Scope insensitivity: The limits of intuitive valuation of human lives in public policy

(1)

Scope insensitivity: The limits of intuitive

valuation of human lives in public policy

Stephan Dickert, Daniel Västfjäll, Janet Kleber and Paul Slovic

Linköping University Post Print

N.B.: When citing this work, cite the original article.

Original Publication:

Stephan Dickert, Daniel Västfjäll, Janet Kleber and Paul Slovic, Scope insensitivity: The limits

of intuitive valuation of human lives in public policy, 2015, Journal of Applied Research in

Memory and Cognition, (4), 3, 248-255.

http://dx.doi.org/10.1016/j.jarmac.2014.09.002

Copyright: 2015 Published by Elsevier Inc on behalf of Society for Applied Research in

Memory and Cognition. This is an open access article under the CC BY-NC-ND license.

http://www.elsevier.com/

Postprint available at: Linköping University Electronic Press

(2)

Journal

of

Applied

Research

in

Memory

and

Cognition

j ou rn a l h o m epa g e : w w w . e l s e v i e r . c o m / l o c a t e / j a r m a c

Original

Article

Scope

insensitivity:

The

limits

of

intuitive

valuation

of

human

lives

in

public

policy

Stephan

Dickert

a,b,∗

_,

_Daniel

_Västfjäll

b,c

_,

_Janet

_Kleber

d

_,

_Paul

_Slovic

c,e a_Department_of_Marketing,_WU_Vienna_University_of_Economics_and_Business,_Austria

b_Psychology_Department,_Linköping_University,_Sweden c_Decision_Research,_USA

d_Psychology_Department,_University_of_Vienna,_Austria e_Department_of_Psychology,_University_of_Oregon,_USA

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Received15October2013 Accepted30September2014 Availableonline7October2014 Keywords:

Valueoflife Scopeinsensitivity Informationprocessing

a

b

s

t

r

a

c

t

Acriticalquestionforgovernmentofﬁcials,managersofNGOs,andpoliticiansishowtorespondto situationsinwhichlargenumbersoflivesareatrisk.Theoriesinjudgmentanddecisionmakingaswell aseconomicssuggestdiminishingmarginalutilitywithincreasingquantitiesofgoods.Inthedomain oflifesaving,thisformofnon-linearityimpliesdecreasingconcernforindividuallivesasthenumberof affectedpeopleincreases.Inthispaper,weshowhowintuitivevaluationsbasedonprosocialemotions canleadtoscopeinsensitivityandsuboptimalresponsestolivesatrisk.Wepresentbothnormative anddescriptivemodelsofvaluationsoflivesanddiscusstheunderlyingpsychologicalprocessesasthey relatetojudgmentsanddecisionsmadeinpublicpolicyandbyNGOs.

1. Introduction

Peopleinorganizationsoftenhavetomakedecisionsthataffect thewelfareofothers.Thisincludesallocationsofresources(e.g., money,time,andservices)bynationalandinternational organi-zationsaswellas bygovernmentagencies.In situations where thewelfareoflargesegmentsofapopulationisthreatened,both government and non-government organizations (e.g.,charities; NGOs)arecalledupontoreact.Recentexamplesofsuchlarge-scale threatsincludetheongoinghumanitariancrisesinAfricaaswellas armedconﬂicts(e.g.,thecivilwarinSyria)andnaturalcatastrophes (e.g.,hurricanes and earthquakes). In response to such calami-ties, international humanitarian aid in 2011 consisted of both government-provided($12.5billion)andprivatevoluntary($4.6 billion)contributions(GlobalHumanitarianAssistance,2013).In ordertounderstandhowmanagers,civilservants,politicians,and otheradministratorsmakedecisionsregardingthewelfareof peo-plewhoselivesareindanger,weneedtobetterunderstandhow peoplevaluethelivesofothers.

Inthisarticlewegiveexamplesofhowhumanlivesarevalued andhowthisdeviatesfromhowtheyshouldbevaluedaccording

∗ Correspondingauthorat:DepartmentofMarketing,WUViennaUniversityof EconomicsandBusiness,Welthandelsplatz1,1020Vienna,Austria.

Tel.:+431313364522;fax:+43131336904522. E-mailaddress:stephan.dickert@wu.ac.at(S.Dickert).

toegalitariannorms.Wearguethatvaluationsoflivesareprone towell-documentedbiaseswhendoneintuitively.Followingdual processmodels,wedeﬁneintuitiveinformationprocessingtobe primarilyautomaticandaffectiveinnatureanddeliberative infor-mationprocessingtobecontrolledandreason-based(Evans,2008; Kahneman,2003;Stanovich&West,2000).Weﬁrstpresenttwo examplesof valuations that follow normative moral principles, thencontrast themwithdescriptive valuationmodelsresulting from intuitive processing. Finally, we present a short studyto demonstratehowvariationsinelicitationmethodscanleadto dif-ferentvaluationsoflivesandclosewithadiscussiononhowthese valuationscouldbeimproved.

Experimentalresearchhasshownthatvaluationsoflivesare oftenscopeinsensitive,whichisthetendencytoberelatively unre-sponsivetothenumberofpeopleatriskinlarge-scalehumanitarian catastrophes.Scopeinsensitivitycanleadtosuboptimaldecision outcomesinpublicpolicy.Thisisthecaseinsituationswherethe goalistoimprovethewelfareofasmanypeopleaspossible,but thedecisions(andunderlyingvaluations)donotcorrespondtothis goal(Baron&Szymanska,2011).Beforeexaminingthereasonsfor scopeinsensitivity,webrieﬂyreﬂectonsomenormativeaspectsof valuationsoflife.

1.1. Normativevaluationsoflives

Howshouldhumanlivesbevalued?Thisisacomplex ques-tionwhoseanswerdependslargelyontheadoptedphilosophical

http://dx.doi.org/10.1016/j.jarmac.2014.09.002

(3)

Fig.1.Linearfunction.

viewpoint.Whatisconsiderednormativemaydependonseveral differentcriteriaandgoalspursued.Herewefocusontwo exam-plesofnormativeprinciplesthatcanbejustiﬁedwithegalitarian moralvalues andthegoalofgroupsurvival.Accordingto egali-tarianmoralperspectives(andvariousformsofutilitarianism),all livesshouldbevaluedequally(Baron&Szymanska,2011;Dickert, Västfjäll,Kleber,&Slovic,2012;Sinnott-Armstrong,2011).In addi-tion tothis moralprinciple, it is alsopossible to acknowledge thatlossesoflivesmaysometimesbesolargeastothreatenthe survivabilityofapopulation.Insuchcases,thevalueofeach addi-tionalendangeredlifeincreasesdisproportionally(Slovic,Fischoff, &Lichtenstein,1982).

Ifalllivesareinherentlyofequalvalue,onewouldexpectpolicy decisionstobefaithfultothisegalitarianmoralprinciple when-everpossible.1 _Such_valuations_are_captured_by_a_simple_formula

inwhich thelevel ofresponse is givenby R=X×N,whereR is theaidresponse(measuredeitherinﬁnancialcontributions, vol-unteertime,volunteernumbers,etc.),Xistheresponseforone victim,andNisthenumberofvictims.Theresultinglinear func-tionisdepictedinFig.1andshowsthateachadditionallifeatrisk shouldincreaseresourceallocationtothesameextent.An under-lyingassumptionofthisvaluationfunctionisthattheefﬁciencyof lifesavingremainsconstantsuchthatsavinganadditionallifedoes notbecomecheaperifmorelivesareatrisk.Thisassumptionis reasonableinsituationsinwhichtheimpactofhowmuchan addi-tionaldollarcandotosavealifedoesnotchange(e.g.,ifthecost andeffectivenessofanadditionalvaccineorabowlofriceremains constant).

ThesecondnormativevaluationfunctionisillustratedinFig.2

andis linearuntil thenumberof livesat riskreachesa critical pointTatwhichthesustainabilityofthegroupisthreatened.After thispoint,thevalueofeachadditionallifeatriskincreases expo-nentially,whichcanbemodeledbyavaluefunctionofR=X×Nb_,

withb=1forallN≤Tandb>1forN>T.Asthenumberoflivesat riskincreases,progressivelymoreaidisgiventoeachone.Both functionscanbeconsiderednormativebecausetheyrepresent val-uationsthatarebasedongenerallyacceptedmoralprinciples(i.e., equalityoflivesandsurvivalofthegroup).However,itshouldalso benotedthatotherpossiblenormativevaluationfunctionsexist. Forexample, ifthegoalis tosavea speciﬁcnumberof livesin ordertoreachacriticalthresholdneededforthesurvivalofagroup,

1_Naturally,_constraints_or_competing_objectives_may_prevent_policy_decisions_to

alwaysfollowsuchaprinciple.

Fig.2.ExponentialfunctionwiththresholdT.

valuationsmightincreasesharplyuntilsuchthresholdisreached andthenleveloffordrop.2,3

1.2. Psychophysicalnumbing

Evidenceexiststhatvaluationsoflivesunderlyingaidresponses donotalwaysfollowsuchnormativemodels.Asthenumberoflives atriskincreases,peopletendtoexhibitvaluationsthat become progressivelyless sensitive tochanges in victim numbers. This diminished sensitivity to thevalue of life wasdocumented by

Fetherstonhaugh,Slovic,Johnson,andFriedrich(1997)byassessing individuals’willingnesstoaidgroupsofdifferentsizes.Inonestudy thatcomparedtheeffectofthesizeofrefugeecamps,participants statedthatitwouldbemorebeneﬁcialtosave4500liveswhen thesizeofthecampwassmaller(11,000refugees)comparedto alargercamp(250,000refugees).Thissuggeststhatrespondents valuedsaving4500livesinthesmallercampmorethansavingthe sameamountwhenmoreliveswereatrisk.Ifeachindividuallife thatcanbesavedisvaluedtothesameextent(asproposedbya lin-earvaluationfunction),thenthesizeoftherefugeecampshouldnot makeadifference.However,theresultsbyFetherstonhaughetal. (1997)suggestthatparticipantswerelesssensitivetothenumber ofliveswhentheproportionwaslow(2%saved)comparedtohigh (41%saved).4

Inaccordancewithsimilarinsensitivitytochangesinquantity inthedomainofperception,Fetherstonhaughetal.(1997)termed thistypeofvaluation“psychophysicalnumbing”.Itcanbecaptured by themathematicalformula of R=X×Nb_,_with_an_exponential

coefﬁcient0<b<1. Thediminishingsensitivity (with increasing quantity of a stimulus) gives rise to a distinct functional form whichdescribesvaluationsofseveraldomains,includingvisualand auditoryperception,thevalueofmoney,aswellasthevalueof humanlives(seeFig.3).Alsoknownasdiminishingmarginalutility amongeconomists(e.g.,vonNeumann&Morgenstern,1944)and

2_We_would_like_to_thank_an_anonymous_reviewer_for_this_suggestion. 3_An_entirely_different_{philosophical}_approach_to_normative_valuations_was

sug-gestedbyTaurek(1977),whoassertedthattheutilityofsavingonelifecannotbe meaningfullyaddedtosavinganotherlife.Accordingtothisperspective,savinglives isnotautility-maximizingproblem.Instead,thewaylivesshouldbevaluedisto giveeachlifeatriskthesamechanceofsurvival,regardlessofhowmanylivesarein danger.Whengiventhechancetoeithersaveoneor50people,an“equalchance” entailsﬂippingacointodeterminewhoisbeingsaved.

4_Although_not_normative_according_to_a_linear_valuation,_an_exponential

normat-ivemodelcouldintheoryexplaintheseﬁndingsifparticipantsperceivedtheaffected proportionofthesmallercamptobelargeenoughtothreatenthesurvivabilityof theentirecamp.

(4)

Fig.3.Psychophysicalnumbingfunction.

depictedasaconcavevaluefunctioninprospecttheory(Kahneman &Tversky,1979),psychophysicalnumbingimpliesthatsavingan additionallifeisalwaysvaluedlessthanthepreviouslife(fora sim-ilarargumentseeAndreoni,2007).Thisinsensitivityisatoddswith thenormativemodelspresentedearlier.Avaluationpatternwith decreasingmarginalutilitycanbeappropriatefortypicalmarket goodsandisgenerallyconsideredrationalamongutilitytheorists. However,applyingtheconceptofdiminishingmarginalvalueinthe domainofhumanitarianaidcanleadtounwantedconsequences. Besidesvaluingeachadditionallifelessthanthepreviouslife,more weightisgiventosavinghigherproportionsratherthanahigher absolutenumberoflives.

Nonetheless,evaluatingthequalityofhumanitarianaidbythe proportionoflivessavedseemstoholdintuitiveappeal.Research onproportionalreasoningshowsthatpeoplecanpreferhelpinga largerproportionofvictimsevenifthismeanshelpingasmaller absolutenumber (Bartels, 2006; Bartels &Burnett, 2011). Fur-thermore,makingpeopleawareofthistradeoffdoesnotseemto changetheirpreferences.Onereasonforthis“proportion domi-nance”isrelatedtothefactthatproportionsareintuitivelyeasierto understandthanabsolutenumbers(e.g.,Slovic,Finucane,Peters,& MacGregor,2002).Indeed,proportionsassertstrongerinfluenceon valuationswhenthequantitytobejudgedisnotreadilyevaluable byitself(Hsee&Zhang,2010;Hsee,1996,1998),asisthecasefor valuationsoflives(Slovic,2007;Slovic,Zionts,Woods,Goodman,& Jinks,2013).Forexample,ahumanitarianaidinterventionprogram saving80%oflives atriskis clearlygood whereasan interven-tionprogramsavingaspecificabsolutenumberoflives(e.g.,80) isdifficulttoevaluatewithoutadditionalinformation(suchasthe groupsizeofallthatareaffectedorthenumberoflivessavedby anotherinterventionprogram).Proportionscarryinherent affec-tivemeaningthatinformsvaluationsbuttheycangiverisetoscope insensitivitywhenlargechangesinproportionsareneededto sub-jectivelyfeelthedifferenceandtakeaction.

1.3. Psychicnumbingandtheroleofemotionsinscope insensitivity

Psychophysicalnumbingdescribesthelackofscalingof human-itarianresponseswithrespecttoanincreasingnumberoflivesat risk.However,itcannotexplainsituationsinwhichpeoplefailto respondentirely.Thiscanbethecasewhenpeopleareconfronted withhumanitariancrisesthataffectlargegroupsofvictims(e.g., genocides,wars,orfamines).AnexampleisthegenocideinDarfur, whichhasevokedlittlepublicconcern.Thisisespecially problem-aticonthelevelofgovernmentinstitutionsandnon-government

Fig.4.Psychicnumbingfunction.

organizations.However,theseseeminglyuncaringresponsesare alsoevidentinjudgmentsanddecisionsmadebyindividuals.

Giventhegreatextentthathumansarecapableofcaringfor individuallives, it issurprising thatone shouldremainpassive andsilentwhenlargenumbersoflivesarethreatened.Itseems asthoughpeoplearenotabletomultiplythecaringtheyfeelfor oneindividualbythenumberofpeopleatrisk(Slovic,2010).Infact, beingconfrontedwithtoomuchsufferingcanleadtothecollapse ofcompassion,whichisresponsiblefortheinactivityinthefaceof largecatastrophes.Tomodelthisbreakdownofcompassion,Slovic (2007)hasproposeda“psychicnumbing”function,whichis pre-sentedinFig.4.Itfollowsafunctionalformsimilartoaninverse exponentialfunctionR=X×N−b,with0<b<1.Accordingtothis valuationmodel,humanitarianaidresponsesarelargestwhenonly fewlivesareatriskanddecreaseasthenumberincreases.This valu-ationmodelwasdevelopedtodescribeemotionalreactionstoward peopleinneed,howeversincepeople’swillingnesstohelpisoften basedonemotionsitcanalsoreﬂectaidresponses.

Tounderstandwhyandwhencompassioncollapsesintheface oflargevictimnumbers,itishelpfultolookathowpro-social emo-tionsare generatedaswellasregulated.Ampleevidenceexists thatemotionslikesympathyandcompassionarestrongerwhen itiseasytovisuallyattendtoormentallyimaginethevictims.For example,presentingvictimsaloneratherthanaspartofalarger groupusuallycreatesmorepro-socialemotions(Dickert&Slovic, 2009;Dickert, Kleber,Peters, &Slovic,2011;Dickert, Sagara,& Slovic,2011;Loewenstein&Small,2007;Slovic&Västfjäll,2010; Small, Loewenstein, &Slovic, 2007).This is especially the case whenthefocusisonidentifiedsinglevictimsversuslargegroups of victims(Kogut&Ritov,2005a,b, 2007; Slovic,2007). Neuro-scientific evidence shows that the emotional reward center of thebrain,thestriatum,wasmoreactivatedforidentifiedvictims (comparedtonon-identifiedvictims),andfurther,thatthis acti-vationpredictedhowmuch participantswerewillingtodonate (Genevsky,Västfjäll,Slovic,&Knutson,2013).Also,whenvictims are perceived as a coherent unit (Hamilton & Sherman, 1996; Smith,Faro,&Burson,2013)andphysicallyorpsychologicallyclose (Batson,1990;Loewenstein&Small,2007)usuallyprosocial emo-tionsarestronger.Sympathyisalsoeasiertoevokethroughsudden changes(comparedtolong-lastingeffects)inthewellbeingof oth-ers(Small,2010).Together,thisevidencesuggeststhatcontextual informationenhancesthegenerationofprosocialemotionswhen thisinformationhighlightsthepressingneedofafewidentified victims.

In situations where many people needhelp, prosocial emo-tions may not be sufﬁciently generated in the ﬁrst place or

(5)

regulateddowntoreduceempathicoverload(Cameron&Payne, 2011;Rubaltelli&Agnoli,2012).Someevidenceexiststhatother, sometimeshighly pertinentinformation hassimilardeleterious effects onthegeneration ofpro-socialemotionsand valuations of individual lives. For example, presenting information in the form of background statistics about the numbers of lives at risk can decrease people’s willingness to contribute to chari-tablecauses. Small et al. (2007) presented an identified single victimtogetherwithbackgroundstatisticsaboutageneral human-itarian aid crisis and found that financial contributions to the singlevictimdecreased compared towhen thevictimwas pre-sented withoutstatistics. These authorsargue that information aboutidentified single victimsis processed affectively whereas adding non-affective information such as background statistics leadstostrongerdeliberativeprocessingthatreducessympathetic responses.Theseexamplesalsoshowthattheeffectsofemotional valuationsarenotalwaysideal.Emotionsdirectattentiontowhatis consideredthemostpressingneed,buttheymayalsocreatescope insensitivity.Toomuchortoolittleemotionalinvolvementcanlead tofailuretoreactaswellascompassioncollapse.5

Itisquitepossiblethatvaluationsascribedtotheidentifiedand singlevictimeffects arealsorelatedtotheperceivedefficacyof potentialdonors.Someevidencesuggeststhatproportion dom-inancein charitablegiving is mediatedby theperceivedutility ofadonation(Erlandsson,Björklund,&Bäckström,2013). Appar-ently,savingahigherpercentageoflivesissometimesseenasmore useful.Asdonorsdoliketofeelthattheirdonationmakesa differ-ence,describinggeneralhumanitarianaidprojectsinmoredetail ledtoanincreasedwillingnesstodonateinCryder,Loewenstein, andScheines(2013).Conversely,situationsinwhichtheperceived efficacyofadonationisunderminedbycontextualinformation sug-gestingthathelpingresponsesarelikelytobefruitlessdecrease caringandprosocialaction.Researchhasshownthatinformation aboutothervictims (i.e.,thenumber of victimsthat cannotbe helped)cancreateastateofpseudo-inefficacywhichdemotivates helping(Västfjäll,Slovic,&Mayorga,submittedforpublication). Pseudo-inefficacyandcompassioncollapsecanexplainwhyand whenvaluationsbecomescopeinsensitiveand followapsychic numbing function. In a recent study, participants’ compassion towardchildreninneeddecreasedassoonasmorethanonechild wasindanger(Västfjäll,Slovic,Mayorga,&Peters,2014).Thisisin linewiththenotionthatbothemotionalreactionsandperceived efficacyaregreatestwhenonlyfewlivesareatrisk,whichcanlead toanintuitivepreferenceforlifesavinginterventionsthatbenefit smallergroupsofaffectedpeople.

1.4. Valuationpreferences

Weconductedasmallstudytoexamineifpeoplehave system-aticpreferencesforspecificvaluationfunctionsandwhetherthis dependsonhowthevaluationpreferencesareelicited.Thestudy wasnotdesignedtotestwhich valuationfunctionisbestinall kindsof situationsand underallpossiblecircumstances.Itwas designedtoprovideasimpleexampleofhowstatedpreferences forvaluations canbedisconnected fromactualchoicebehavior inlifesavingcontexts.Asthenatureofnormativeanddescriptive valuationmodelsisofspecialinterestforpublicpolicydecisions, wewantedtoseewhetherpreferencesofindividualpeoplecan bereverseddependingontheelicitationmethod(Lichtenstein& Slovic,2006).Specifically,weaimedtotest(1)whetherspecific valuationfunctionsarepreferredoverotherswhenusingratingsas elicitationmethod,(2)whichfunctionpeoplewouldchoose,and(3)

5_See_Grant_and_Schwartz₍₂₀₁₁₎_for_a_similar_argument_on_various_{psychological}

constructsthatfollowanon-monotonicinverted-Ushape.

whichvaluationfunctionbestdescribesactualallocationsbyNGOs andgovernmentagencies.Finally,wegaveparticipantsachoice scenarioinwhichtheycouldchoosetosaveahigherproportion ofvictimsversusalargernumberofvictimstoseewhethertheir revealedpreferenceswouldmatchtheirstatedpreferences.

In the study,we asked participants(n=41; Mage=22.9;68%

female)recruitedfromtheUniversityofViennatoimagineworking foracharitableorganizationandbeingresponsibleforthe alloca-tionoffundsinanAfricanvillagefacingaseverefoodshortage. Participantsweretoldthattheycoulddistributeamaximumof 5000donatedEurosandthat100peopleliveinthevillage.The distributionofdonationscanfollowoneofthreedifferentplans (ProgramsA–C).In awithin-subjectsdesign,wethenpresented threedifferentvaluationmodelstoallparticipantswhichdepicted functionssimilartothosepresentedinFig.1(ProgramA:linear val-uations),3(ProgramB:psychophysicalnumbing),and4(Program C:psychicnumbing).Thedetailsofeachvaluationfunctionwere explainedundertheﬁgures(e.g.,valuationsincreaseataconstant rateinFig.1,increaseatadecreasingrateinFig.3,anddecreasein

Fig.4;fordetailsseeAppendixA).Foreachfunction,participants ratedhowmuchtheyagreedthatthedonationmoneyshould gen-erallybedistributed asdepictedin theﬁgures (onascale from 1=do notagree atallto7=agree verymuch), whichwe inter-pretedastheir“normativepreferences”.Nextweaskedparticipants tochooseoneofthesefunctionsastheirpreferredoptionto dis-tributethedonationmoney,whichweinterpretedastheirstated “subjectivepreference”.Afterthat,participantsjudgedhow dona-tionsareactuallyallocatedinsuchacontext,whichweinterpretas “judgedallocations”.Finally,wepresentedthesameparticipants withadifferentdecisionscenariounrelatedtotheﬁrstinwhich theywereaskedtochoosebetweentwocharitableprograms pro-vidingfoodtopeopleinneed.“ProjectA”providedfoodfor102 outof115peopleatriskofstarvation,whereas“ProjectB” pro-videdfoodfor 105out of247peopleat riskof starvation.This taskfollowedthetypicalformatofproportiondominancestudies (e.g.,Bartels,2006;Fetherstonhaughetal.,1997),andassumesthat thecostforbothprogramsisthesame.Alinearvaluationfunction shouldfavorProjectB,sincesaving105livesisbetterthansaving 102.

Arepeatedmeasuredanalysisofvarianceshowedthat partic-ipants’normativepreferencesdifferedsigniﬁcantlybetweenthe valuationfunctionsF(2,80)=46.3,p<.001,2

p=.537.Ascanbeseen

in Fig. 5,preferences were higher for thelinear function com-paredtothepsychophysicalandpsychicnumbingfunctions.Also, toindicatetheirsubjectivepreferenceparticipantschosethelinear functionsigniﬁcantly moreoftenthan thepsychophysical func-tion,X2 _(1,_N₌₄₁₎₌_10.8,_p₌_.001_(and_no_one_chose_the_psychic

numbingfunction;seeFig.6).Whenaskedwhichfunctionbest describesactualallocationsbyNGOs, participants’choices were alsonotequallydistributed,X2_(2,_N₌₄₀₎₌_5.2,_p₌_.076._{Speciﬁcally,}

inthis casemostparticipantsselectedthepsychophysical func-tion.Lastly,theresultsofthesecondchoicescenariobetweenthe twocharitableprojectsindicatedthatmostparticipantspreferred theprojectwhichsavedthehigherproportion,X2_(1,_N₌₃₆₎₌_4.0,

p=.045(seeFig.7),whichisconsistentwiththepsychophysical function.

Theseresultssuggest that,onaverage,although participants preferalinearvaluationfunctionwhenaskedabouttheir normat-iveandsubjectivepreferences,theydonotnecessarilyapplythese preferenceswhengivenachoicescenarioinwhichsavingahigher proportionispittedagainstsavingahighernumberofvictims.This proportionalresponseisconsistentwithpsychophysicalnumbing andislikelydrivenbytheintuitiveappealofproportiondominance (i.e.,helpingalargerproportionfeelsbetter)andtherelative insen-sitivityofsaving105versus102lives.Theresultsofoursmallstudy serveasademonstrationofhowstatedpreferencesarenotalways

(6)

Fig.5. Normativepreferences(ratings).Errorbarsdenotestandarderrorsofthe mean.

connectedtochoicesinlifesavingcontexts,evenwhenbeingaware ofthedifferentvaluationfunctions.Insituationswheremorelives andmoremoneyareatplay,adivergencebetweenvaluationsand choicescouldfailtoaccomplishthegoalofsavingasmany peo-pleaspossible.Oneofthequestionsthatresearchneedstoanswer regardingresultssuchastheseishowpeoplecanbemademore scopesensitive(i.e.,morelinearintheirvaluations).

1.5. Waystoincreasescopesensitivity

Suggestions to increase scope sensitivity may be based on improvingdecisionmakers’affectiveresponsesaswellasengaging appropriatedeliberativeresponses.Notethataffectiveresponses anddeliberationshouldnotbeseenascompletelyseparatemental processesforthispurpose,asincreaseddeliberationcanalsolead tostrongeraffectiveresponsesandviceversa.Althoughprosocial emotionsarebounded(i.e.,onecannotfeellimitlesscompassion;

Batson,1990;Batson,O’Quin,Fultz,Vanderplas,&Isen,1983),it ispossibletoraise emotionalconcernfor victimsoflarge-scale catastrophesbyvivid descriptionsofaffectedsingleindividuals.

Fig.7. Preferencesforlifesavingprojects.

Narratives,personalstoriesofvictims,andvisualimagesgenerally motivatehelpingresponsesmorethanusingabstractnumerical ﬁgures.Whenfacedwithlargenumbersthataredifﬁcultto men-tallyrepresent,peoplemaysometimesimaginearepresentative and more vivid prototype that keeps their emotions engaged (Kahneman&Frederick,2005).Whilesomeevidenceexiststhat affect-rich stimuli tend to decrease scope sensitivity (Hsee & Rottenstreich,2004;butseeGong&Baron,2011),inthecontextof lifesavingitmightbeusefultoenhanceaffectiveprocessing. Emo-tionsbasedonmoralintuitionscanbeagoodguideinvaluations whenonlyfewlivesareatrisk,andiftheseintuitionsdonotscale upwellotherstepsneedtobetakentokeepemotionsaroused whenmanylivesareatrisk(Slovicetal.,2013).

Otherattemptstoincreasescopesensitivityarebasedonmore carefuldeliberationandreflection.Thisshouldespeciallybethe casefor publicpolicydecisionsthat arecomplexanddeal with multipleconflictingobjectivesbesideslifesaving.Perhapsthemost directwayofcounteractingscopeinsensitivitycouldbetoincrease individualandpublicawarenessofitsexistenceandinfluenceon decisions.However,itislikelythatmerelyinformingpeopleabout thenegativeeffectsofintuitivevaluationsongivingisnotsufficient.

Smalletal.(2007)foundthat debiasingparticipantsby inform-ingabouttheidentifiedvictimeffectactuallydecreaseddonations to theidentified victim rather than increase donations toward unidentifiedstatisticalvictims.Anotherwaytoengagedeliberative

(7)

processeswouldbetomakethenumbersoflivesatriskmore evalu-able(Slovicetal.,2013).Thiscouldincludepresentingcomparative statisticsofvictimnumbersaswellasthepredictedimpactofa humanitarianaidintervention.Statisticscanbebroughttolifeby makingthemmoredescriptive,splittingnumberof victimsinto women,children,elderly,victimsofassault,etc.Amoreindirect waytoincreasescopesensitivityisbasedonthefactthatpeople liketobeconsistentintheirevaluations(Ariely,Loewenstein, & Prelec,2003),whichcanleadtoamoredeliberativeapproachto valuationsbymeansofcalculations.Forexample,Hsee,Zhang,Lu, andXu(2013)proposethatinitialscopeinsensitivitycanbeturned intoscopeconsistencybyﬁrstaskingaboutvaluationsofasingle lifeandthenaboutvaluationsofmultiplelives.Providinga valua-tionforonelifecanleadpeopletorealizethattheirvaluationforN livesisnotlinearandcorrectforthisbyincreasingtheirvaluations. Anotherrecentapproachfoundthatwhilepeopleareinsensitive tothenumberoflivesatrisk,theymaynotbeasinsensitiveto thenumberoffatalitiesthathavealreadyoccurred(Evangelidis& vandenBergh,2013).Thiscouldimplythatthenumberoffatalities informsdecisionmakersabouttheseverityofthehumanitarianaid crisismorethanthenumberoflivesatriskdoes.

Atthedecision-makinglevelofgovernmentagenciesandNGOs, otherinstitutionalchangesandrationalorganizationalstructures mayalsoincreasescopesensitivityandhelptoavoidfallingpreyto thelimitsofmoralintuitionswhenmanylivesareindanger.Public policyismorethanasimplereactiontowardissuesrelatedtorisks. UsuallygovernmentagenciesandNGOshavetochoosebetween manydifferentcoursesofactionandevaluatedifferenttradeoffs, allofwhichshouldbedonereflectively.Forexample,organizations canmakeuseofcost-benefitanalysesorimplementathreshold modelwhichautomaticallytriggersactionwhenaspecific num-beroflivesareatrisk(Slovicetal.,2013).Itisnotalwaysclear, however,howcostsandbenefitsshouldbemeasured,calculated andweightedindecisions.Possiblecostsincludemoney,time,the livesofrescueworkers,internationalpoliticalconsequences,and othernegativeimplicationsof(in)actions.Benefitsinclude(among others)theimprovementinnutrition,healthandsafety,and gen-eralstandardofliving.6_If_valuations_of_these_dimensions_are_done

adhocandintuitively,theymaybeinconsistentandbiased(e.g.,

Kahneman,Slovic,&Tversky,1982).Requiringofﬁcialstogive rea-sonsforinactionaswellasactioninthefaceofhumanitariancrises wouldencourageamorereﬂectiveapproachtopublicpolicyand couldreducetheseinconsistencies.Decisionanalysisandusinga consequencematrixforclarifyingtheinterventiontradeoffscould alsobeausefultooltoreachthisgoal(Slovic,Västfjäll,&Gregory, 2012).However,calibratingtherightlevelofaidresponsesis a particularchallengeforpublicpolicy.Someevidenceexiststhat althoughknowledgeandcommitmenttoproblemswithina par-ticulardomaincanreducebiasesinvaluations,itmaynotincrease scopesensitivity(Markowitz,Slovic,Västfjäll,&Hodges,2013).

Adecisionstrategythatcanpotentiallyhelporganizationsavoid aninefficientallocationoffundsforcharitableprojectswouldbeto usetheaverageimpact(i.e.,benefit)aswellasthemarginalbenefit ofeachadditionaldollarinvested(Baron&Szymanska,2011).Ifthe goalistohelpasmanypeopleaspossible(orincreasewelfaretothe maximumextent),thevaluationoflivescanbecapturedasa util-itymaximizingproblemwithintheframeworkofmostformsof utilitarianism.Accordingtothisperspective,toassurethatfunds areusedinthebestpossiblewayorganizationsshouldusethem tomaximizethewelfareofallpeopleaffected.Humanitarianaid projectsthathelpmorepeopleatthesameorlowercostofless

6_The_tradeoff_between_costs_and_beneﬁts_obviously_includes_many_more

dimen-sions.Forexample,age,income,andnationalitymayalsobeconsideredwhen decidinghowtoallocatefundingtopeopleinneed(e.g.,Lietal.,2010).

efficientprojectsshouldthereforebepreferredbecausetheyhave ahigheraverageimpact.Conversely,theideaofusingthemarginal benefit per additional dollar invested is based onthe observa-tionthatcharitableorganizationsandprojectshavespecifictarget fundinglevels,atwhichtheavailablefundsareusedoptimally. Additionalfundingmightdomoregoodinprojectsororganizations thathavenotreachedthisoptimallevel.

2. Concludingremarks

Inthis article,we presentedfourdifferentwaystovaluethe livesofothersandarguedthatintuitivevaluationscanleadtoscope insensitivity,whichrepresentsadeviationfromnormative valua-tions.Animportantassumptioninthisdepictionofvaluationsis thattheyunderlieand guidedecisionsregardingthewelfareof others.Itispossible,ofcourse,thatvaluationsoflifearenotalways reflectedintheamountofhelp(i.e.,money,food,medicine, educa-tion,etc.)giventopeopleinneed.Moreover,animplicitassumption inmostresearchontheeffectsofvictimnumbersondonationsis thattheeffectivenessofanadditionaldollarspentremains con-stantregardlessofthetotalnumberofpeopleinneed.Itislikely thatthisisnotalwaysthecase.Forexample,iffundingisusedto buildawelltoprovidecleanwater,theeffectivenessofthe spend-ingdependsonhowmanypeoplebenefitfromit.Also,certaincosts remainfairlyconstantregardlessoftheamountofpeopleinneed (e.g.,thetransportationcostsoffoodforoneorone-hundred peo-ple).Inordertoefficientlyallocateresources,charityorganizations andinstitutionsneedtoconsiderthesefixedcostsintheir calcu-lationsofhowmuchalifeisworthtothem.Indeed,thevaluation oflivesisonlyoneconsiderationinresponsestorisk(Fischhoff, Lichtenstein,Slovic,Derby,&Keeney,1981).Someevidenceexists thatpeopletendtoevaluateprogramstosavehumanlivesbytheir cost-effectiveness(e.g.,thenumberoflivessavedpermoneyunit), butonlyifthisdimensioniseasytoevaluate(Caviola,Faulmüller, Everett,Savulescu,&Kahane,2014).

Whenonlylookingatthenumbersofaffectedpeople,itis diffi-culttounderstandwhysomehumanitariancatastrophesgeneratea greateraidresponsethanothers.Forexample,in2010the human-itarianaidresponse intheaftermathoftheearthquakein Haiti (which affectedan estimated3 millionpeople) wasmore than 3billiondollars.In thesame year,theflood inPakistan(which affectedanestimated20millionpeople)receivedonly2.2billion dollars(GlobalHumanitarianAssistance,2013).Thevaluationofa singlelifewas10timeshigherinHaitithaninPakistan.Itis,of course,possiblethataidinPakistanwasmoreavailableandmore efficientinhelpingaffectedpeoplethaninHaiti,whichcautions againstusingonlythevolumeoffinancialcontributionsasaproxy forvaluations.

Perhapslikefewotherdecisionmakingdomains,thevaluation oflivesisinherentlyintertwinedwithissuesofmorality, norma-tivityandrationality.Thesevaluationsareoftenguidedbymoral intuitionswhichare,inturn,inﬂuencedmorebyemotionsrather thanrationalconsiderations(Haidt,2007).Inthespeciﬁccaseof valuationsunderlyingpublicpolicydecisions,onewouldexpect that each individuallife at riskshould begiven thesame con-siderationand value,whichis a moralprincipletowhich most individualsinwesterncountrieswouldprobablyagreeto. Nonethe-less,intuitivetradeoffsandthelimitsofmoralintuitionsunderlying scopeinsensitivityin lifesavingcontextscanoftenleadto non-normativeandirrationalvaluations(Reyna&Casillas,2009).We wouldalsoliketopointoutthatdeviationsfromnormativemodels canpresentaparticularproblemforpublicpolicywhenithasto jus-tifythereasonsforallocationdecisionstothepublic.Itisnotalways clearwhichnormativestandardsshouldguidedecisions,though. Forexample,itispossibletoargueforproportiondominanceto

(8)

Futureresearchshouldaimatgeneratingabetter understand-ingofthereasonsandmechanismsofscopeinsensitivityandways toreduce it.Thesearchfor scopesensitivitycouldlookat how tradeoffsbetweenlivesandmoneyaredoneincomplexpublic pol-icydecisionsregardingthewelfareofothers.Oneinterestingway forwardwouldbetomodeltheaffectiveandcognitiveprocesses underlyingsuchvaluations.Giventhelinktoperceptualtheories

Conﬂictofinterest

Theauthorsdeclarenoconﬂictofinterest.

AppendixA. Materials

Instructions:

Pleaseimaginethatyouworkforacharityorganizationandare responsiblefordistributingdonationsinavillageinAfricainwhich 100peoplelivethatfaceaseverefoodshortage.Youhaveupto5000 Eurosavailablebutdonotexactlyknowhowmanyofthe100 peo-pleareaffectedandneedhelp.Threeprogramstocombatthefood shortageareavailable(ProgramA–C).Theydifferintheamount ofmoneydistributeddependingonhowmanypeopleareatrisk ofstarvation.Yourtaskistochooseoneoftheseprograms.Each ofthemwillalleviatethehungeroftheaffectedpeople(although notnecessarilytothesameextent)andcostsdifferentamountsof money,dependingonthenumberofpeopleinneedofhelp.

References

Andreoni,J.(2007).Givinggiftstogroups:Howaltruismdependsonthenumberof recipients.JournalofPublicEconomics,91,1731–1749.

Ariely,D.,Loewenstein,G.,&Prelec, D. (2003).Coherentarbitrariness:Stable demandcurveswithoutstablepreferences.QuarterlyJournalofEconomics,118, 73–106.

Baron,J.,&Szymanska,E.(2011).Heuristicsandbaisesincharity.InD.M. Oppen-heimer,&C.Y.Olivola(Eds.),Thescienceofgiving:Experimentalapproachesto thestudyofcharity(pp.215–235).NewYork,NY:PsychologyPress,US. Bartels,D.M.(2006).Proportiondominance:Thegeneralityandvariabilityof

favor-ingrelativesavingsoverabsolutesavings.OrganizationalBehaviorandHuman DecisionProcesses,100,76–95.http://dx.doi.org/10.1016/j.obhdp.2005.10.004 Bartels,D.M.,&Burnett,R.C.(2011).Agroupconstrualaccountof

drop-in-the-bucketthinkinginpolicypreferenceandmoraljudgment.JournalofExperimental SocialPsychology,47,50–57.http://dx.doi.org/10.1016/j.jesp.2010.08.003 Batson,C.D.(1990).Howsocialananimal?Thehumancapacityforcaring.American

Psychologist,45,336–346.

theextenttowhichexperts(i.e.,publicpolicymakers)can pro-tectthemselvesagainstintuitivevaluationsofhumanlives.Scope insensitivitypresentsauniquechallengeforpublicpolicyand man-agersofgovernmentandnon-governmentorganizations.Itisour hopethatresearchwillprovidesomeanswerstoguidepublic pol-icydecisionsandreducescopeinsensitivityinthisimportantrisk domain.

ProgramA(Fig.1):

Heredonationswillbedistributedinawaysuchthattheamount increasesataconstantratethemoreneedypersonsexist.

ProgramB(Fig.3):

Heredonationswillbedistributedinawaysuchthattheamount increasesthemoreneedypersonsexist,butatadecreasingrate.

ProgramC(Fig.4):

Heredonationswillbedistributedinawaysuchthattheamount ishighifonlyfewneedypersonsexistbutdecreasesoverallifthe numberofneedypersonsincreases.

Pleaseindicatehowmuchyouagreetoeachofthese statements:

Donot agree

Agreevery much 1.Donationsshouldgenerallybe

distributedaccordingtoProgramA

1 2 3 4 5 6 7

2.Donationsshouldgenerallybe distributedaccordingtoProgramB

1 2 3 4 5 6 7

3.Donationsshouldgenerallybe distributedaccordingtoProgramC

1 2 3 4 5 6 7

Ifyouhadtochooseoneoftheseprojects,whichonewouldyouchoose?

OProgramA OProgramB OProgramC

Inyouropinion,whichoftheseProgramsdescribesactualallocationofdonationsbyNGOsintheabovesituationbest?

OProgramA OProgramB OProgramC

Pleasenowimagineasituationinwhichyouwanttodonatetoahumanitariancause.Pleaseindicatewhichofthe followingtwoprojectsyouwouldratherdonateto.Bothoftheseprojectsaredesignedtosupportaspeciﬁcnumber ofpeoplewithfoodwhoareatriskofstarvation.

䊐ProjectAsaves102peopleof115peopleatriskofstarvation. 䊐ProjectBsaves105peopleof247peopleatriskofstarvation.

Batson, C. D., O’Quin, K., Fultz, J., Vanderplas, M., & Isen, A. (1983). Self-reported distress and empathy and egoistic versus altristic moti-vation for helping. Journal of Personaly and Social Psychology, 45(706), 718.

Cameron,C.D.,&Payne,B.K.(2011).Escapingaffect:Howmotivatedemotion reg-ulationcreatesinsensitivitytomasssuffering.JournalofPersonalityandSocial Psychology,100,1–15.

Caviola,L.,Faulmüller,N.,Everett,J.A.C.,Savulescu,J.,&Kahane,G.(2014).The evalu-abilitybiasincharitablegiving:Savingadministrationcostsorsavinglives? JudgmentandDecisionMaking,9,303–315.

Cryder,C.E.,Loewenstein,G.,&Scheines,R.(2013).Thedonorisinthedetails. OrganizationalBehaviorandHumanDecisionProcesses,120,15–23.

Dickert,S.,&Slovic,P.(2009).Attentionalmechanismsinthegenerationof sympa-thy.JudgmentandDecisionMaking,4,297–306.

Dickert,S.,Kleber,J.,Peters,E.,&Slovic,P.(2011).Numericabilityasaprecursor topro-socialbehavior:Theimpactofnumeracyandpresentationformatonthe cognitivemechanismsunderlyingdonations.JudgmentandDecisionMaking,6, 638–650.

(9)

Dickert,S.,Sagara,N.,&Slovic,P.(2011).Affectivemotivationstohelpothers:A two-stagemodelofdonationdecisions.JournalofBehavioralDecisionMaking, 24,361–376.

Dickert,S.,Västfjäll,D.,Kleber,J.,&Slovic,P.(2012).Valuationsofhumanlives: Nor-mativeexpectationsandpsychologicalmechanismsof(ir)rationality.Synthese, 189,95–105.

Erlandsson,A.,Björklund,F.,&Bäckström,M.(2013).Perceivedutility(not sympa-thy)mediatestheproportiondominanceeffectinhelpingdecisions.Journalof BehavioralDecisionMaking,http://dx.doi.org/10.1002/bdm.1789

Evangelidis, I.,& vanden Bergh, B. (2013). The number of fatalities drives disasteraid:Increasingsensitivity topeopleinneed.PsychologicalScience, http://dx.doi.org/10.1177/0956797613490748

Evans,J. (2008).Dual-processing accounts of reasoning,judgment, andsocial cognition. Annual Review of Psychology, 59, 255–278. http://dx.doi.org/ 10.1146/annurev.psych.59.103006.093629

Fetherstonhaugh,D.,Slovic,P.,Johnson,S.M.,&Friedrich,J.(1997).Insensitivityto thevalueofhumanlife:Astudyofpsychophysicalnumbing.JournalofRiskand Uncertainty,14,283–300.

Fischhoff,B.,Lichtenstein,S.,Slovic,P.,Derby,S.L.,&Keeney,R.L.(1981).Acceptable risk.Cambridge,UK:CambridgeUniversityPress.

Genevsky,A.,Västfjäll,D.,Slovic,P.,&Knutson,B.(2013).Neuralunderpinnings oftheidentiﬁablevictimeffect:Affectshiftspreferencesforgiving.Journalof Neuroscience,33,17188–17196.

GlobalHumanitarianAssistance.(2013).GHAReport2012.

Gong,M., &Baron,J. (2011).Thegenerality oftheemotion effecton magni-tudesensitivity.JournalofEconomicPsychology,32,17–24.http://dx.doi.org/ 10.1016/j.joep.2010.10.002

Grant,A.M.,&Schwartz,B.(2011).Toomuchofagoodthing:Thechallenge andopportunityoftheinvertedU.PerspectivesonPsychologicalScience,6, 61–76.

Haidt,J.(2007).Thenewsynthesisinmoralpsychology.Science,316,998–1002. Hamilton,D.L.,&Sherman,S.J.(1996).Perceivingpersonsandgroups.Psychological

Review,103,336–355.

Hsee,C. K.(1996).Theevaluabilityhypothesis:Anexplanationforpreference reversalsbetweenjointandseparateevaluationsofalternatives.Organizational BehaviorandHumanDecisionProcesses,67,247–257.

Hsee,C. K. (1998).Less is better: Whenlow-value options are valuedmore highlythan high-value options.Journal of BehavioralDecision Making, 11, 107–121.

Hsee,C.K.,&Rottenstreich,Y.(2004).Music,pandas,andmuggers:Onthe affec-tivepsychologyofvalue.JournalofExperimentalPsychology-General,133,23–30. http://dx.doi.org/10.1037/0096-3445.133.1.23

Hsee,C.K.,&Zhang,J.A.(2010).Generalevaluabilitytheory.Perspectiveson Psycho-logicalScience,5,343–355.http://dx.doi.org/10.1177/1745691610374586 Hsee,C.K.,Zhang,J.,Lu,Z.Y.,&Xu,F.(2013).Unitasking:Amethodtoboost

donationsandbeyond.PsychologicalScience,24,1801–1808.

Kahneman, D. (2003). A perspective on judgment and choice: Mapping bounded rationality. American Psychologist, 58, 697–720. http://dx.doi.org/ 10.1037/0003-066x.58.9.697

Kahneman,D.,&Frederick,S.(2005).Amodelofheuristicjudgment.InK.J.Holyoak, &R.G.Morrison(Eds.),TheCambridgehandbookofthinkingandreasoning(pp. 267–293).CambridgeUniversityPress.

Kahneman,D.,&Tversky,A.(1979).Prospecttheory:Analysisofdecisionunderrisk. Econometrica,47,263–291.http://dx.doi.org/10.2307/1914185

Kahneman,D.,Slovic,P.,&Tversky,A.(1982).Judgmentunderuncertainty:Heuristics andbiases.NewYork:CambridgeUniversityPress.

Kogut,T.,&Ritov,I.(2005a).The“Identiﬁedvictim”effect:Anidentiﬁedgroup, orjustasingleindividual?JournalofBehavioralDecisionMaking,18,157–167. http://dx.doi.org/10.1002/bdm.492

Kogut,T.,&Ritov,I.(2005b).Thesingularityeffectofidentiﬁedvictimsinseparate andjointevaluations.OrganizationalBehaviorandHumanDecisionProcesses,97, 106–116.http://dx.doi.org/10.1016/j.obhdp.2005.02.003

Kogut,T.,&Ritov,I.(2007).“Oneofus”:Outstandingwillingnesstohelpsavea sin-gleidentiﬁedcompatriot.OrganizationalBehaviorandHumanDecisionProcesses, 104,150–157.http://dx.doi.org/10.1016/j.obhdp.2007.04.006

Li,M.,Vietri,J.,Galvani,A.P.,&Chapman,G.B.(2010).Howdopeoplevaluelife? Psy-chologicalScience,21,163–167.http://dx.doi.org/10.1177/0956797609357707 Lichtenstein,S.,&Slovic,P.(2006).Theconstructionofpreference.Cambridge

Uni-versityPress.

Loewenstein,G.,&Small,D.A.(2007).Thescarecrowandthetinman:The vicissi-tudesofhumansympathyandcaring.ReviewofGeneralPsychology,11,112–126. http://dx.doi.org/10.1037/1089-2680.11.2.112

Markowitz,E.M.,Slovic,P.,Västfjäll,D.,&Hodges,S.D.(2013).Compassionfadeand thechallengeofenvironmentalconservation.JudgmentandDecisionMaking,8, 397–406.

Reyna,V.F.,&Casillas,W.(2009).Developmentanddualprocessesinmoralreasoning: Afuzzy-tracetheoryapproach.In:B.H.Ross(SeriesEd.)&D.M.Bartels,C.W. Bauman,L.J.Skitka,&D.L.Medin(Eds.),Psychologyoflearningandmotivation. vol.50,Moraljudgmentanddecisionmaking(pp.207–239).SanDiego,CA: ElsevierAcademicPress.

Rubaltelli,E.,&Agnoli,S.(2012).Theemotionalcostofcharitabledonations. Cogni-tion&Emotion,26,769–785.

Sinnott-Armstrong,W.(2011). Consequentialism.In E.N.Zalta(Ed.),Stanford encyclopediaofphilosophy.Stanford:StanfordUniversity.http://www.plato. stanford.edu/archives/win2011/entries/consequentialism

Slovic,P.(2007).IfIlookatthemassIwillneveract”:Psychicnumbingandgenocide. JudgmentandDecisionMakingJournal,2,79–95.

Slovic,P.(2010).Themorewhodie,thelesswecare.InE.Michel-Kerjan,&P.Slovic (Eds.),Theirrationaleconomist:Makingdecisionsinadangerousworld(pp.30–40). NewYork:PublicAffairs.

Slovic,P.,&Västfjäll,D.(2010).Affect,moralintuition,andrisk.PsychologicalInquiry, 21,387–398.http://dx.doi.org/10.1080/1047840x.2010.521119

Slovic,P.,Fischoff,B.,&Lichtenstein,S.(1982).Responsemode,framing,and infor-mationprocessingeffectsinriskassessment.InR.Hogarth(Ed.),Newdirections formethodologyofsocialandbehavioralscience:Questionframingandresponse consistency(vol.11).SanFrancisco:Jersey-Bass.

Slovic,P.,Finucane,M.,Peters,E.,&MacGregor,D.G.(2002).Theaffectheuristic.In T.Gilovich,D.Grifﬁn,&D.Kahneman(Eds.),Heuristicsandbiases:Thepsychology ofintuitivejudgment(pp.397–420).NewYork:CambridgeUniversityPress. Slovic,P.,Västfjäll,D.,&Gregory,R.(2012).Informingdecisionstopreventgenocide.

SAISReview,32,33–47.

Slovic,P.,Zionts,D.,Woods,A.K.,Goodman,R.,&Jinks,D.(2013).Psychicnumbing andmassatrocity.InE.Shaﬁr(Ed.),Thebehavioralfoundationsofpublicpolicy (pp.126–142).NJ:PrincetonUniversityPress.

Small,D.A.(2010).Reference-dependentsympathy.OrganizationalBehaviorand HumanDecisionProcesses,112,151–160.

Small,D.A.,Loewenstein,G.,&Slovic,P.(2007).Sympathyandcallousness:The impactofdeliberativethought ondonationsto identiﬁableandstatistical victims.OrganizationalBehaviorandHumanDecisionProcesses,102,143–153. http://dx.doi.org/10.1016/j.obhdp.2006.01.005

Smith,R.W.,Faro,D.,&Burson,K.A.(2013).Moreforthemany:Theinﬂuenceof entitativityoncharitablegiving.JournalofConsumerResearch,39,961–976. Stanovich,K.E.,&West,R.F.(2000).Individualdifferencesinreasoning:

Impli-cationsfortherationalitydebate?BehavioralandBrainSciences,23,645–665. http://dx.doi.org/10.1017/s0140525x00003435

Taurek,J.M.(1977).Shouldthenumberscount?PhilosophyandPublicAffairs,6, 293–316.

Västfjäll,D.,Slovic,P.,Mayorga,M.,&Peters,E.(2014).Compassionfade:Affect andcharityare greatestforasinglechildinneed.PLOSONE, 9,e100115. http://dx.doi.org/10.1371/journal.pone.0100115

Västfjäll,D.,Slovic,P.,&Mayorga,M.(2014).Whoeversavesonesavestheworld: Confrontingthechallengeofpseudo-inefﬁcacy.(submittedforpublication). vonNeumann,J.,&Morgenstern,O.(1944).Theoryofgamesandeconomicbehaviour.