVIDUALANDCORPORATESOCIALRESPONSIBILITY

17

9.Thisappliesatleastdomestically;givingtopoorcountriesissomewhatdifferent,inpartbecauseof

nabouandTirole2006b),andinpartdifferentattributionsmadeforthecausesoftheirpoverty(seeBe

becausesuchgivinghasmuchgreatervisibility,asittakesplaceona‘global’stage.

10.ThespecificationofpreferencesunderlyingFigure4isU¼(vaþvyy)aþmaE[va|a,y]–C(a),wherevaandvyparametrizeanindividual’sdegreeofaltruismandmarginalutilityofincome,aishislevelofcontribution,C(a)istheassociatedcost,andyistheper-unitincentiverate.E[va|a,y]isthebestestimateofvaconditionalonthechosenactionanditsreward,andmaistheintensityoftheimageconcern,commontoallagentsforsimplicity.ThejointdistributionofvaandvyistakentobeGaussian,andC(a)isquadratic.ThecurvesshowninFigure4correspondtodifferentvaluesofma.

nabouandTirole(2006a).11.SeeBe

12.Onthispointsee,forexample,HolmstromandTirole(1993).13.AsimilarconclusionisreachedbyGeczyetal.(2005).14.Ofcourse,NGOscrutinyisitselfendogenous.

15.SeeAmericanInternationalGroup,Inc.(2007,pp.4–5).

16.Amongotherinstances,Jennings(2006)relatesthat:‘from1992to2002,whileDennisKoslowskiserved

asCEO,Tycogave35milliontocharitiesdesignatedbyMr.Koslowski...Mr.KoslowskiservedontheboardoftheWhitneyMuseumofAmericanArt,and,asaresultTycodonated4.5milliontothetravellingmuseumshowsthatWhitneysponsored.’

17.Itisinterestingtonotethatthisnotionismorepopularincivil-lawcountries,whichputmoreemphasis

onstakeholderparticipationingovernanceandabroaderdutyoffirmstosociety,thanincommon-lawcountries,whereshareholdervaluehasprimacy.18.ArecententryinthisliteratureisKruger(2009),whoconductseventstudiesontheannouncementof

ratingsbyKLDResearch&Analytics,afirmthatprovidesenvironmental,socialandgovernance(ESG)researchforinstitutionalinvestors.Stockpricesfallonaverageby1%withinaweekfollowinganegativeannouncement.Ontheotherhand,theredoesnotseemtobeanabnormalreturnafterapositiveannouncement.

19.Atleastuptoapoint.BarneaandRubin(2006)arguethatsmallamountsofCSRexpenditureswill

raiseprofitability,whilelargeramountswillnot.20.Inhisdescriptivestatistics,Kruger(2009)findsthat‘irresponsiblecompanies’Fdefinedasthosefor

whichKLDrecordsmoreeventsofbadbehaviourFtendtohavebeenlessprofitableinthepastthanresponsibleones.

21.Inthelongrun,however,certaintechnologies(e.g.cleanpowermorelikelythanreforestation)may

benefitfromeconomiesofscaleandlearning-by-doing,bringingcostsbackdownandyieldsup.22.See,forexample,MaskinandTirole(2004).

REFERENCES

AMERICANINTERNATIONALGROUP,INC.(2007).PuttingOurStrengthBehindSustainability.Corporate

ResponsibilityReport.Availableathttp://media.corporate–ir.net/media_files/irol/76/76115/AIG_CRe-port_0708.pdf

ANDRIES,M.(2008).Socialresponsibilityandassetprices:istherearelationship?Mimeo,Universityof

Chicago.

ARIELY,D.,BRACHA,A.andMEIER,S.(2009).Doinggoodordoingwell?Imagemotivationandmonetary

incentivesinbehavingprosocially.AmericanEconomicReview,forthcoming.

AVLONAS,N.(2009).TheoriginsofsocialresponsibilityinancientGreece.Mimeo,AmericanCollegeof

Greece.

BARNEA,A.andRUBIN,A.(2006).Corporatesocialresponsibilityasaconflictbetweenshareholders.EFA

2006ZurichMeetings.

BARON,D.(2001).Privatepolitics,corporatesocialresponsibilityandintegratedstrategy.JournalofEconomics

andManagementStrategy,10,7–45.BENABOU,R.(2008a).Ideology.JournaloftheEuropeanEconomicAssociation,6(2–3),321–52.

FFF(2008b).Groupthink:collectivedelusionsinorganizationsandmarkets.Mimeo,PrincetonUniversity.FFFandTIROLE,J.(2006a).Incentivesandprosocialbehavior.AmericanEconomicReview,96(5),1652–78.FFFandFFF(2006b).Beliefinajustworldandredistributivepolitics.QuarterlyJournalofEconomics,

121(2),699–746.

FFFandFFF(2007).Identity,moralsandtaboos:beliefsasassets.CEPRDiscussionPaperno.6123.FFFandFFF(2010).Lawsandnorms.Mimeo.

BESLEY,T.andGHATAK,M.(2005).Competitionandincentiveswithmotivatedagents.AmericanEconomic

Review,95(3),616–36.

FFFandFFF(2007).Retailingpublicgoods:theeconomicsofcorporatesocialresponsibility.Journalof

PublicEconomics,91(9),1645–63.

rTheLondonSchoolofEconomicsandPoliticalScience2010

18

ECONOMICA

[JANUARY

BREKKE,K.A.andNYBORG,K.(2008).Moralhazardandmoralmotivation:corporatesocialresponsibilityas

labormarketscreening.ResourceandEnergyEconomics,30(4),509–26.

BURASCHI,A.andCORNELLI,F.(2003).Donations.Mimeo,LondonBusinessSchool.

CARPENTER,J.andMYERS,C.(2007).Whyvolunteer?Evidenceontheroleofaltruism,reputation,and

incentives.IZADiscussionPaperno.302.

CESPA,G.andCESTONE,G.(2007).Corporatesocialresponsibilityandmanagerialentrenchment.ECGI

FinanceWorkingPaperno.157/2007.

DANA,J.,CAIN,D.M.andDAWES,R.(2006).Whatyoudon’tknowwon’thurtme:costly(butquiet)exitina

dictatorgame.OrganizationalBehaviorandHumanDecisionProcesses,100(2),193–201.

FFF,KUANG,J.andWEBER,R.(2007).Exploitingmoralwriggleroom:experimentsdemonstratingan

illusorypreferenceforfairness.EconomicTheory,33(1),67–80.

DELLAVIGNA,S.,LIST,J.andMALMENDIER,U.(2009).Testingforaltruismandsocialpressureincharitable

giving.Mimeo,UniversityofChicago.

DEWATRIPONT,M.,JEWITT,I.andTIROLE,J.(1999).Theeconomicsofcareerconcerns.II:Applicationsto

missionsinorganizations.ReviewofEconomicStudies,66(1),199–217.

FREEMAN,R.B.(1997).Workingfornothing:thesupplyofvolunteerlabor.JournalofLaborEconomics,15,

140–66.

FRIEDMAN,M.(1970).Thesocialresponsibilityofbusinessistoincreaseitsprofits.NewYorkTimesMagazine,

13September,122–6.

FUNK,P.(2008).Socialincentivesandvoterturnout:evidencefromtheSwissmailballotsystem.Journalofthe

EuropeanEconomicAssociation,forthcoming.

GECZY,C.,STAMBAUGH,R.andLEVIN,D.(2005).Investinginsociallyresponsiblemutualfunds.Working

Paper,TheWhartonSchool,October.

GLAZER,A.andKONRAD,K.(1996).Asignalingexplanationforcharity.AmericanEconomicReview,86(4),

1019–28.

HAMMAN,J.,WEBER,R.andLOEWENSTEIN,G.(2009).Self-interestthroughdelegation:anadditionalrationale

fortheprincipal–agentrelationship.AmericanEconomicReview,forthcoming.

HEAL,G.(2005).CorporatesocialresponsibilityFeconomicandfinancialperspectives.GenevaPapers,30,387–

409.

M,B.andTIROLE,J.(1993).Marketliquidityandperformancemonitoring.JournalofPoliticalHOLMSTRO

Economy,101,678–709.

HONG,H.andKACPERCZYK,M.(2009).Thepriceofsin:theeffectsofsocialnormsonmarkets.Journalof

FinancialEconomics,forthcoming.

ISHERWOOD,C.(2007).Thegraffitiofthephilanthropicclass.NewYorkTimes,2December.

JENNINGS,M.(2006).TheSevenSignsofEthicalCollapse:HowToSpotMoralMeltdownsinCompanies...

BeforeItIsTooLate.NewYork:StMartin’sPress.

KITZMUELLER,M.(2008).Economicsandcorporatesocialresponsibility.Mimeo,EuropeanUniversityInstitute.KRUEGER,A.andMAS,A.(2004).Strikes,scabs,andtreadseparations:laborstrifeandtheproductionof

defectiveBridgestone/Firestonetires.JournalofPoliticalEconomy,112(2),253–89.

GER,P.(2009).Corporatesocialresponsibilityandtheboardofdirectors.Mimeo,ToulouseSchoolofKRU

Economics.

LACETERA,N.andMACIS,M.(2008).Socialimageconcernsandpro-socialbehavior.IZADiscussionPaperno.

3771.

LANDIER,A.andNAIR,V.(2008).InvestingforChange:ProfitfromResponsibleInvestment.Oxford:Oxford

UniversityPress.

MARGOLIS,J.andELFENBEIN,H.(2007).Dowellbydoinggood?Don’tcountonit.HarvardBusinessReview,

SocialResponsibility,SpecialIssueonHBSCentennial,86(1),19.

FFF,FFFandWALSH,J.(2007).Doesitpaytobegood?Ameta-analysisandredirectionofresearchon

therelationshipbetweencorporatesocialandfinancialperformance.WorkingPaper,RossSchoolofBusiness,UniversityofMichigan.

MASKIN,E.andTIROLE,J.(2004).Thepoliticianandthejudge:accountabilityingovernment.American

EconomicReview,94,1034–54.

MAZAR,N.,AMIR,O.andARIELY,D.(2008).Thedishonestyofhonestpeople:atheoryofself-concept

maintenance.JournalofMarketingResearch,45,633–4.

MONIN,B.(2007).Holierthanme?Threateningsocialcomparisoninthemoraldomain.InternationalReviewof

SocialPsychology,20(1),53–68.

FFFandMILLER,D.(2001).Moralcredentialsandtheexpressionofprejudice.JournalofPersonalityand

SocialPsychology,81(1),33–43.

rTheLondonSchoolofEconomicsandPoliticalScience2010

2010]

INDIVIDUALANDCORPORATESOCIALRESPONSIBILITY

19

FFF,SAWYER,P.J.andMARQUEZ,M.J.(2008).Therejectionofmoralrebels:resentingthosewhodothe

rightthing.JournalofPersonalityandSocialPsychology,95(1),76–93.

ORLITZKY,M.,SCHMIDT,F.andRYNES,S.(2003).Corporatesocialandfinancialperformance:ameta-analysis.OrganizationStudies,24(3),403–41.

PIGOU,A.(1920).TheEconomicsofWelfare.London:Macmillan.

REINHARDT,F.,STAVINS,R.andVIETOR,R.(2008).Corporatesocialresponsibilitythroughaneconomiclens.

FEEMWorkingPaperno.84.2008.

SHAVELL,S.(2002).Lawversusmoralityasregulatorsofconduct.AmericanLawandEconomicsReview,4(2),

227–57.

SOCIALINVESTMENTFORUM(2007)ReportonSociallyResponsibleInvestingTrendsintheU.S.Availableat

http://www.socialinvest.org/resources/research

rTheLondonSchoolofEconomicsandPoliticalScience2010