Who Deserves Help Evolutionary Psychology Social Emoti - PDF document

WhoDeservesHelp?EvolutionaryPsychology,SocialEmotions,andPublicOpinionaboutWelfare pops_883395..418 MichaelBangPetersenAarhusUniversityDanielSznycerUniversityofCalifornia,SantaBarbaraLedaCosmidesUniversityofCalifornia,SantaBarbaraJohnToobyUniversityofCalifornia,SantaBarbara Evidencesuggeststhatourforagingancestorsengagedinthesmall-scaleequivalentofsocialinsuranceasanessentialtoolofsurvivalandevolvedasophisticatedpsychologyofsocialexchange(involvingthesocialemotionsofcompassionandanger)toregulatemutualassistance.Here,wehypothesizethatpoliticalsupportformodernwelfarepoliciesareshapedbytheseevolvedmentalprograms.Inparticular,thecompassionatemotivationtosharewithneedynonfamilycouldnothaveevolvedwithoutdefensesagainstopportunistsinclinedtotakewithoutcontributing.Cognitively,suchparasiticstrategiescanbeidentiÞedbytheintentionalavoidanceofproductiveeffort.Whendetected,thispatternshouldtriggerangeranddown-regulatesupportforassistance.Wetestedpredictionsderivedfromthesehypothesesinfourstudiesintwocultures,showingthatsubjectsÕ KEYWORDS:evolutionarypsychology,publicopinion,socialwelfare,socialemotions,socialexchangeWhenindividualsformopinionsaboutsocialwelfare,aprimaryconcerniswhetherwelfarethebeneÞtstheyreceive(Cook&Barrett,1992;Gilens,1999;Iyengar,1991;Larsen,2006;Petersen,Slothuus,Stubager,&Togeby,2011;Sniderman,Brody,&Tetlock,1991,chap.5).Indecidingwhetherrecipientsdeservewelfare,individualspayattentionprincipallytotherecipientsÕeffortsinalleviatingtheirownneed(Gilens,1999;Oorschot,2000).Ifwelfarerecipientsareseenasabletowork,butpreferringnotto(i.e.,theyareÒlazyÓ),theyareperceivedasundeservingandwelfareisopposed.Incontrast,ifwelfarerecipientsareseenasunluckyvictimsofexternalcircumstances,theyareperceivedasdeservingandwelfareissupported.While 0162-895X©2012InternationalSocietyofPoliticalPsychologyPublishedbyWileyPeriodicals,Inc.,350MainStreet,Malden,MA02148,USA,9600GarsingtonRoad,Oxford,OX42DQ,andPOBox378CarltonSouth,3053Victoria,Australia strongevidencehasbeenproducedthatestablishesanempiricallinkbetweenwelfareopinionsandjudgmentsofrecipientsÕeffort,extantresearchlacksempiricallywell-supportedexplanationsforwhyandhowthesejudgmentssostronglycolorwelfareopinions(cf.Oorschot,2006;Petersen,Existingresearchhasfocusedontwogeneralexplanationsoftheroleofdeservingnessjudg-mentsinpublicopinionaboutwelfare.First,afewresearchershaveattemptedculturallyorinsti-tutionallyspeciÞcexplanations.Thus,Gilens(1999,p.63)linkstheimportanceofdeservingnessjudgmentstotheindividualisticcultureofAmericans.Incontrast,RothsteinarguesthatafocusonrecipientsÕdeservingnessisfosteredininstitutionalcontextswithmeans-testedwelfareprograms(Rothstein,1998).Theparticularisticnatureoftheseexplanations,however,rendersthemincon-sistentwiththeavailableevidenceaboutthetransnationaldistributionofthesephenomena.Analy-sesofcross-culturaldatafromtheWorldValuesSurveyshowthattheperceptionthatpovertyiscausedbylazinessÑi.e.,alackofmotivationtoputineffortÑisauniversaldriverofoppositiontogovernmenteffortstoreducepoverty(seeAppendix).Dataisavailablefor49countriesfromallpartsoftheworldandinallbutonecountry,theeffectisintheexpecteddirectionandsigniÞcant.For39ofthecountries,theeffectsizeisabove.2andmostoftensubstantiallyso.Alargenumberofthesecountriesarenotnotablyindividualistic,nordotheyhavethekindsofmeans-testedwelfareinstitutionspinpointedasresponsiblebyRothstein(1998).Perceptionsofdeservingnessarealsocentralinexplainingdifferencesbetweennationsinsupportforwelfare.Thatis,muchofthecross-nationalvariationinpublicsupportforwelfare(and,indeed,inactualsocialspending)canbeexplainedbycross-nationaldifferencesinperceptionsofwelfarerecipientsaslazy(Alesina,Glaeser,&Sacerdote,2001;Larsen,2006).Basedonthisevidence,thereisacasethatperceptionsofwelfarerecipientsÕmotivationtoworkareauniversaldriverofpersonalandpublicsupportforwelfareacrossdifferentnationsanddifferentwelfaresystems.Thesecondapproachtoexplainingthedependenceofwelfareopinionontheperceivedeffortofrecipientshasfocusedonindividual-levelexplanationsthatlinkthejudgmentstohigher-orderreasoningstructuressuchasideology.Indeed,severalstudieshavedocumentedthatpeopleonthepoliticalrightwing(i.e.,conservatives)aremorelikelytoattributewelfarerecipientsÕneedtolaziness,whileliberalsaremorelikelytoviewrecipientsÕneedastheproductofexternalcir-cumstances(seeSkitka&Tetlock,1993;Skitka,Mullen,GrifÞn,Hutchinson,&Chamberlin,2002).BasedonsuchÞndings,researchershavearguedthatdeservingnessjudgmentsaresome-whateffortfulcognitiveprocessesundertakentosupportpreexistingideologicalcommitments(e.g.,Skitkaetal.,2002).Whilethisproposalishighlyplausible,severalobservationssuggestthatthepreoccupationwiththeeffortofneedyindividualsisgroundedinpsychologicalprocessesthatpreexistideology.First,ifpeopleengageindeservingnessjudgmentsfromsomeculturallyspeciÞcideology,thenpeoplewhohavedifferentoropposingideologiesoughtnottoprovideparalleljudgments.Yettheydo.Inarecentstudy,Petersenetal.(2011)demonstratedthatwhileegalitar-iansandnonegalitariansmightdisagreeintheabstractaboutwelfarerecipientdeservingness,ideologicaldifferencesvanishwhenaskedtojudgethedeservingnessofspeciÞcwelfarerecipi-ents.Inachievingthiseffect,deservingnessjudgmentswereshowntooperateinanautomatedfashion,pickingupcuesandinformingwelfareopinionseffortlessly(Petersenetal.,2011).More-over,researchinpsychologyondeservingnessjudgmentsshowthatindividualsacrossculturesspontaneouslyjudgethemotivationsofneedyindividualsinallkindsofeverydayinteractionsÑfromlendingexamnotestofellowstudentstohelpingadrunkenpersoninthesubway(Weiner,1995).Theimportanceofdeservingnessjudgments,then,isnotconÞnedtoissues(suchaswelfare)drawnfromthedomainofpoliticalideology(seealsoPetersen,2012).Suchresultscastdoubtonclaimsthatsuchjudgmentsaresolelytheproductsofculturallylearnedideologies.Indeed,theyraisethequestionofwhetherelementsofanunderlyinghumanuniversalpsychologymightbeparticipatingaswell.Petersenetal. WeproposethatanapproachthatdrawsonrecentÞndingsinevolutionarypsychologyandhunter-gathererstudiescanhelpexplainwhypeoplespontaneouslyconnectsupportforwelfaretowelfarerecipientsÕeffort.AccumulatingevidencefromevolutionarypsychologyandneuroscienceindicatesthathumannatureÑouruniversal,reliablydevelopingpsychologicalarchitectureÑincludesanarrayofevolvedcognitiveandemotionprogramstailoredbynaturalselectiontosolverecurrentadaptiveproblemsfacedbyourgroup-livingancestors(Sell,Tooby,&Cosmides,2009;Tooby&Cosmides,2008;Tooby,Cosmides,Sell,Lieberman,&Sznycer,2008).Hereweexploretheimplicationsof(1)thehypothesisthatthesocialemotionsofangerandcompassionweredesignedbynaturalselection,inpart,toregulatewhetherandtowhatextentwewanttohelpaneedyperson,and(2)thehypothesisthattheseemotionprogramsareembeddedinasystemofcognitivemechanismsthatcollectivelyimplementalogicofsocialexchangethatevolvedtoadvantageouslymanagemutualassistanceamongourancestorsinsmall-scaleforaginggroups.Fromthistheoreticalframework,wearguethatthepervasiveeffectofperceptionsofwelfarerecipientsÕeffortregardingworkonsupportforwelfarearisesbecausetheseperceptionsÞttheinputsystems(i.e.,resemblethetriggers)thatthetwosocialemotionprograms,angerandcompassion,aredesignedtomonitorandrespondto.Below,weßeshoutthisargumentandthenreportaseriesoffourstudiesconductedtotestpredictionsdrawnfromthistheory.InStudy1,wedemonstratethatwelfareopinionsareinfactpowerfullyshapedbysocialemotions.Whenactivated,theseemotionprogramsinßuencenotonlywelfareopinionsbutalsotheeasewithwhichtheyareformed.InStudy2,weprovideevidencethatangerandcompassionmediatethelinkbetweeneffortcuesandattitudesaboutwelfareandthattheydosoindependentlyoftheideologyoftheobservers.InStudy3weshowthattheÞtbetweenangerandcompassionontheonehandandperceptionsofwelfarerecipientsÕeffortontheotherarehighlyspeciÞc.Thatis:(1)itisspeciÞcallyperceptionsaboutwelfarerecipientsÕeffortratherthanothertypesofperceptionsthatregulateangerandcompassion;and(2)perceptionsofwelfarerecipientsÕeffortregulatetheactivationofangerandcompassionratherthanemotionssuchasanxiety,con-tempt,anddisgust.Study4demonstratesthatthisÞtbetweeneffortperceptionsandangerandcompassionisrobustacrosstwohighlydifferentcountriesandwelfaresystems:theUnitedStatesandDenmark.OuraimistoprovideevidenceforthehypothesesthatangerandcompassionconstituteacoresetofthroughwhichperceptionsandcuesofwelfarerecipientsÕeffortstranslateintoopinionsonsocialwelfare;and,second,thatthesemechanismsoperateinsimilarwaysforindi-vidualsacrossideologiesandcountries.Inthatregard,ouraiminthisarticledivergefromtraditionalgoalssuchaspredictingindividualdifferencesinwelfareopinionsorexplainingwhysomepeoplebelievethatwelfarerecipientsarelazyandothersbelievetheyareunlucky.Rather,weseektoelucidatethesharedpsychologicalmechanismsthatareresponsibleformediatingbetweenthetwo.Suchananalysispresupposesratherthanunderminestheimportanceofotherkindsofanalyses.InunderstandingwhetheraspeciÞcindividualperceiveswelfarerecipientsaslazyornot,oneneedstorelyonanalysesof,forexample,thecontentoftheindividualsÕideology(Skitka&Tetlock,1993),theagendaofthemedia(Gilens,1999),thestructureofpoliticalinstitutions(Larsen,2006),andthelevelofethnicdiversityintheindividualÕscountry(Alesinaetal.,2001).Ourcontributionhereistoilluminatethepsychologicalmechanismsthatprocesssuchindividualandcontextualfactorsandcausethemtoinßuencewelfareopinions.Wereturntothisthemeinthediscussion.EvolvedSocialEmotionsandPublicOpinionResearchinevolutionarypsychologysuggeststhatemotionsarenotvagueandcrudefeelingsandurgesasintuitionsuggests(Tooby&Cosmides,2008).Rather,ouremotionsconstituteanarrayofdistinctandsophisticatedinformation-processingmechanisms,eachdesignedbynaturalselectionWhoDeservesHelp? tosolvespeciÞcproblemsfacingourancestors(Petersen,2010;Selletal.,2009;Toobyetal.,2008).Formillionsofyears,ourancestorslivedinsmall-scalegroups(Alford&Hibbing,2004;Cosmides&Tooby,2006;deWaal,1989),andsomeemotionsÑthesocialemotionsÑhavedesignfeaturesthatevolvedforsuccessfullysolvingrecurrentadaptiveproblemsofgrouplivingÑsuchassharing,exploitation,coalitions,powerrelations,hierarchy,collectiveaction,punishingnormviolatorsandmanagingintergrouprelations(Petersen,2009).Thesocialemotionofcompassion,forinstance,seemstobedesignedformotivatinginvestinginsocialpartnersinneed(Goetz,Keltner,&Simon-Thomas,2010;Petersen,Sell,Tooby,&Cosmides,2010).Theevolvedfunctionofanothersocialemotion,anger,istodefendagainstexploitationandbargainforbettertreatment.ItistriggeredwhenotherpeopleplacetoolowvalueononeÕswelfare(Selletal.,2009).Inthisway,angerisalsodistinguishedfromthetwootherwiserelatedemotionsofcontemptanddisgust(Ekman,2004;Rozin,Lowery,Imada,&Haidt,1999).Angerisdesignedtoincreasethesocialinvestmentsofthetargetofanger,andhenceitmotivatesapproach(Petersenetal.,2010;Toobyetal.,2008).Incontrast,contemptanddisgustmotivateavoidance.Forexample,thesameneuralcircuitsareactivatedinmoraldisgustasindisgustforcontaminatingmatter(e.g.,blood,excrement)and,inparallel,bothkindsofdisgustmotivateavoidanceoffurthercontactwithitstarget(Rozin,Haidt,&McCauley,2000).Incontrasttobothangeranddisgust,contemptoperatesprimarilyinthedomainofstatus,andtheelicitationofcontemptfunctionstocommunicatethatthetargetisoflowerstatusandtofacilitateavoidanceacrosslevelsofsocialhierarchies(Rozinetal.,1999).Hence,whileanger,contempt,anddisgustareallimportanttosocialinteractionandfacilitatetreatingtheobjectinanegativemanner,eachservesindependentevolvedfunctionsandhencemotivatesdifferentactsandtrade-offs.Theexistenceofmultiple,qualitativelydistinctsocialemotionprogramscarrymajorimplica-tionsforpoliticalscienceresearch.Wesuggestthatmanymodernpoliticalissues,suchaswelfare,taxpayments,criminalsanctions,redistribution,revolution,immigration,andracerelationscontainbasicdilemmasthatourancestorsevolvedtodealwithinordertosuccessfullynavigatesocialrelationships(Alford&Hibbing,2004;Cosmides&Tooby,2006;Petersen,2009,2012).Byimplication,themindofmoderncitizensisendowedwithatoolboxofspecializedmechanisms,includingthesocialemotions,thatcouldassistandfacilitatetheirpoliticaldecisionmaking.Yet,thesecognitiveandemotionprogramsweredesignedinancestralenvironmentstorespondtocuesthatwerepredictiveinthoseenvironments.Consequently,theyshouldbeactivatedorinactivatedwhenexposedtocuesthatmimictheseancestralcuesÑwhetherornotthesecuesarerationallyrelevantinevolutionarynovelsituationssuchasmodernmasspolitics.Itiswithinthisgeneralframeworkweseektoelucidatetheunderpinningsofdeservingnessjudgmentsinwelfareopinions.AncestralSubsistence,NaturalSelection,andtheLogicofSocialExchangeModernwelfareinstitutionsentailredistribution,i.e.,thetransferralofwealthfromthemosttotheleastwell-off.Inthissense,modernwelfarestatesarelarge-scalesocialexchangesystems.YetsocialexchangeisfarfromarecentWesternculturalinvention.Anthropologicalstudiesoflivingforagershaveuncoveredcomplexandpervasivesystemsofsharingbothwithinandbetweenfamilies(Kaplan&Gurven,2005).Takentogether,hunter-gathererstudies,paleoanthropologicalevidence,Inunderstandingthestructureofthehumanmindanditsconstituentmechanisms,ancestralconditionsareemphasizedbecausethemajorityofourspeciesonlygaveupforagerlifeafewthousandyearsago(Rindos,1987).Thisistoolittletimeforselectiontoengineercomplexspecies-typicaladaptationstonovelpostforagerconditions(instead,oneobservesgeographicallylimitedresponsestointenseselectionpressuressuchaslocaldiseaseslikemalariaorfoodslikemilk,involvingsmallnumbersofalleles,orsmallrandomlydispersedgeneticdifferencesattheindividuallevel;seeTooby&Cosmides,1990).Hence,whateverspecies-typicalpsychologicalmechanismsexistevolvedinresponsetolifeinthepreagriculturalenvironment.Petersenetal. andprimatologicalevidencesupporttheviewthatourancestorshavebeenengaginginsocialexchangeforhundredsofthousandsormillionsofyears(Cosmides&Tooby,1992).SocialExchangeintheContextofForagingOuradaptationsforsocialexchangeevolvedtooperateinaworldofsmallforaginggroups.Importantly,theforagingnichethatancestralhumansoccupiedinvolvedtheexploitationoflargegame,ahigh-quality,nutrient-dense,large-packagedfoodresource(Kaplan,Hill,Lancaster,&Hurtado,2000).SuchresourcesaredifÞculttoacquire,and,hence,hunter-gatherersregularlyexperiencedhighvarianceinhuntingsuccessduetochance,illness,orotheradversity(Hill&Hawkes,1983;OÕConnell,Hawkes,&Jones,1991;Sugiyama,2004).Suchinterruptionsintheßowofcaloriesposedanacuteadaptiveproblemforourancestors(Kaplanetal.,2000).Atthesametime,huntingsuccesseswouldoftenprovidemorenutrientsthanasingleindividualorhisfamilycouldconsumeatonetime.Whilethecapacitytostoreexcessfoodwassharplylimited,thisenabledstorageintheformofsharingandthroughtheimpositionofreciprocalsharingobligationsfromthoseonehadsharedwith(Cosmides&Tooby,1992;Lee&DeVore,1968).Formalmodelingindicatesthatsuchformsof-personsocialexchangecanbeanadaptivesolutiontotheresourcevarianceproblemandbufferriskamongforagers(Axelrod&Hamilton,1981;Cosmides&Tooby,1992,2006;Kameda,Takezawa,Tindale,&Smith,2002;Kaplan,Hill,&Hurtado,1990).Consistentwiththis,anumberofstudieshaveshownthathumanforagers(and,perhaps,nonhumanprimates)sharearesourcetoagreaterextentiftheacquisitionoftheresourceissubjecttorandomvariance(Cashdan,1980;deWaal,1996;Kaplan&Hill,1985).Ancestrally,sharinginsituationsofresourcevarianceproducedaveragenetgainstoparticipantsbecauseresourceswereshiftedfromthosewithsmallmarginalreturnsonadditionalresources(becausetheyhadmore)tothosewithgreatermarginalreturns(becausetheyhadless).Evenforthosewhohavemoreatanyonetime,suchredistributivestrategieswouldbeadaptiveifreversalsofconditionoccurwithsufÞcientfrequency,andifanexchangeisindeedreciprocal,i.e.,ifoverthelongrunthosewhoreceivealsogive(Cosmides&Tooby,1992;Trivers,1971).CheatersasanAdaptiveProblemOnesigniÞcantchallenge,inthelatterrespect,isthatpracticingsharingexposessharerstoopportunisticexploitationbythosethatreapthebeneÞtsofothersÕproductiveeffortswithoutincurringthecostsofcontributing(herecalled)(Cosmides&Tooby,1992;Price,Cosmides,&Tooby,2002).ThisincreasestheÞtnessofthecheaterbutdiminishestheadaptivevalueofsocialexchangeand,hence,givesrisetoaselectivecounterpressurefortheevolutionofmecha-nismsfordetectingcheaters,redirectionofassistanceawayfromcheaters,andresponsesdesignedtorecalibratethemotivationsofcheaters.Thus,socialexchangeentailsanevolutionaryarmsracebetweencooperatorsandcheaters.Thisentailsthepredictionthatthehumanmindiswell-adaptedtodetectingandreactionagainstcheaters.Andaspredicted,30yearsofexperimentshaveshownthatthehumanmindincludesreasoningspecializationsfordetectingcheatersinsocialexchanges,i.e.,forsolvingtasksinvolvingtheidentiÞcationofindividualswhotakesbeneÞtswithoutpayingtherequiredcosts(Cosmides&Tooby,2005).Insupportoftheviewthatthesespecializationsareadaptations,researchershavedemonstratedtheirexistencecross-culturally,includinginsmall-scalesocieties(Sugiyama,Tooby,&Cosmides,2002);thatperformanceappearsjustasgoodinculturallyunfamiliarasculturallyfamiliarcontexts(Cosmides&Tooby,2005);thatcapacitiesforcheaterdetectionisfoundinveryyoungchildren(Harris,Nunez,&Brett,2001);andthatsuchcapacitieshavedistinctneuralunder-pinnings(Stone,Cosmides,Tooby,Kroll,&Knight,2002).Importantly,thedetectionofcheatersWhoDeservesHelp? alsohasthepredictedbehavioraleffects.Hence,evidencefromexperimentaleconomicgamesshowsthatpeopleceasetocontributetoapublicgoodifothersdonÕtfollowsuit(see,e.g.,Fehr&GŠchter,2000).Similarly,observationalstudiesofforagers(Kaplan&Gurven,2005;Kaplan&Hill,1985)indicatethatfoodsharingamongnonkinistoasigniÞcantdegreereciprocal,i.e.,conditionalinthesensethatAshareswithB,ifBshareswithA.Indiscriminatingbetweencheatersandnoncheaters,evolutionaryanalysissuggeststhatourmindshavebeendesignedtoespeciallyattendtocuesofothersÕmotivationtotakepartinthesystemofsocialexchange,i.e.,theirwillingnesstoaccrueandexchangeresources(Cosmides,Barrett,&Tooby,2010;Delton,Cosmides,Guemo,Robertson,&Tooby,2012).Suchcuesinclude,forexample,theeffortsspentwhenaccruingresourcesorthegratitudeexpressedwhenreceivingbeneÞts.Theobviousalternativecue,actualforagingsuccess,isinferiorfortworeasons.First,therandomvariationinforagingsuccessmakesitdifÞculttogaugewhetherothersÕlackofhuntingsuccessstemsfromloweffortorbadluck.Second,studiesoflivingforagersshowthatindividualsdifferintheirforagingcompetence(Kaplan&Gurven,2005),and,hence,somehaveconsistentlylesssuccessthanotherswithoutthisnecessarilybeingtheresultsofparasiticmotivations.Anumberofstudiesdocumentthathighfoodproducersobtain,forexample,morematingopportunitiesandgreateroffspringsurvivorshipandhenceseemtoberepaidinothercurrenciesthanfood(Kaplan&Gurven,2005).Hence,fromtheperspectiveofahighproducer,itcanpayofftosharewithlowproducersbut,again,thisisonlythecaseifthelatteraremotivatedtoputineffortandreciprocate(e.g.,inothercurrencies).Inlinewiththesearguments,experimentalstudiesshowthathumansrepresentbadoutcomeswithdifferentmentalcategoriesdependingonwhetherthoseoutcomesareattributabletoincompe-tenceorlackofmotivation(Deltonetal.,2012).Similarly,arangeofstudiesinneurosciencehaveprovidedevidencefortheimportantroleofintentionsincheaterdetection(foranoverviewofthisresearch,seePetersen,Roepstorff,&Serritzlew,2009).Forexample,fMRIstudieshavedemon-stratedthatcheater-detectiontasksengagedistincttheory-of-mind-relatedneuralcircuits(i.e.,cir-cuitsinvolvedingaugingtheintentionsofothers),whicharenotengagedbyotherlogicallyequivalenttasks(seeErmer,Guerin,Cosmides,Tooby,&Miller,2006).TheEvolvedFunctionofAngerandCompassionintheContextofSocialExchangeGiventheselectionpressuresactingonsocialexchange,wesuggestthatthetwosocialemo-tions,compassionandanger,arecentrallyinvolvedincarryingoutthemotivationalagendamandatedbythestrategyofsharing.Conditionalsharingisasocialinvestmentstrategy,andÑasdescribedaboveÑcompassionandangerregulatesocialinvestments.SpeciÞcally,theneedforinsuranceagainstgapsintheßowofcalories(e.g.,insituationsofillness)madeitimportanttocultivateandinvestinpotentiallyvaluablesocialrelationships(Tooby&Cosmides,1996).Inthefaceofneedynoncheaters(individualswhoareableandwillingtoreciprocateonfutureoccasions),sharingistheadaptiveresponse.Giventhemotivationalfunctionsofcompassion,weexpectcompassiontofacilitatethisresponsetowardsnoncheaters.Inthefaceofcheaters(individualswithparasiticmotivations),however,theadaptiveresponseentailsavoidingsharingbutalsoattemptstorecalibratethecheaterÕsmotivationalsystemtobemorecooperative.Thelatterisimportant.Inancestralsmall-scalegroups,therewouldbeonlyalimitednumberofpotentialvaluablesocialrelationships,andhenceitwouldbeimportantnotjusttodismissstrategiccheatersbuttorecalibratethemsothattheybecomebettercooperators.ThisstronglysuggeststhatthedetectionofcheatersinasharingsituationshouldtriggerangerÑtheemotiondesignedtodefendagainstexploitationandincentivizetheup-regulationofinvestmentsbyothers(e.g.,Selletal.,2009)Ñratherthanthemoreavoidance-orientedemotionsofcontemptanddisgust.Onlyinseverecasesshouldindividualsfeelcompelledtoshunthetargetaltogether.Consistentwiththis,anumberofanthropologicalaccountsreportthat,whileindividualswhodonotsharesometimesareostracized,theyareallowedreentryinthePetersenetal. communityiftheirsharinglevelsincrease(Kaplan&Gurven,2005).Moregenerally,cross-culturalstudiesofeverydaymoralityshowthatangerandcompassionareinfactregulatedbyeffort-relatedperceptionsinthefaceofneedyindividuals.Hence,Weiner(1995)reportsstudiesfromtheUnitedStates,Canada,andJapanthatconsistentlyshowthatsubjectsrespondwithhighlevelsofangerandlowlevelsofcompassiontoalackofeffortamongindividualsrequestinghelp.Predictions:EmotionsandWelfareRecipientsasPotentialCheatersCitizensÕresponsestothepoliticalissueofwelfare,weargue,areatleastpartlyshapedbythisinheritedpsychology.Totheevolvedmind,assistanceprovidedtowelfarerecipientsisacueofpossibleexploitationbyunproductiveothers.Onthistheory,welfarejudgmentsshouldautomaticallyrecruittherelevantpsychologicalmachineryandarriveatadecisionaboutwhetherhelpingislegitimate,basedontheperceivedcooperativeintentionsofwelfarerecipientsÑasifindividualswereinfactpersonallyengagedwithwelfarerecipients(Petersen,2012).Asdemonstratedbytheliteratureondeservingnessjudgmentsandwelfareopinionsandourowninitialanalysis(seeAppendix),thisisinfactwhathappensincountryaftercountry.Anevolutionaryframeworkprovidesanaturalexplanationforwhyperceptionsoflazinessregulatejudgmentsaboutwhetherrecipientsdeservewelfare.Moreimportantly,theevolutionaryframeworkallowsustomovebeyondcurrentknowledgebyelucidatingthepreciserolethatsocialemotionssuchasangerandcompassionplayinthispsychology.Fromthistheory,asdescribedindetailbelow,wederive14predictions.ThepredictionsarelistedinTable1.Ifwelfareissuesareprocessedbyemotionprogramsthatevolvedtoregulatesocialinteractions,welfareopinionsshouldbeinßuencedmorestronglybysocialemotionsthanbyotherlessrelevantkindsofemotions.SpeciÞcally,weexpectfeelingsofcompassiontowardswelfarerecipientstoincreasesupportforwelfareandredistribution(Prediction1),andfeelingsofanger,contempt,anddisgusttodecreasesuchsupport(Prediction2).Incontrast,priorresearchonemotionsinpoliticalsciencehas,ingeneral,focusedonnonsocialemotionssuchasanxiety(seeMarcus,Neuman,&MacKuen,2000;Neuman,Marcus,Crigler,&MacKuen,2007;Redlawsk,2006).Currentevidencesuggeststhatanxietyevolvedasaprecautionaryprogramtomotivatevigilanceandpreparationtowardhazardsofdiversekinds(Fiddick,2004;Petersen,2010)and,hence,islessrelevanttosocialTable1.OverviewofPredictions #Prediction1Feelingsofcompassiontowardwelfarerecipientsincreasesupportforwelfare.2Feelingsofanger,contemptanddisgusttowardswelfarerecipientsdecreasesupportforwelfare.3Feelingsofanxietywhenthinkingaboutwelfarerecipientshavelittleornoeffectonsupportforwelfare.4Feelingsofanger,contempt,anddisgusttowelfarerecipientsmakeopinionformationonwelfareissuesfaster.5Feelingsofcompassionforwelfarerecipientsmakeopinionformationonwelfareissuesfaster.6Feelingsofanxietywhenthinkingaboutwelfarerecipientsdonotmakeopinionformationonwelfareissuesfaster.7Ideologicalpredispositiondoesnotinßuencethespeedwithwhichindividualsformopinionsonwelfare8Welfarerecipientswithlittlemotivationtolookforworkelicitanger.9Welfarerecipientsmotivatedtolookforworkelicitcompassion.10Theactivationofangerpartiallymediatestheeffectofeffortcuesonsupportforwelfare.11Theactivationofcompassionpartiallymediatestheeffectofeffortcuesonsupportforwelfare.12Compassionandangermediatetheopinioneffectsofeffortcuesindependentlyofpoliticalideology.13Cuesofeffortregulateanger(andcompassion)ratherthananger-relatedemotionssuchasanxiety,contempt,anddisgust.14Angerandcompassionareregulatedbycuesofeffortratherthanbyeffort-relatedcuesofcompetence. WhoDeservesHelp? welfareifprocessedbyourevolvedlogicofsocialexchanges.Tocorroboratethisargument,wecontrasttheeffectsofthesocialemotionswiththeeffectofanxietythatwepredicttobesmallinthecontextofwelfareopinions(Prediction3).Convergingresultsfromseveralareasofresearchsuggeststhatevolveddedicatedcircuitstendtooperatemorerapidlyandeffortlesslywithintheirnaturaldomainthandoacquiredskillsordedicatedcircuitsusedoutsidetheirdomain(Ermeretal.,2006;New,Cosmides,&Tooby,2007).Forthisreason,weexpectsocialemotionswillinßuencewelfareopinionsinwaysthataredistinctfromotheremotionsandopinionfactors.Forexample,ifgeneratingsupportfororoppositiontoassistanceisanevolvedfunctionoftheseemotions,thentheactivationoftheseemotionsispredictedtofacilitatesuchjudgmentsÑthatis,activationshouldmakewelfarejudgmentsfastandintuitive.SpeciÞcally,weexpectthatcitizensfeelingaversion(anger,contempt,ordisgust)toorcompassionforwelfarerecipientsshouldrespondfasterwhenaskedfortheirwelfareopinions(Predictions4and5).Incontrast,neithertheprecautionaryemotionofanxiety(seealsoMarcusetal.,2000;Prediction6)nortraditionalfactorsofopinionsuchasideology(Prediction7)shouldfacilitateopinionformationinthisfashion.Predictions8and9testourcoreargument:thatangerandcompassiontowardswelfarerecipi-entsaretriggeredbycuesoftheabsenceorpresenceofeffortinalleviatingtheirownorothersÕneed.Hence,weexpectwelfarerecipientscharacterizedaslackingworkmotivationtoelicitanger,andwelfarerecipientswithsuchmotivationtoelicitcompassion.Indeed,theactivationofcompassionandangershouldmediateeffectsofthesecuesonsupportforwelfare(Predictions10and11).Furthermore,weexpectcompassionandangertomediateeffortcuesindependentlyofpoliticalideology(Prediction12).Asnotedintheintroduction,Skitka,Tetlock,andcolleaguesshowedthatpoliticalideologyhasanimpactonwelfareopinions(Skitka&Tetlock,1993;Skitkaetal.,2002).Yet,ifitisthelogicofsocialexchange,builtintothestructureoftheangerandcompassionsystems,thatmakesmoderncitizenssensitivetoeffortcues,weshouldexpecttheiroperationstobeinde-pendentoftheeffectsofpoliticalideology.Conservativesandliberalsareexpectedtosharethesamespecies-typicalmentalarchitectureand,hence,shouldbeangeredorfeelcompassionatebyexposuretothesameancestrallyrelevantcues.Thatis,whileconservativesandliberalsdisagreeintheabstractaboutwelfarerecipients,thesegeneraldifferencesshoulddropinimportancewhenindividualsacrossthepoliticalspectrumareprovidedwiththesamecues,aslongasthesecuesÞttheinputconditionsoftheemotionalprograms(foramoreextendeddiscussion,seePetersen,2009).Asdiscussed,previousstudiesÑmostnotably,Weiner(1995)andFeather(2006)Ñhavepre-dictedandfoundrelationshipsbetweeneffort-relatedperceptionsandfeelingstowardstheneedyinpersonaleverydaysituationsofhelp-giving.Inthatperspective,amainempiricalcontributionofthepresentstudyistoshowthatmodernindividualsalsoenterandprocesstheevolutionarilynovelphenomenaofimpersonalmasspoliticsasifitwere,infact,apersonalexchangeofhelp,usingthesamesetofhighlystructuredcognitiveandemotionalsystems.Moreover,theadaptationistapproachallowsustoprovideevenmorespeciÞchypothesesontherelationshipbetweeneffortperceptionsandemotionsthanthoseofferedbyWeiner(1995)andFeather(2006).First,theargumentthatemotionalsensitivitytoeffortcuesreßectsadaptivestrategiesdesignedtofacilitateinvestmentsinreciprocalsocialexchangesallowedustopinpointangerasthedistinctiveaversiveemotiontriggeredinthefaceoflack-of-effortindividualsrequestinghelp.Feather(2006,p.46)arguesthatlackofefforttriggersfeelingsofresentment.Thisincludesangerbutcouldalsoincludeothernegativeemotionssuchascontempt.Similarly,whileWeiner(1995)doesfocusmorespeciÞcallyonanger,hedoesnotdifferentiateclearlybetweenangerandotheraversiveemotions(e.g.,heoccasionallysubstitutesdisgustforangerinhisanalysesandnevercomparestheeffectsofsuchdistinctemotions;seep.156).Here,wepredictthatthesensitivitytoeffortcuesisspeciÞctoangerandcompassion,comparedtoothernegativeemotionssuchasanxietyandevenmoreanger-relatedemotionssuchascontemptanddisgust(Prediction13).Petersenetal. Second,bothFeather(1999,p.41)and,inparticular,Weiner(1995)focusonthecontrollabilityoftheneedasamainfactorintheregulationofangerandcompassiontowardstheneedy.SpeciÞ-cally,lackofeffort(acontrollablecauseofneed)andlackofability(anuncontrollablecauseofneed)arejuxtaposedaskeyelicitorsofangerandcompassion,respectively.Theevolutionaryperspective,however,leadstodifferentpredictions.Ifwearedesignedtodirectsocialinvestmentstowardsneedyindividualswhowillreciprocateonfutureoccasions,uncontrollablebutchroniclowlevelsofabilityintheformofincompetenceshould,ifanything,decreaseourwillingnesstohelp.Atthesametime,however,suchadecreaseshouldnotbedirectlymediatedbyangerorcompassion.Hence,iftheroleofangerandcompassioninthefaceofrequestsforhelpis,ontheonehandtorecalibratethemotivationsofstrategiccheatersand,ontheotherhandtodirectinvestmentstowardscooperativelymotivatednoncheatersthenthecompetenceoftheindividualshouldbeirrelevantfortheirexecutionasarguedinthediscussionontheroleofactualforagingsuccessincheaterdetection.Whereafocusoncontrollabilityleadstotheexpectationthatpeoplefeelcompassionatetowardsanincompetentindividualinneed,we,inotherwords,predictthatangerandcompassiontowardwelfarerecipientsarespeciÞcallyregulatedbyperceptionsoftheircooperativeintentions(asreßectedintheireffortstoÞndwork)ratherthanbyperceptionsoftheircompetence(Prediction14).Aswewilldemonstrate,themoreavoidance-orientedsocialemotionsdo,however,pickupcuesaboutcompetence.OverviewofStudiesWetestourpredictionsinfourstudies.Studies1Ð3wereconductedinDenmarkandStudy4wasconductedintheUnitedStates.BoththeUnitedStatesandDenmarkarewealthydevelopeddemoc-raciesbut,atthesametime,theydifferonanumberofimportantdimensions.TheUnitedStatesisalargeethnicandraciallyheterogeneouscountrywithasmallwelfarestatebasedonmeanstestingandahighlyindividualisticculture.Incontrast,DenmarkisasmallhomogenouscountrywithacomparativelylargewelfarestatewherewelfarebeneÞtsareuniversallyprovidedratherthangrantedonthebasisofmeanstesting.Inthisway,acomparisonoftheinßuenceofsocialemotionsonsocialwelfareopinionsamongDanesandAmericansgivesusimportantinformationabouttheuniversalityoftheunderlyingmechanismsacrossimportantculturaldifferences.Study1isdesignedtotestthepredictionsregardingthegeneralimportanceofsocialemotionstowelfareopinions(Predictions1Ð7).Study2isanexperimentalstudydesignedtotestPredictions8Ð12and,hence,establishthevalidityofthecoreargumentthatourfocalemotions,angerandcompassion,mediatetheopinioneffectofeffortandmotivationcues.Predictions13and14testthefunctionalspeciÞcityoftherelationshipbetweenangerandcompassion,respectively,andpercep-tionsofworkmotivation.ThevalidationoftheseÞnalpredictionsisimportantforestablishingtheevolutionaryinformedproposalthateachemotionrepresentsafunctionallydistinctandwell-designedinformation-processingsystem.Duetotheirimportance,weseektoestablishthattheyarecross-culturallyvalid.Hence,Study3isbasedonaDanishsample,andStudy4isbasedonaU.S.sample,butotherwisethetwostudiesuseexactlythesamemeasurestotestPredictions13and14.Study1DesignandMeasuresAmongasampleofDanishcitizens,asurveyonwelfareopinionswascollectedovertheinternetinthefallof2008.ThesampleconsistedofDanesagedbetween18and70andisbasedonaOnthecontrary,fromanadaptationistperspective,temporarylossesofability,e.g.,incapacityduetoillness,shouldincreasethemotivationtohelp,asitindicatesthattheneedwasnottheproductoflazinessÑthatis,notduetoalackofeffort.WhoDeservesHelp? representativequotasampledesignedtoachievenationalrepresentativenessonsex,age,andgeo-graphicallocation.ThenumberofÞnalsubjectswas1,537,andtheresponseratewas34%.TotestPredictions1Ð3,i.e.,whethersocialemotionsinßuencewelfareopinions,subjectswerepresentedwithabatteryofsixLikert-scaleditemsaboutredistribution,welfare,andtheroleofthestate.Subjectswereaskedtoindicatedisagreementwiththefollowingstatementsona7-pointscalerangingfromÒCompletelyAgreeÓ(0)toÒCompletelyDisagreeÓ(6):ÒHighincomesshouldbetaxedmorethaniscurrentlythecaseÓ(reversecoded),ÒWeshouldresistthedemandsforhigherwelfarebeneÞtsfrompeoplewithlowincomes,ÓÒThewealthyshouldgivemoremoneytothosewhoareworstoffÓ(reversecoded),ÒThegovernmentspendstoomuchmoneyontheunemployed,ÓÒThestatehastoolittlecontroloverthebusinessworldÓ(reversecoded),andÒInpolitics,oneshouldstrivetoassurethesameeconomicconditionsforeveryone,regardlessofeducationandemploymentÓ(reversecoded).Theanswerstotheseitemswereaddedtogethertoformascaleofsupportforthewelfarestaterangingfrom0to1(TogetherwiththesubjectsÕresponsestothewelfareitems,responsetimesfortheitembatteryasawholewerecollected.Inpsychology,responsetimeisastandardmeasureofthedegreeofprocessingusedtoformaresponseand,hence,thecollectionofresponsetimesallowsustotestPredictions4Ð7,thatis,whethertheactivationofsocialemotionsmakesopinionformationeasier.However,usingresponsetimescollectedovertheweb(comparedtothelaboratory)entailssomecomplications(seePetersenetal.,2011).VariabilityintraveltimethroughtheInternetintroducessomenoiseintothesemeasures,and,hence,measurementswillbeinßuencedby,forexample,thespeedofthesubjectÕsInternetconnection.Tocompensate,wethereforeusedrankedresponsetimesratherthantheexactresponsetimeinmilliseconds(i.e.,thefastestresponsetimewasassignedavalueof1,thesecond-fastestwasassignedavalueof2andsoforth).Therankinghasbeenrescaledfrom0(fastestresponsetime)to1(slowestresponsetime).Tomeasureemotionalreactionstowelfarerecipients,subjectsÕwereaskedÒHowdoyoufeel,whenyouhearorreadaboutpeopleonsocialwelfare?Óandpresentedwithalistwithsevenemotions:anger,disgust,contempt,compassion,sympathy,anxiety,andfear.Subjectswereaskedtoanswerona7-pointscalewiththeendpointslabeledÒNotatallÓandÒVerystrongly.ÓFollowingMarcus,MacKuen,Wolak,andKeele(2006),threeemotionalscaleswerecreated.Aswedonotexpectdifferencesintheireffectshere,anger,disgust,andcontemptwereputtogetherinascalemeasuringtheintensityofaversivesocialemotions(0.89).Compassionandsympathywereputtogetherinscalemeasuringtheintensityofcompassionatesocialemotions(0.82).Finally,thetwoprecautionaryemotions,anxietyandfear,wereputtogetherinananxietyscale(0.78).Allscaleshavebeencodedfrom0(lowintensity)to1(highintensity).Asbasiccontrolvariables,wemeasuredthesubjectsÕsex,age,andeducation.Ageismeasuredinyears,whilelevelofeducationhasbeencodedfrom0to1.Finally,asameasureofthegeneralideologyofthesubjects,weaskedthemtoplacethemselvesonan11-pointpoliticalleft-rightscale.Thismeasureofideologyhasalsobeenscaledfrom0to1.Table2presentsthetestsofPredictions1Ð3.Predictions1and2entailthataversiveandcompassionatesocialemotionsstronglyinßuencewelfareopinions.InModel1,welfareopinionsareregressedonthethreebasiccontrolvariables(sex,age,andeducation)andthethreeemotionscales.Aspredicted,weÞndstrongandsigniÞcanteffectsofthetwoscalesofsocialemotionalThewholebatterywaspresentedtosubjectsonasinglescreenduringthewebsurvey.Thecollectedresponsetimesconstitutethetimelapsefromwhenthesubjectentersthisscreenuntilthesubjecthascompletedallitemsandmovestothenextscreen.Petersenetal. reactionstowelfarerecipients.Withallvariablesscaledfrom0to1,thecoefÞcientof.19inthecaseofaversivesocialemotionsimpliesthatashiftfromnoaversivefeelingstostrongaversivefeelingsdecreasesupportforwelfarebyaboutone-Þfthofthefullscale.Inthecaseofcompas-sionatesocialemotions,thecoefÞcientof.30impliesthatashiftfromnocompassionatefeelingstostrongcompassionatefeelingsincreaseswelfaresupportbyaboutone-thirdofthefullscale.AsexpectedfromPrediction3,althoughthetwosocialemotionscalesstronglyinßuencewelfareopinions,theeffectoftheprecautionaryanxietyscaleisnotsigniÞcant.ThispatternofresultsprovidesstrongsupportfortheideathatwhenmodernpoliticalproblemsresembleancestralsocialproblemsÑasisthecasewiththewelfareissueÑtheactivationoftherelevantemotionsprovidesconsiderableguidancewhencitizensformopinions.Model2includesthepredictorsofModel1plusthesubjectsÕgeneralideology.SupportingtheSkitka,Tetlock,andcolleaguesproposal,wealsoÞndthatideologyisapowerfulpredictorofwelfareopinionsÑindeedstrongerthanotherfactors(Skitka&Tetlock,1993;Skitkaetal.,2002).Nonetheless,theeffectsofbothaversiveandcompassionateemotionsremainsigniÞcantaftertheinclusionofthisclassicexplanatoryfactor.Whilebothsocialemotionsandideologypredictwelfareopinions,theevolutionaryperspectivesuggestthattheyoperateindifferentwaysintheopinion-formationprocess.Socialemotionsareevolvedpsychologicalsystemsdesignedtofacilitatesocialresponsesthatwouldhavebeenadaptiveancestrally.Incontrast,theideologiesofleftandrightarerelativelyrecent,culturallyelaboratedconstructswhosecomplexspeciÞcsmustbememorized.Thesetwoprocessesofopinionformation(triggeringexistingcircuitsversusreasoningfrommemorizeddatastructures)arefundamentallydifferentinkind.Thus,Predictions4Ð7entailthatsocialemotionsnotonlyshapethecontentofsubjectsÕwelfareopinionsbut(unlikeideology)themoretheyareactivated,themorerapidlytheyshouldorganizewelfareopinions.Thus,increasingintensityofemotionalreactionstowelfarerecipientsshouldbeassociatedwithlowerresponsetimeswhenansweringquestionsaboutwelfareTable2.TheEffectofEmotionalReactionstoWelfareRecipientsandIdeologyonOpiniononWelfareIssues(Models1and2)andResponseTimesWhenFormingWelfareOpinions(Models3and4) WelfareOpinionWelfareOpinionResponseTimeResponseTimeModel1Model2Model3Model4Intercept.41***.21***.47***.46***(.03)(.02)(.04)(.05)Sex(male).03**.00.03.03(.01)(.01)(.02)(.02)Age.002***.003***.002***.002***(.000)(.000)(.001)(.001)(.02)(.02)(.03)(.03)AversionScale(.03)(.03)(.03)(.05)CompassionScale.29***.19***(.02)(.02)(.03)(.03)AnxietyScale.05.02.02(.03)(.03)(.05)(.05)IdeologyÐ.45***.04.001(.02)(.03)(.15)Ideology*IdeologyÐÐÐR2(adj.).25.46.02.02 Notes.N1356.UnstandardizedOLSregressioncoefÞcients,standarderrorsinparentheses.Allvariables,exceptage,varybetween0and1.Ageismeasuredinyears.*.05,**.01,***WhoDeservesHelp? opinions.Ideologyshouldnot.Totestthisprediction,Model3inTable2regressessubjectsÕresponsetimestowelfareopinionsonthedifferentexplanatoryvariables.Aspredicted,weseesigniÞcantnegativeeffectsofthetwosocialemotionsscalesonresponsetimes.Themoreintenselysubjectsfeelaversiveorcompassionateemotionstowardswelfarerecipi-ents,thefastertheyrespondtoquestionsaboutwelfare.Incontrast,therearenoeffectsofeitheranxietyorideologyonresponsetime.Onecould,forexample,havearguedthatthewelfarestatehasbeenthecentralpoliticalprojectoftheleft,and,therefore,thosebelongingtotheideologicalleftshouldbeabletoanswerquestionsaboutwelfarerapidly.Asrevealedinthemodel,thisisnotthecase.Therelationshipbetweenideologyandresponsetimecan,however,alsobemodeleddiffer-ently.Hence,asthewelfarestatehasbeenafocalissueinthepoliticalconßictbetweenthetwopoliticalwings,onecouldarguethatwhatshouldmatteristheintensityofoneÕsideologicalcommitment(i.e.,ratherthanwhetheronebelongstotheleftortheright).Thispredictsacurvilinearrelationshipbetweenideologyandresponsetimesuchthatsubjectswithhigherabsoluteideologyscores(whetherleftorright)wouldanswerwelfareopinionquestionfaster.InModel4,weincludeaquadratictermtomodelacurvilineareffect.ThistermisinsigniÞcant.Insum,Study1demonstratesthatopinionsonwelfareissuesareassociatedwiththeintensityofsocialemotionsbutnotofprecautionaryemotions(anxiety).Socialemotionscontributetowelfareopinionformationintwoways.First,socialemotionspullwelfareopinionsinoneortheotherdirection.Compassionateemotionstowardswelfarerecipientsincreasesupportforwelfareandredistribution,whileaversiveemotionsdecreasesupport.Second,theactivationlevelofsocialemotions,butnotofprecautionaryemotions,isassociatedwithfasterresponsetimesinwelfareopinions.Socialemotionprogramswereengineeredbyselectiontomotivateorinhibitassistance,andregulatingsupportforwelfareisthereforeamoderncaseofitsevolvedfunction.Anxietywasnotancestrallyrelevanttothistask,andsohasnoeffectonit.Hence,itseemslikelythatwhenevolvedcircuitsaretriggeredbyrelevantinputs,theytendtocarryouttheirfunctionsrapidly.Weconcludethatwhenemotionsproducetheoutputsforwhichtheyaredesigned,theyfunctioninwaysthatarequitedifferentfromhowtraditionalopinionfactors(suchasideology)operate.Whileideologyisapowerfulpredictorofwelfareopinions,citizensdonotmakeinferencesfromideologywiththesameeaseastheydofromtheirmoralfeelings.Study2Study2testswhetherperceptionsofwelfarerecipientsÕcooperativeintentionsfacilitateopinionformationbecausetheseperceptionsÞtinputsystemsthatangerandcompassionevolvedtomonitor.DesignandMeasuresWedesignedanexperimentinwhichsubjectsarepresentedwithaseriesofshamwelfarerecipientswhoarecharacterizedasvaryingintheirefforttoalleviatetheirunemployment.WethenmeasuredthesubjectsÕopinionstowardstherecipient,aswellastheirfeelingsofangerandHere,wefocusonindividualsÕself-perceivedideologicalposition.Anotherapproachtoassessingideologyentailsmeasur-ingtheextenttowhichindividualshaveadeepunderstandingofwhattheirpositionentails(cf.Converse,1964).Giventhatsophisticatedideologuescouldhaveautomatedpartsoftheiropinionformationprocessonpoliticalissues(cf.Schreiber,2007),itispossiblethatwewouldÞndeffectsonresponsetimeusingsuchameasure.Inthesurvey,thismeasureisnotavailablenoristhetraditionalmeasureofpoliticalsophistication,politicalknowledge,available.Theeffectsare,however,controlledforeducationÑanoften-usedproxyforsophistication(cf.Snidermanetal.,1991).Theeffectofeducationisnegative(i.e.,themoreeducatedrespondfaster),buttheeffectissmallandinsigniÞcant.Petersenetal. TwohundredandsevenDanishundergraduatesinpoliticalsciencecompletedanexperimentwiththreeconditions.Ineachcondition,thesubjectswerepresentedwithaspeciÞcrecipientofsocialwelfare,butantecedentinformationwasmanipulatedacrossconditionssuchthattherecipientsvariedinwhethertheirneedwasattributableeithertobadluckortolaziness.Inthecontrolcondition,subjectswereaskedtoÒImagineamanwhoreceivessocialwelfarebeneÞts.ÓIntheconditionwiththeunluckyrecipient,subjectswereaskedtoÒImagineamanwhoreceivessocialwelfarebeneÞts.Hehasalwayshadaregularjob,buthasnowbeenthevictimofawork-relatedinjury.Heisverymotivatedtogetbacktoworkagain.ÓFinally,intheconditionwiththelazyrecipient,subjectswereaskedtoÒImagineamanwhoreceivessocialwelfarebeneÞts.Hehasneverhadaregularjob,butheisÞtandhealthy.Heisnotmotivatedtogetajob.ÓToobtainanopinionmeasure,subjectswereaskedwhethertheyagreedordisagreedwiththefollowingstatement:ÒTheactivationrequirementsshouldbemadestricterforhim.ÓInDenmark,recipientsofsocialwelfarehavetherightandobligationtotakepartinjobactivationprogramssuchasjobtraining.Byfocusingonwhethertheserequirementsshouldbemadestricter,weareessen-tiallyaskingwhethertherecipientshouldpaygreatercostsinreturnforhiswelfarebeneÞts(seealsoPetersenetal.,2011).Toobtainmeasuresofemotions,subjectswereaskedtoagreeordisagreewiththetwofollowingstatements:ÒIoftenfeelangertowardspeoplelikehim,ÓandÒIoftenfeelcompassiontowardspeoplelikehim.ÓToobtainameasureofthesubjectsÕpoliticalideology,subjectswereaskedtoagreeordisagreeon5-pointscaleswiththreestatements:ÒThestatehastoolittlecontroloverprivateinvestments,ÓÒInpolitics,oneshouldstrivetoprovideallwiththesameeconomicconditions,ÓandÒHighincomesshouldbetaxedmoreheavilythaniscurrentlythecase.ÓTheanswerswereaddedtogethertoformascale(0.76)suchthathighervaluesindicatesamoreliberal(i.e.,egalitarian)ideology.Allvariablesarescaledbetween0and1andallt-testsareone-sidedasallperformedt-testsrelatetodirectionalhypotheses.UsingOLSregression,weinvestigatedwhethertheemotionalmeasuresmediatetherelationshipbetweencuestotheeffortoftherecipientandopinionsaboutwelfare(seeTable3).InModel1,weregresssupportforstricteractivationrequirementsontheexperimentalconditions.Asreferencecategory,weusethecontrolcondition,and,hence,themodeltestswhetherthelazyandtheunluckyrecipient,respectively,aretreateddifferentlyfromthisneutralrecipient.Consistentwithpreviousstudies,model1demonstratesthatcuesofeffortaffectopinionstowardsarecipient.SubjectsaresigniÞcantlymoresupportiveoftighteningthelazyrecipientÕsrequirementsandsigniÞcantlylesssupportiveoftighteningtheunluckyrecipientÕsrequirements.InModels2Ð4,wetestwhetherthiseffectismediatedbyfeelingsofangerandcompassion.InModels2and3,weregressangerandcompassion,respectively,ontheexperimentalconditions.AspredictedbyPredictions8and9,subjectsfeelmoreangerandlesscompassiontowardsthelazyrecipientandlessangerandmorecompassiontowardstheunluckyrecipient.Theimportantquestion,then,iswhethertheseaffectiveresponsesmediatetheopinioneffectsoftheexperimentalmanipu-lations.InModel4,weseethattheydo.Hence,whencontrollingforourtwosimplemeasuresofemotions,weremoveanydirecteffectsoftheÒlazyrecipientÓmanipulationonopinion.Hence,theeffectchangesfrom.20andhighsigniÞcanceinmodel1to.09andinsigniÞcanceuponinclusionoftheemotionmeasures.Thiscontrolalsoremovesaboutone-thirdoftheeffectoftheÒunluckyrecipientÓmanipulation(i.e.,theeffectchangesfrom.22to.16).Formaltests(Sobeltests)ofmediationcorroboratetheseconclusions.Hence,angersigniÞcantlymediatestheopinioneffectoftheÒlazyrecipientÓmanipulation(.001)andtheÒunluckyrecipientmanipulationÓWhoDeservesHelp? .01).Similarly,compassionsigniÞcantlymediatestheopinioneffectsoftheÒlazyrecipientÓmanipulation(.03)andtheÒunluckyrecipientÓmanipulation(.04).TheseÞndingssupportPredictions10and11.Iftheseeffectsindeedarerootedintheemotionsofangerandcompassion,thenweshouldexpecttheemotionalreactionsofindividualstobesensitivetotheexperimentalmanipulationsregardlessoftheindividualsÕideologicalviews(Prediction12).WhetherthisisthecaseistestedinModels5Ð7.Here,thepoliticalideologyofthesubjectstogetherwithtwo-wayinteractionsbetweenpoliticalideologyandtheexperimentalmanipulationsareregressedonsupportforstricterrequire-mentsandthefeelingsofangerandcompassion.Hence,thesemodelstestwhetherindividualsacrossthepoliticalspectrumreactanydifferentlytothethreemanipulations.Model5revealsthatwhileideologyhasalargeeffectinthecontrolcondition(cf.themaintermforideology),thiseffectissubstantiallyandsigniÞcantlyreducedinthetwoexperimentalconditionswheresubjectsarepre-sentedwithspeciÞcandecologicallyvalideffortcues(cf.thecoefÞcientsandsignsoftheinteractionterms).Infact,furtheranalysesshowthattheeffectofideologyisinsigniÞcantinthetwotreatmentgroups(lazyrecipient:.14;unluckyrecipient:Models6and7suggestthatareasonwhyideologydoesnotconditionopinionsinthefaceofspeciÞccuesisthatangerandcompassionareactivatedinsimilarwaysinsubjectsacrosstheideologicalspectrum.Hence,inbothmodels,weseethattheinteractiontermsareinsigniÞcant.Thisindicatesthatforsubjectsacrossthepoliticalspectrum,thedifferencesinfeltangerandcompassiontowardsthelazyandunluckyrecipients,ontheonehand,andthecontrolcondition,ontheotherhand,arejustthesame.TheseÞndings,inotherwords,supportPrediction12.Table3.AngerandCompassionasMediatorsofEffortCues VariableSupportforAngerCompassionSupportforSupportforAngerCompassionModel1234567Intercept.57***.23***.60***.58***.80***.35***.39***(.03)(.03)(.04)(.06)(.06)(.06)(.06)ExperimentalManipulationLazyRecipient.20***.26***.20***.09.04.34***(.05)(.04)(.05)(.05)(.08)(.08)(.08)UnluckyRecipient.11*.15**17*.22**(.05)(.04)(.05)(.04)(.08)(.07)(.08)AngerÐÐÐ.34***ÐÐÐCompassionÐÐÐ.15*ÐÐÐIdeologyÐÐÐÐ.30**.53**(.12)(.11)(.12)Ideology*LazyRecipientÐÐÐÐ.40*(.17)(.16)(.18)Ideology*UnluckyRecipientÐÐÐÐ.64***.13(.17)(.16)(.18)(adj.).30.28.21.42.37.37.33 Notes.N199.UnstandardizedOLSregressioncoefÞcients,standarderrorsinparentheses.Thecontrolconditionisthereferencecategorytowhichtheexperimentalmanipulationsarecompared.Theexperimentalmanipulationsareenteredasdummyvariablesmeasuringwhetherthesubjecthas(1)orhasnot(0)beentreatedwiththegivenmanipulation.Allvariablesvarybetween0and1..05,**.01,***Petersenetal. Insum,thisstudydemonstrates,Þrst,thattheopinioneffectsofcuesaboutwelfarerecipientsÕeffortare(atleastpartly)mediatedbyfeelingsofangerandcompassion.ThisisconsistentwithWeinerÕs(1995)studiesinpsychologyshowingthatsimilarcuesfeedintoangerandcompassioninordinarysocialinteraction.Here,weshowthatthisrelationshipholdsalsointhedomainofsocialwelfare.Asexpected,whenthecuessurroundingmasspoliticsÞttheinputconditionsoftheemotionprogramsthatguideoureverydaybehavior,theseemotionsstartprovidingguidanceinpolitical-opinionformation.Eveninimpersonalsituationssuchasmasspolitics,ourevolvedemotionsareengagedwheneverancestrallyrelevantcuesarepresent.Second,wehavedemonstratedthatthecausaleffectofthesecuesonemotionsexistsirrespectiveofindividual-leveldifferencesinpoliticalideology.Inthisregard,itshouldbenotedthatthesamplewasbasedonpoliticalscienceunder-graduates.Suchasampleshouldbeparticularlypronetobeingaffectedbytheirideologicalcom-mitments,iftherewereeffectstobefound.Giventhis,theseÞndingsstronglysuggestthatcitizensÕpreoccupationwithwhetherwelfarerecipientsarelazyornotdonotemergefromideologicalconcernsbutfromtheevolvedstructureofouremotionalsystems.Studies3and4InourtwoÞnalstudies,Studies3and4,wedemonstratethatthelinkbetweenangerandcompassionandperceptionsofwelfarerecipientsÕmotivationtoworkarespeciÞcintwosenses.First,wedemonstratethatperceptionsofwelfarerecipientsÕlazinesselicitangerratherthananxiety,contempt,anddisgust(Prediction13).Second,wedemonstratethatfeelingsofangerandcompas-sionareregulatedbyperceptionsofwelfarerecipientsÕworkmotivationratherthancompetence(Prediction14).Althoughearlierstudieslinkeddeservingnessjudgmentsandemotions,thisresultsupportsourproposalaboutthespeciÞcroleofangerandcompassionintheprocessofopinionformation.Furthermore,theclaimthatamechanismorsetofmechanismsareadaptations(typically)requiresevidenceofuniversality.Studies3and4aredesignedtotestPredictions13and14cross-culturally.Usingthesameexactmeasures,Study3isbasedonaDanishsample,whileStudy4isbasedonaU.S.sample.DesignandMeasuresThedataforStudy3werecollectedaspartofthesurveyusedinStudy1.Hence,thebasisforStudy3isanapproximatelyrepresentativenationallyrepresentativesampleofDanes(n1,537).InStudy4,asurveywasconductedamong274undergraduatesfromaU.S.universityinthespringofTotestpredictions13and14,weneedmeasuresofsubjectsÕperceptionsofwelfarerecipientsÕeffortandcompetenceandmeasuresofsubjectsÕemotionalreactionstowelfarerecipients.Tomeasureperceptions,weadaptedstandardmeasuresofstereotypicalbeliefsfromtheAmericanNationalElectionStudies.PerceptionsofwelfarerecipientsÕeffortweremeasuredusingthefollow-ingquestion:ÒNowwehavesomequestionsabouthowyouperceivepeopleonsocialwelfare.Inyouropinion,aremostpeopleonwelfaremakinganeffortoraretheylazy?ÓSubjectswereaskedtoprovidetheiropinionona7-pointscalewiththeendpointslabeledÒMakinganeffortÓ(0)andÒLazyÓ(6).Similarly,perceptionsofwelfarerecipientsÕcompetenceweremeasuredbyasking:ÒInyouropinion,aremostpeopleonwelfareintelligentorunintelligent?ÓA7-pointscalewasusedwithendpointslabeledÒIntelligentÓ(0)andÒUnintelligentÓ(6).ThisfocusonintelligenceasameasureofchroniclevelsofabilityisinlinewithWeinerÕs(1995,p.31)operationalization.Bothmeasureswerereversedsuchthathighscoresindicatethatwelfarerecipientsareperceivedashighineffortandhighincompetence,respectively.WhoDeservesHelp? Tomeasureemotions,weprovidedsubjectswithalistofemotionsandaskedÒHowdoyoufeelwhenyouhearorreadaboutpeopleonsocialwelfare?ÓHerewefocusonanger,compassion,anxiety,disgust,andcontempt.Aswepredictthatdifferentemotionswillberegulatedbydifferentperceptions,weassesseachemotionseparately,incontrasttostudy1.Subjectsansweredon7-pointscaleswithendpointslabeledÒNotatallÓandÒVerystrongly.ÓWebeginbycontrastingangerandcompassionwithanxiety,untilnowthefocalemotioninmuchofpoliticalscienceresearch.Ifthetheoreticalargumentisvalid,perceptionsofwelfarerecipientsÕeffortshouldregulatefeelingsofangerandcompassiontowardswelfarerecipients,butnotanxiety.Figure1(PanelsAandB)displaystheserelationshipsfortheDanishandU.S.samples.Thesimilaritiesbetweenthesetwocountriesarestriking.Inbothsamplesthezero-ordercorrelationsbetweenangerandcompassionontheonehand,andperceptionsofeffortontheotherhand,arelargeandhighlysigniÞcant.ThosecorrelationsremainhighwhenonestatisticallyremovestheeffectofperceptionsofwelfarerecipientsÕcompetence.InbothDenmarkandtheUnitedStates,thepercep-tionthatwelfarerecipientsarelazyactivatesanger,whiletheperceptionthattheyaremotivatedtoalleviatetheirownneedactivatescompassion.InneitherDenmarknortheUnitedStatesisanxietystronglyassociatedwithperceptionsofeffort.Whileangerandanxietyaresimilarlyvalenced,theeffectofperceptionsofeffortisspeciÞctoanger.Hence,initialsupportisprovidedforPredictions13and14. Figure1.TheeffectofperceptionsofwelfarerecipientsÕeffortandcompetenceonanger,anxiety,andcompassiontowardswelfarerecipients.Zero-orderandpartialcorrelations(controlledfortheotherperception).Notes.N1421(Study3)/266(Study4).ThepartialcorrelationscontrolforperceptionsofwelfarerecipientsÕeffort.StarsindicatewhethercorrelationissigniÞcantlydifferentfromzero(two-tailedtests).*.05,**.01,***Petersenetal. WithregardtoPrediction14,however,theimportantquestioniswhetheritisspeciÞcallyperceptionsofeffortandmotivationthatfeedintoangerandcompassion,orwouldanytypeofnegativeorpositiveperceptionhavethesameimpact?Inparticular,ourargumentsfocusonpercep-tionsofthecompetenceofwelfarerecipients.OtherresearchershavearguedthatuncontrollablecausesofneedÑsuchaschroniclowlevelsofabilityintheformofincompetenceÑtriggercom-passion.Yet,evolutionaryanalysisandanthropologicalevidencesuggestthatwhentagginganindividualasacheateroranoncheater,hiscompetenceisalesspredictivecuethanhiseffort.Therefore,iftheroleofangerandcompassioninwelfareopinionsareunderwrittenbyanevolvedpsychologyforsocialriskbuffering,weshouldexpectperceptionsofcompetencetoregulatefeelingsofangerandcompassion.Figure1(PanelsCandD)displaystheserelationshipsfortheDanishandU.S.samples.InbothsamplesweÞndsigniÞcantbutmoderatezero-ordercorrelations:Theperceptionsofcompetencearepositivelycorrelatedwithcompassionandnegativelycorrelatedwithanger.IfanythingÑandincontrasttopredictionsfromacontrollabilityperspectiveÑincompetencereducespeopleÕswilling-nesstohelp.Importantly,however,inourdatasetstheperceptionsofeffortandcompetencearehighlyintercorrelated(Danishsample:.000;U.S.sample:.000).WhenwestatisticallyremovetheeffectofperceptionsofwelfarerecipientsÕeffort,thesecorrelationsaregreatlyreduced,andallbutonebecomeinsigniÞcant(theremainingsigniÞcantcorrelationisbetweencompassionandperceptionsofcompetenceintheDanishsample,wherethecoefÞcientof.05juststayssigniÞcantduetothelargesamplesize).FurtherstatisticaltestscorroboratetheseÞndings.Inbothsamples,theeffectofperceptionsofeffortissigniÞcantlystrongerthantheeffectofperceptionsofcompetenceonbothanger(Study3:.000;Study4:11.13,.000)andcompassion(Study3:.000;Study4:.002).InneithersampledoweÞndanysigniÞcantcorrelationbetweenfeelingsofanxietyandperceptionsofTheseÞndingsprovidestrongsupportforPrediction14.Acrossthetwosamples,angerisspeciÞcallyactivatedbyneedyindividualswhoarenotmotivatedtoputineffortandreciprocate,whilecompassionisspeciÞcallyactivatedbyneedyindividualswithcooperativemotivations.Itis,inotherwords,notanykindofnegativeorpositiveperceptionthatregulatestheseemotions,norisitdirectlyamatterofcontrollability.Asshownabove,perceivingwelfarerecipientsaslazydoesnottriggeranxiety.Whilemuchpriorresearchinpoliticalsciencehasfocusedonthevalenceofemotions,thisÞndingcorroboratesotherstudiesshowingthatemotionsneedtobedistinguishedbeyondvalence(seePetersen,2010).Thequestion,however,ishowfarthisprocessofdistinctionandcharacterizationneedstogo?Recentstudiesinpoliticalsciencehave,forexample,distinguishedbetweenaversiveemotions(suchasanger,contempt,anddisgust)ontheonehandandanxietyontheotherhand(Marcusetal.,2006).However,ifdistinctemotionsevolvedtosolvedistinctadaptiveproblems,evencloselyrelatedemotionssuchasanger,contempt,anddisgustareexpectedtoshowsharplydifferentiatedactivationpatternsandoutputsinresponsetodifferentclassesofevents.Giventhelargepotentialgainsofcooperationandthelimitednumberofavailablesocialpartnersintheancestralsocialenvironmentofhumans,evolutionaryanalysissuggestsasequenceofcounter-measuresthathumansshouldfollowwhenconfrontedwithsomeonewhoisexploitive.TheÞrstresponsevis-ˆ-vischeaterswouldhavebeentoattempttorecalibratetheircooperativedispositionsupwards(Selletal.,2009).Ifthatfailed,withdrawalfromcooperativearrangementsandshunningwouldensue.Recalibrationofcooperativedispositions(e.g.,effort)isthedomainofangerratherthantheavoidance-motivatingemotionsofdisgustandcontempt(Rozinetal.,1999;Selletal.,2009).Thus,Prediction13entailsthateffortperceptionsarespeciÞcallylinkedtothemobilizationofanger.Figure2comparesthecorrelationsbetweenanger,disgust,contempt,andcompassiontowardswelfarerecipientsontheonehandandperceptionsofwelfarerecipientsÕeffortontheother.Ascan411WhoDeservesHelp? beseen,thezero-ordercorrelationsbetweeneffortperceptionsandcontemptanddisgustarehighlysigniÞcantandareaboutaslargeasthecorrelationsbetweentheseperceptionsandangerandcompassion.However,feelingsofanger,disgust,andcontemptarehighlyintercorrelatedinboththeDanish(average.73)andtheU.S.sample(average.52).ThegraybarsinFigure2showwhathappenswhenonestatisticallyremovestheeffectsoftheotheremotionsandofcompetenceperceptions:Althoughitsmagnitudedecreases,thecorrelationbetweenperceptionsofeffortandanger(controllingforcontempt,disgustetc.),remainshighandsigniÞcantinboththeDanishandtheU.S.samples.Inthecaseofcontemptanddisgust,incontrast,mostoftheeffectsdisappearaftercontrollingfortheotheremotions.OnlythecorrelationbetweenperceptionsofeffortanddisgustintheDanishsampleremainssigniÞcant,butitseffectisweak(partial.08).Statisticaltestscorroboratetheseconclusions.InbothsamplesthecorrelationbetweenangerandperceptionsofeffortissigniÞcantlystrongerthanthecorrelationsbetweentheseperceptionsandcontempt(Study.001;Study4:.05)anddisgust(Study3:.02;Study4:.04),respectively.Theeffectofcompassionis,generallyspeaking,unaffectedbycontrolfortheotheremotions.TheseÞndingssupportPrediction13.Asexpected,therearecross-culturallyrobustlinksbetweenperceivinganeedyindividualaslazyandexperiencingangertowardsthisindividual,aswellasbetweenperceivinganeedyindividualasunluckyandexperiencingcompassiontowardsthisindividual.Perceptionsofeffortdonotmobilizedisgustandcontemptdirectly,althoughtheseemotionscanbeco-activatedwithanger.Alackofcooperativeeffortinthetargetmightjustbeoneinaseriesofcuesthatneedtobepresentbeforedisgustandcontemptaretriggeredandthetargetisshunnedaltogether.Infact,furtheranalysesshowthat,acrossthetwosamples,thereisaconsistenttwo-wayinteractioneffectbetweenperceptionsofeffortandperceptionsofcompetenceontheavoidance-orientedemotionofcontempt(Study3:.004;Study4:Substantively,thissigniÞcantinteractioneffectexpressesthatifwelfarerecipientsareseenaslazy,contemptincreasesstronglywiththeperceptionthattheyalsoareincompetent.If,however,welfarerecipientsareseenascooperativelymotivated,competencejudgmenthasnoeffectoncontempt. Study 3 – Danish Sample –.47***–.44***–.44***.36***–.20***–.08**–.04.28***–0.5–0.3–0.1AngerDisgustContemptCompassionEffect SizeStudy 4 – US Sample –.44***–.42***–.30***.43***–.27***–.04–.10.36***–0.5–0.3–0.1AngerDisgustContemptCompassion Zero-Order Correlations Partial Correlations Figure2.TheeffectofperceptionsofwelfarerecipientsÕeffortonfeelingsofanger,contempt,disgust,andcompassiontowardswelfarerecipients.Zero-orderandpartialcorrelations.Notes.N1421(Study3)/266(Study4).Thepartialcorrelationshavebeencontrolledfortheeffectsofthethreeotheremotions(e.g.,thecorrelationbetweeneffortperceptionsandangerhasbeencontrolledforcontempt,disgust,andcompassion)aswellasperceptionsofwelfarerecipientsÕcompetence.StarsindicatewhethercorrelationissigniÞcantlydifferentfromzero(two-tailedtests).*.05,**.01,***Petersenetal. Hence,whenneedyindividualsareneithermotivatedtoreciprocatehelpnorhavevaluablecompe-tences,theysufferastronglossofsocialrespectÑpresumably,becausetheyareoflowvalueascooperativepartners.Evolutionarypsychologistsandhunter-gathererresearchershavedevelopedmultiple,converg-inglinesofevidencethatsupporttheviewthatsocialexchangefunctionedamongourancestorsasasocialinsurancestrategythroughwhichindividualscouldguardagainstinterruptionsinthefoodsupplyduetoinjuryorbadluck.Bysharingwithothers,individualsinvestinfuturehelp.Bymakingthatsharingconditionalonwhetherpotentialrecipientsweredisposedtocontributewhentheycould,sharersprotectedthemselvesagainstexploitation.TherecurrentpayoffstoconditionalcooperationÑextendingoverhundredsofthousandsofyearsÑselectedforpsychologicalmecha-nismsinourspeciesthatreliablyguidedourancestorstoimplementthiswinningstrategy.Onthisview,thesocialemotionsofangerandcompassionevolved,inpart,tomotivatetheseinvestmentOurresultsaccordwiththeviewthatmoderncitizensformpoliticalopinionsusinganarrayofemotionprogramsthatevolvedtoprocessandsolveancestralsocialproblems.Herewefoundthatangerandcompassionarecausallyimplicatedintheformationofwelfareopinions,whencitizensencountercuesthattheseemotionprogramsevolvedtoprocess.ToriskoversimpliÞcation,justaswehaveevolvedspecializationsthatcauseustofearsnakesandspiders,weevolvedspecializationsthatmakeusangryatthelazybutcompassionatetowardtheneedy.Inthisway,withevolvedcognitiveandemotionprogramsastheintermediatelink,theconÞgurationofpastadaptiveproblemsisresponsibleforstructuringaspectsofpublicopiniononpresentpoliticalissuesinawaythathasnotbeenwidelyappreciated.Thatis,publicopinionturnsouttobesensitivetocuesthatwererelevantforsocialnavigationinancestralsmallgroupsÑeventhoughthesecuesmightnotbeimportant(ormayevenbecounterproductive)torespondtoinmodernsocieties.Inthecaseofwelfare,itistheperceptionofwelfarerecipientsÕmotivationtowork,andnottheircompetence,thatisthemorepowerfuldeterminantofwelfareopinion.Fromaneconomicperspective,moderntechnologicallybasedadvancedsocietiesplaceapremiumonanindividualÕscompetence,butthemoralintuitionsoftheircitizensdonot.Theseresultshelpelucidatethepsychologicalmechanismsthroughwhichmodernindividualsformopinionsaboutwhetherwelfarerecipientsmeritassistance.Giventheevolutionarilylong-standingselectionpressuresfordiscriminatingbetweencheatersandnoncheaters,wehypothesizedthatthestructureoftheseemotionalmechanismsisspecies-typical.Thatis,individualvariationsintheiroperationsshouldberelativelysmallandrandomlydistributed(Tooby&Cosmides,1990).Consistentwiththisargument,wehavedemonstratedthatangerandcompassionoperateinequalwaysacrosspeoplewithdifferentideologiesanddifferentnationalbackgrounds.However,whileindividualandcross-nationaldifferencesdonotchangethestructureoftheemotionalmechanisms,wedonotinanywayintendtosaythatsuchdifferencesareunimportant(see,e.g.,Alford,Funk,&Hibbing,2008;Jost,Nosek,&Gosling,2010).Rather,wewanttoemphasizethattheirimportancerestsinprovidinginputtotheseandothermechanismsintheabsenceofanyexternallyprovidedandvividcuesabout,forexample,speciÞcwelfarerecipients.Insuchsituations,weshouldexpectouremotionalsystemstofallbackonextractingthecuesnecessaryfortheirexecutionfrominternallyprovidedperceptions,images,andstereotypesaboutthemotivationsofwelfarerecipients(see,e.g.,Petersen,2009;Petersenetal.,2011).Suchperceptionsaremostlikelycoloredbyideology-relevantdifferencesinpersonalityrelatingto,forexample,right-wingauthoritarianism(Altemeyer,1988)andsocialdominanceorientation(Sidanius&Pratto,2001).Inthatregard,theimplicationofthecurrentstudyisthat,totheextentsuchpersonalitydifferencesinßuenceperceptionsofeffortamongWhoDeservesHelp? socialwelfarerecipients,theywillinßuencepoliticalattitudesonsocialwelfarebyregulatingadistinctsetofsocialemotions,angerandcompassion.Giventhatatleastsomeoftheabove-mentionedpersonalitydifferencesrestongeneticallyheritabletraits,theheritabilityofopinionsoneconomicequality(Bell,Schermer,&Vernon,2009)couldreßectprocessesofreactiveheritabilityoperatingthroughthespecies-typicalmechanismsofangerandcompassion(Tooby&Cosmides,1990).Thatis,becauseheritableindividualdifferencescouldinßuencethepropensitytoviewothersascooperativelymotivatedabsentvividcues,thesensitivitytocuesofmotivationintheangerandcompassionprogramswouldmaketheseheritabledifferencesinßuencewelfareopinionsreactively.Inthatway,species-typicalmechanismsandindividualdifferences(heritableandnonheritable)thatserveasinputtothesemechanismscaninteractingeneratingpoliticalbehaviorandorientations.Byelucidatingthestructureofthepsy-chologicalmechanismsthatmediatebetweeninputsandoutputs,evolutionarypsychologicalresearchfacilitatesthestudyofthisinteractionbyhelpingidentifywhichdifferencescouldserveasinputtoaspeciÞcmechanismand,hence,createdifferencesinoutput.Notonlyindividual-leveldifferencesbutalsostructuraldifferencesatthenationallevelcanserveasinputtoevolvedspecies-typicalmechanismssuchasangerandcompassionprograms.Empirically,wehavemadeuseofsamplesfromDenmarkandtheUnitedStates.Oneofmanydifferencesbetweenthesetwocountriesrelatestoracialandethnichomogeneity.Denmarkisrelativelyhomogeneous,andneitherracenorethnicityplaysamajorroleindiscussionsonsocialwelfare(Larsen,2006).Incontrast,theUnitedStatesishighlyheterogeneousand,inaddition,adisproportionatelylargenumberofblackAmericansareonsocialwelfare(Alesinaetal.,2001).Byimplication,raceplaysakeyroleinopinionsonsocialwelfareintheUnitedStates,andwhiteAmericansÕoppositiontosocialwelfareseemstobedrivenprimarilybytheperceptionthatblackAmericansarelazy(Gilens,1999).Thisdifferenceprovidesanillustrationofhowmacrostructuralconditionsshapeourbehaviorsthroughanumberofinteractingpsychologicalmechanisms.Previousresearchinevolutionarypsychologysuggeststhatthehumanminddoesnotincludeadedicatedsystemforcategorizingbyrace.Instead,differencesinappearanceassociatedwithracearepickedupbypsychologicalmechanismsdesignedfortrackingcoalitionsmorebroadly,creatingorenhanc-ingmentalcategoriesofraceandethnicity(Kurzban,Tooby,&Cosmides,2001).Moreover,motivationalmechanismspromptindividualstothinkintuitivelyofreciprocalsocialexchangesasorganizedalongcoalitionallines(Yamagishi,Jin,&Kiyonari,1999).InthemindsofwhiteAmeri-cans,theinteractionofthesetwosetsofmechanismswouldtagsocialwelfarerecipientsasunlikelytoreciprocategiventheirdisproportionateafÞliationwithanothercoalition.Moreover,thisinfor-mationwouldbefedintotheangerandcompassionsystems,down-regulatingsupportforsocialwelfare.Incontrast,inDenmark,thelargerracialhomogeneityimpliesthatsocialwelfarerecipientsaremorelikelytobetaggedascomembersofoneÕscoalitionand,hence,supportforsocialwelfareisonaveragestrong(see,e.g.,Larsen,2006,forevidenceofthesedifferencesinwelfaresupportamongDanesandAmericans).ThepresentÞndingscontributetoextantresearchinanumberofways.Firstandforemost,theycontributetothepoliticalscienceliteratureondeservingnessandsocialwelfarebyofferinganexplanationofhowandwhyperceptionsofwhetherrecipientsarelazyornotsostronglycolorpublicopiniononsocialwelfare.Essentially,thephenomenaofmodernsocialwelfareactivateemotionalandcognitivesystemsdesignedforregulatingancestralsmall-scaleexchangeofhelpand,giventheselectionpressuresoperatingonthesesystems,theyareparticularlysensitivetoeffort-relatedinformation.Second,theÞndingscontributetothelargerpsychologicalliteratureondeservingnessbyshowingthat(inthedomainofhelpandgiving)lackofcontrolintheformofchronicincompe-tencedoesnottranslateintogreaterdeservingness(assometheoriespredict).Fortheactivationofangerorcompassion,competenceorlackthereofappearsirrelevant;butwhenincompetenceiscoupledwithnoncooperativemotivationstheavoidance-orientedemotionofcontemptistriggered.Petersenetal. Third,theseÞndingscontributetothepsychologicalandpoliticalscienceliteratureonemotions.ThisstudydemonstratestheneedforpoliticalresearcherstomakeÞne-graineddistinctionsbetweenemotions,takingtheirfunctionsandtheirevolvedinformationprocessingdesignsintoaccount.Mostpriorresearchonemotionsinpoliticalsciencehasfocusedonemotionalvalence,i.e.,whethertheemotionsarepositiveornegative(e.g.,Brader,Valentino,&Suhay,2008;Kuklinski,Riggle,Ottati,Schwarz,&Wyer,1991;LodgeandTaber,2005;Marcusetal.,2000;Snidermanetal.,1991).Whilesomerecentresearchhasdistinguishedbetweendifferenttypesofnegativeemotions,thisresearchhasstilltreatedanger,contempt,anddisgust(forexample)asequivalents(see,e.g.,Marcusetal.,2006).Inlinewiththeargumentthatdistinctemotionsaredesignedtosolvedistinctproblems,wehavedemonstratedthistobeproblematic.Moreover,priorresearchonemotionshasgenerallyfocusedonprecautionaryemotionssuchasanxiety(cf.Marcusetal.,2000).Suchafocusunderes-timatestheimportanceofotheremotionstopublicopinionformation.Ifmodernmasspoliticsisrepresentedusingevolvedmechanismsdesignedtomanagesocialinteractionsinsmall-scalesoci-eties,publicopinionshouldbeespeciallyinßuencedbysocialemotions.IfthepresentÞndingsareareliableindication,thisisindeedthecase. FigureA1.PerceivingpovertyascausedbylazinesscorrelatessigniÞcantlywithoppositionagainstgovernmentinvolvementinpovertyreductionin48outof49countries.Zero-ordercorrelations.Notes.N59,144.DataisfromtheWorldValuesSurvey1994Ð1999andthevariablesaree131(ÒWhyarepeopleinneed?Becauseoflazinessandlackofwillpowerorbecauseofanunfairsociety?Ó)ande133(ÒHowmuchisthegovernmentdoingagainstpoverty?Toomuch,abouttherightamountortoolittle?Ó)intheEuropeanandWorldValuesSurveysfour-waveintegrateddataÞle,1981Ð2004.SubjectsinthecategoriesÒotheranswersÓhavebeendeletedfromtheanalysis.AllcorrelationsexceptfortheDominicanRepublicandVenezuelaaresigniÞcantatthe.001-level.ThecorrelationfortheDominicanRepublicissigniÞcantatthe.05-level,whilethe-valueforVenezuelaisWhoDeservesHelp? Alesina,A.,Glaeser,E.,&Sacerdote,B.(2001).WhydoesnÕttheU.S.haveaEuropean-stylewelfaresystem?BrookingsPapersonEconomicActivity,2,187Ð254.Alford,J.,Funk,C.L.,&Hibbing,J.(2008).Beyondliberalsandconservativestopoliticalgenotypesandphenotypes.PerspectivesonPolitics,6,321Ð328.Alford,J.,&Hibbing,J.(2004).Theoriginofpolitics:Anevolutionarytheoryofpoliticalbehavior.PerspectivesonPolitics,,707Ð723.Altemeyer,B.(1988).Enemiesoffreedom:Understandingright-wingauthoritarianism.SanFrancisco:Jossey-BassAxelrod,R.,&Hamilton,W.(1981).Theevolutionofcooperation.Science,211,1390Ð1396.Bell,E.,Schermer,J.A.,&Vernon,P.A.(2009).Theoriginsofpoliticalattitudesandbehaviours:Ananalysisusingtwins.CanadianJournalofPoliticalScience,42,855Ð879.Brader,T.,Valentino,N.,&Suhay,E.(2008).Whattriggerspublicoppositiontoimmigration?Anxiety,groupcues,andimmigrationthreat.AmericanJournalofPoliticalScience,52,959Ð978.Cashdan,E.(1980).Egalitarianismamonghuntersandgatherers.AmericanAnthropologist,82,116Ð120.ConverseP.(1964).Ideologyanddiscontent.NewYork:TheFreePress.Cook,F.L.,&Barrett,E.J.(1992).SupportfortheAmericanwelfarestate.NewYork:ColumbiaUniversityPress.Cosmides,L.,Barrett,H.C.,&Tooby,J.(2010).Adaptivespecializations,socialexchange,andtheevolutionofhumanProceedingsoftheNationalAcademyofSciences,107,9007Ð9014.Cosmides,L.,&Tooby,J.(1992).Cognitiveadaptationsforsocialexchange.InJ.Barkow,L.Cosmides,&J.Tooby(Eds.),Theadaptedmind(pp.163Ð228).Oxford:OxfordUniversityPress.CosmidesL.,&Tooby,J.(2005).Neurocognitiveadaptationsdesignedforsocialexchange.InD.M.Buss(Ed.),handbookofevolutionarypsychology(pp.584Ð627).Hoboken,NJ:JohnWiley.Cosmides,L.,&Tooby,J.(2006).Evolutionarypsychology,moralheuristics,andthelaw.InG.GigerenzerandC.EngelHeuristicsandthelaw(pp.182Ð212).Cambridge,MA:MITPress.Delton,A.W.,Cosmides,L.,Guemo,M.,Robertson,T.E.,&Tooby,J.(2012).Thepsychosemanticsoffreeriding:Dissectingthearchitectureofamoralconcept.JournalofPersonalityandSocialPsychology.Advanceonlinepubli-cation.doi:10.1037/a0027026deWaal,F.(1989).Chimpanzeepolitics.Baltimore:JohnsHopkinsUniversityPress.deWaal,F.(1996).Goodnatured;Theoriginsofrightandwronginhumansandotheranimals.Cambridge,MA:HarvardUniversityPress.Ekman,P.(2004).Emotionsrevealed.London:Weidenfeld&Nicolson.Ermer,E.,Guerin,S.,Cosmides,L.,Tooby,J.,&Miller,M.(2006).Theoryofmindbroadandnarrow:ReasoningaboutsocialexchangeengagesTOMareas,precautionaryreasoningdoesnot.SocialNeuroscience,1,196Ð219.Feather,N.T.(1999).Values,achievement,andjustice:StudiesinthePsychologyofDeservingness.NewYork:KluwerFeather,N.T.(2006).Deservingnessandemotions:Applyingthestructuralmodelofdeservingnesstotheanalysisofaffectivereactionstooutcomes.EuropeanReviewofSocialPsychology,17,38Ð73.Fehr,E.,&GŠchter,S.(2000).Cooperationandpunishmentinpublicgoodsexperiments.TheAmericanEconomicReview,,980Ð994.Fiddick,L.(2004).Domainsofdeonticreasoning.QuarterlyJournalofExperimentalPsychology,57A,447Ð474.Gilens,M.(1999).WhyAmericanshatewelfare.Chicago:UniversityofChicagoPress.Goetz,J.L.,Keltner,D.,&Simon-Thomas,E.(2010).Compassion:Anevolutionaryanalysisandempiricalreview.PsychologicalBulletin,136,351Ð374.Harris,P.,Nunez,M.,&Brett,C.(2001).LetÕsswap:EarlyunderstandingofsocialexchangebyBritishandNepalichildren.MemoryandCognition,29,757Ð764.Hill,K.,&Hawkes,K.(1983).NeotropicalhuntingamongtheAcheofEasternParaguay.InR.Hames&W.Vickers(Eds.),AdaptationsofnativeAmazonians(pp.139Ð188).NewYork:AcademicPress.Iyengar,S.(1991).Isanyoneresponsible?Chicago:TheUniversityofChicagoPress.Jost,J.T.,Nosek,B.A.,&Gosling,S.D.(2010).Ideology:Itsresurgenceinsocial,personality,andpoliticalpsychology.PerspectivesonPsychologicalScience,3,126Ð136.Kameda,T.,Takezawa,M.,Tindale,R.S.,&Smith,C.(2002).Socialsharingandriskreduction:Exploringacomputationalalgorithmforthepsychologyofwindfallgains.EvolutionandHumanBehavior,23,11Ð33.Petersenetal. Kaplan,H.,&Gurven,M.(2005).Thenaturalhistoryofhumanfoodsharingandcooperation:AreviewandanewmultiÐindividualapproachtothenegotiationofnorms.InH.Gintis,S.Bowles,R.Boyd,&E.Fehr(Eds.),sentimentsandmaterialinterests:Thefoundationsofcooperationineconomiclife(pp.75Ð113).Cambridge,MA:MITKaplan,H.,&Hill,K.(1985).FoodsharingamongAcheforagers:Testsofexplanatoryhypotheses.CurrentAnthropology,,223Ð246.Kaplan,H.,Hill,K.,&Hurtado,A.M.(1990).Risk,foragingandfoodsharing.InE.Cashdan(Ed.),Riskanduncertaintyinthefoodsupply(pp.107Ð144).Boulder,CO:WestviewPress.Kaplan,H.S.,Hill,K.R.,Lancaster,J.B.,&Hurtado,A.M.(2000).Atheoryofhumanlifehistoryevolution:Diet,intelligence,andlongevity.EvolutionaryAnthropology,9,156Ð185.Kuklinski,J.H.,Riggle,E.,Ottati,V.,Schwarz,N.,&Wyer,R.S.(1991).Thecognitiveandaffectivebasesofpoliticaltolerancejudgments.AmericanJournalofPoliticalScience,35,1Ð27.Kurzban,R.,Tooby,J.,&Cosmides,L.(2001).Canracebeerased?Coalitionalcomputationandsocialcategorization.ProceedingsoftheNationalAcademyofSciences,98,15387Ð15392.Larsen,C.A.(2006).Theinstitutionallogicofwelfareattitudes.London:Ashgate.Lee,R.B.,&DeVore,I.(Eds.).(1968).Manthehunter.NewYork:AldinedeGruyter.Lodge,M.,&Taber,C.S.(2005).Theautomaticityofaffectforpoliticalleaders,groups,andissues:Anexperimentaltestofthehotcognitionhypothesis.PoliticalPsychology,26,455Ð482.Marcus,G.,MacKuen,M.,Wolak,J.,&Keele,L.(2006).Themeasureandmismeasureofemotion.InD.Redlawsk(Ed.),Feelingpolitics(pp.31Ð46).NewYork:PalgraveMacmillan.Marcus,G.,Neuman,R.,&MacKuen,M.(2000).Affectiveintelligenceandpoliticaljudgment.Chicago:UniversityofChicagoPress.Neuman,R.,Marcus,G.,Crigler,A.,&MacKuen,M.(2007).Theaffecteffect.Cambridge:CambridgeUniversityNew,J.,Cosmides,L.,&Tooby,J.(2007).Category-speciÞcattentiontoanimalsreßectsancestralprioritiesnotexpertise.ProceedingsoftheNationalAcademyofSciences,104,16598Ð16603.OÕConnell,J.F.,Hawkes,K.,&Jones,N.G.(1991).HuntingincomepatternsamongtheHadza:Biggame,commongoods,foraginggoalsandtheevolutionofthehumandiet.PhilosophicalTransactions:BiologicalSciences,334Oorschot,W.(2000).Whoshouldgetwhat,andwhy.PolicyandPolitics,28(1),33Ð49.Oorschot,W.(2006).MakingthedifferenceinsocialEurope:DeservingnessperceptionsamongcitizensofEuropeanwelfareJournalofEuropeanSocialPolicy,16(1),23Ð42.Petersen,M.B.(2009).Publicopinionandevolvedheuristics:Theroleofcategory-basedinference.JournalofCognitionandCulture,9,367Ð389.Petersen,M.B.(2010).Distinctemotions,distinctdomains:Anger,anxietyandperceptionsofintentionality.JournalofPolitics,72,357Ð365.Petersen,M.B.(2012).Socialwelfareassmall-scalehelp:Evolutionarypsychologyandthedeservingnessheuristic.AmericanJournalofPoliticalScience,56,1Ð16.Petersen,M.B.,Roepstorff,A.,&Serritzlew,S.(2009).Socialcapitalinthebrain?InG.T.Svendsen&G.L.H.SvendsenHandbookofsocialcapital(pp.75Ð92).Northampton,MA:EdwardElgar.Petersen,M.B.,Sell,A.,Tooby,J.,&Cosmides,L.(2010).Evolutionarypsychologyandcriminaljustice:Arecalibrationaltheoryofpunishmentandreconciliation.InHenrikH¿gh-Olesen(Ed.),Humanmoralityandsociality:Evolutionaryandcomparativeperspectives(pp.72Ð131).Basingstoke,England:PalgraveMacmillan.Petersen,M.B.,Slothuus,R.,Stubager,R.,&Togeby,L.(2011).Deservingnessversusvaluesinpublicopiniononwelfare:Theautomaticityofthedeservingnessheuristic.EuropeanJournalofPoliticalResearch,50,24Ð52.Price,M.E.,Cosmides,L.,&Tooby,J.(2002).Punitivesentimentasananti-freeriderpsychologicaldevice.EvolutionandHumanBehavior,23,203Ð231.Redlawsk,D.(Ed.).(2006).Feelingpolitics.NewYork:PalgraveMacmillan.Rindos,D.(1987).Theoriginsofagriculture:Anevolutionaryperspective.Waltham,MA:AcademicPress.Rothstein,B.(1998).Justinstitutionsmatter.Cambridge:CambridgeUniversityPress.Rozin,P.,Haidt,J.,&McCauley,C.R.(2000).Disgust.InM.Lewis&J.Haviland(Eds.),Handbookofemotions(2nded.,pp.637Ð653).NewYork:GuilfordPress.Rozin,P.,Lowery,L.,Imada,S.,&Haidt,J.(1999).TheCADTriadHypothesis.JournalofPersonalityandSocialPsychology,76,574Ð586.WhoDeservesHelp? Schreiber,D.2007.Politicalcognitionassocialcognition:Areweallpoliticalsophisticates?InW.RussellNeuman,GeorgeE.Marcus,AnnN.Crigler,&MichaelMackuen(Eds.),Theaffecteffect(pp.48Ð70).ChicagoandLondon:UniversityofChicagoPress.Sell,A.,Tooby,J.,&Cosmides,L.(2009).Formidabilityandthelogicofhumananger.ProceedingsoftheNationalAcademyofSciences,106(35),15073Ð15078.Sidanius,J.,&Pratto,F.(2001).Socialdominance:Anintergrouptheoryofsocialhierarchyandoppression.Cambridge:CambridgeUniversityPress.Skitka,L.J.,Mullen,E.,GrifÞn,T.,Hutchinson,S.,&Chamberlin,B.(2002).Dispositions,ideologicalscripts,ormotivatedcorrection?Understandingideologicaldifferencesinattributionsforsocialproblems.JournalofPersonalityandSocialPsychology,83,470Ð487.Skitka,L.J.,&Tetlock,P.E.(1993).Providingpublicassistance:Cognitiveandmotivationalprocessesunderlyingliberalandconservativepolicypreferences.JournalofPersonalityandSocialPsychology,65,1205Ð1223.Sniderman,P.M.,Brody,R.A.,&Tetlock,P.E.(1991).Reasoningandchoice:ExplorationsinpoliticalpsychologyCambridge:CambridgeUniversityPress.Stone,V.,Cosmides,L.,Tooby,J.,Kroll,N.,&Knight,R.(2002).Selectiveimpairmentofreasoningaboutsocialexchangeinapatientwithbilaterallimbicsystemdamage.ProceedingsoftheNationalAcademyofSciences,99,11531Ð11536.Sugiyama,L.(2004).Illness,injury,anddisabilityamongShiwiarforager-horticulturalists:Implicationsofhealth-riskbufferingfortheevolutionofhumanlifehistory.AmericanJournalofPhysicalAnthropology,123,371Ð389.Sugiyama,L.S.,Tooby,J.,&Cosmides,L.(2002).Cross-culturalevidenceofcognitiveadaptationsforsocialexchangeamongtheShiwiarofEcuadorianAmazonia.ProceedingsoftheNationalAcademyofSciences,99,11537Ð11542.Tooby,J.,&Cosmides,L.(1990).Ontheuniversalityofhumannatureandtheuniquenessoftheindividual:Theroleofgeneticsandadaptation.JournalofPersonality,58,17Ð67.Tooby,J.,&Cosmides,L.(1996).Friendshipandthebankersparadox:OtherpathwaystotheevolutionofadaptationsforProceedingsoftheBritishAcademy,88,119Ð143.Tooby,J.,&Cosmides,L.(2008).Theevolutionarypsychologyoftheemotionsandtheirrelationshiptointernalregulatoryvariables.InM.Lewis&J.M.Haviland-Jones(Eds.),Handbookofemotions(3rded.,pp.114Ð137)NewYork:GuilfordTooby,J.,Cosmides,L.,Sell,A.,Lieberman,D.,&Sznycer,D.(2008).Internalregulatoryvariablesandthedesignofhumanmotivation:Acomputationalandevolutionaryapproach.InA.J.Elliot(Ed.),Handbookofapproachandavoidance(pp.251Ð271).Mahwah,NJ:LawrenceErlbaum.Trivers,R.(1971).Theevolutionofreciprocalaltruism.TheQuarterlyReviewofBiology,46,35Ð57.Weiner,B.(1995).Judgmentsofresponsibility.NewYork:GuilfordPress.Yamagishi,T.,Jin,N.,&Kiyonari,T.(1999).Boundedgeneralizedreciprocity:Ingroupboastingandingroupfavoritism.AdvancesinGroupProcesses,16,161Ð197.Petersenetal.