Behavioral Economics and Perverse Effects of The Welfare State
Behavioral Economics and Perverse Effects of The Welfare State
JEL Classifications: I3, D6, H3, D1 Key or!s: "e#a$ioral econo%ics, elfare state, po$erty
Abstract: Critics often ar&ue t#at &o$ern%ent po$erty pro&ra%s per$ersely %a'e t#e poor orse off "y encoura&in& une%ploy%ent, out(of( e!loc' "irt#s, an! ot#er )social pat#olo&ies*) Ho e$er, "asic %icroecono%ic t#eory tells us t#at you cannot %a'e an a&ent orse off "y e+pan!in& #is c#oice set* ,#e current paper ar&ues t#at fa%iliar fin!in&s in "e#a$ioral econo%ics can "e use! to resol$e t#is para!o+* Insofar as t#e stan!ar! rational actor %o!el is ron&, a!!itional c#oices can %a'e a&ents orse off* -ore i%portantly, e+istin& e%pirical e$i!ence su&&ests t#at t#e poor !e$iate fro% t#e rational actor %o!el to an unusually lar&e !e&ree* ,#e paper t#en consi!ers t#e policy i%plications of our alternati$e perspecti$e*
.e oul! li'e to t#an' ,yler Co en, /o"in Hanson, Ke$in -cCa"e, Dan Houser, /on Heiner, se%inar participants at 0eor&e -ason 1ni$ersity, Beloit Colle&e, an! 2e 3or'
*
Scott Beaulier, 4ssistant 5rofessor, Depart%ent of Econo%ics an! -ana&e%ent, Beloit Colle&e, Beloit, .I 63611* E%ail: "eaulies7"eloit*e!u* Bryan Caplan, 4ssociate 5rofessor, Depart%ent of Econo%ics an! Center for Stu!y of 5u"lic C#oice, 0eor&e -ason 1ni$ersity, 8airfa+, 94 ::;3;* E%ail: "caplan7&%u*e!u*
:
1ni$ersity, an! an anony%ous referee for !iscussion an! co%%ents* ,#e stan!ar! !isclai%er applies*
,#e %ost co%pellin& e+planation for t#e %ar'e! s#ift in t#e fortunes of t#e poor is t#at t#ey continue! to respon!, as t#ey al ays #a!, to t#e orl! as t#ey foun! it, "ut t#at e < %eanin& t#e not(poor an! un(!isa!$anta&e! < #a! c#an&e! t#e rules of t#eir orl!* 2ot of our orl!, =ust of t#eirs* ,#e first effect of t#e ne rules as to %a'e it profita"le for t#e poor to "e#a$e in t#e s#ort ter% in ays t#at ere !estructi$e in t#e lon& ter%* ,#eir secon! effect as to %as' t#ese lon&(ter% losses < to su"si!i>e irretrie$a"le %ista'es* .e trie! to pro$i!e %ore for t#e poor an! pro!uce! %ore poor instea!* .e trie! to re%o$e t#e "arriers to escape fro% po$erty, an! ina!$ertently "uilt a trap* ?C#arles -urray, Losing Ground, p* @A
I* I2,/BD1C,IB2
4 recurrin& criticis% of t#e elfare state is t#at it per$ersely #ar%s t#e $ery people it is inten!e! to #elp*1 0i$in& %oney to t#e poor re!uces t#eir incenti$e to enter t#e or'force, acCuire e+perience, an! e$entually =oin t#e %i!!le class* 5ro$i!in& elfare support for c#il!ren "orn out(of( e!loc' encoura&es teen pre&nancy an! !iscoura&es %arria&e, t o serious i%pe!i%ents to escapin& po$erty ?Herrnstein an! -urray, 1@@DE -urray, 1@FDA* 4s -urray /ot#"ar! succinctly puts it, GH,I#e easy a$aila"ility of t#e elfare c#ec' o"$iously pro%otes present(%in!e!ness, un illin&ness to or', an! irresponsi"ility a%on& t#e recipients < t#us perpetuatin& t#e $icious cycle of po$erty( elfare*G ?1@JF, p*16DA ,#ere is a parallel co%plaint a"out affir%ati$e action in #i&#er e!ucation: It alle&e!ly lea!s %inority stu!ents to enroll in o$erly co%petiti$e pro&ra%s* In conseCuence, stu!ents a"le to earn a colle&e !e&ree fro% a less co%petiti$e sc#ool often in! up !roppin& out an! fail to co%plete any !e&ree at all* 4s ,#o%as So ell critically o"ser$es:
,#is H%inorityI stu!ent !oes not &et a "etter e!ucation "ecause #e is at a %ore presti&ious sc#ool* Bn t#e contrary, #e %ay ell &et a %uc# orse e!ucation at suc# fast(pace! #ic# #e is perfectly capa"le of institutions, in t#e sense of failin& to learn t#in&s
learnin&, in a learnin& en$iron%ent t#at procee!s at a nor%al pace* Suc# a %inority stu!ent %ay en! up )confuse!, floun!erin&, an! una"le to 'eep up*) ?1@@3, p* 13JA
1
-urray ?1@FDA is ar&ua"ly t#e %ost pro%inent critic of t#e ).ar on 5o$erty*) 8rie!%an ?1@F: H1@6:IA offere! si%ilar ar&u%ents a&ainst pu"lic #ousin& an! elfare t o !eca!es earlier: t#e &oal of t#ese pro&ra%s as to #elp t#e poor, "ut t#e uninten!e! effect as G=ust t#e re$erseG ?1J@A* So ell ?1@@6 H1@F;I, 1@@;A an! C#a$e> ?:;;:, 1@@:A raise si%ilar o"=ections to affir%ati$e action an! relate! policies*
3 .#ate$er t#e intuiti$e appeal of t#ese sorts of ar&u%ents, t#ey are !ifficult to rationali>e it# stan!ar! %icroecono%ic t#eory* ,o "e %ore precise, te+t"oo' analysis su&&ests t#at eac# of t#e prece!in& clai%s is #alf ri&#t an! #alf ron&* 3es, &i$in& %oney to t#e poor re!uces t#eir incenti$e to enter t#e or'force, acCuire e+perience, an! e$entually =oin t#e %i!!le class t#rou&# t#eir o n efforts* But t#is #ar!ly s#o s t#at it %a'es t#e poor orse offK 4ll t#e transfer pro&ra% !oes is %a'e t#e non( or' option %ore attracti$e* If it re%ains less attracti$e t#an or', t#e poor ill still opt for or'* If, on t#e ot#er #an!, t#e poor !eci!e to ta'e t#e &o$ern%entGs %oney an! re%ain i!le, it %ust "e "ecause < all t#in&s consi!ere! < it %a+i%i>es t#eir e+pecte! lifeti%e utility* In ot#er or!s, if you ere to re%in! elfare recipients t#at t#ey are #urtin& t#eir future =o" prospects, t#ey coul! curtly respon!: )I a% ell(a are of t#at, "ut %y elfare c#ec's %ore t#an co%pensate for %y loss of future inco%e*) 4s ,yler Co en e+plains, t#e Gtra!itional conser$ati$e critiCue of t#e elfare state***G
***in$ol$es an analytic tension* In %ost %atters, conser$ati$es an! li"ertarians ar&ue fro% neoclassical an! C#ica&o sc#ool econo%ic t#eories* In t#ese approac#es, a &ift of cas# al ays %a'es in!i$i!uals "etter off, as e$i!ence! "y t#e classroo% !e%onstration of #o suc# &ifts s#ift in!i$i!uals onto )#i&#er in!ifference cur$es*) ,#is is a "asic lesson of any inter%e!iate course in %icroecono%ics, re&ar!less of t#e political persuasion of t#e instructor* ?:;;:, p*3@A
,#e sa%e &oes for any &o$ern%ent pro&ra% t#at e+pan!s t#e opportunity set of t#e poor* It %ay $ery ell c#an&e t#eir "e#a$ior, "ut ele%entary %icro tells us t#at if you %a'e a !ifferent c#oice after your opportunity set e+pan!s, your utility %ust rise* If affir%ati$e action &i$es un!er(Cualifie! %inority stu!ents t#e c#ance to atten! an I$y Lea&ue sc#ool, t#ey are "etter off if t#ey ta'e a!$anta&e of t#is offer, an! no orse off if t#ey refuse* Bf course, t#is is only true ex anteE a&ents %i&#t co%e to re&ret a c#oice as e$ents unfol!* But t#e stan!ar! assu%ption of "asic %icroecono%ics is t#at a&entsL e+pectations are, on a$era&e, correct* If so%e elfare recipients un!eresti%ate t#e a!$erse effects on t#eir future =o" prospects, t#ere are ot#ers #o o$eresti%ate t#e%* ,#us, for e$ery person #o %ista'enly &oes on elfare, t#ere is anot#er potential elfare recipient #o, contrary to #is o n "est interests, insists on pullin& #is o n ei&#t*
D -icroecono%ic t#eory e%p#asi>es, %oreo$er, t#at c#oices it# i%%e!iate "enefits an! lon&(run costs are still %a!e opti%ally* If a teena&e &irl !eci!es to #a$e an out(of( e!loc' c#il! an! &o on elfare, it is "ecause s#e #as !eter%ine! t#at t#e up(front "enefits of t#e c#il! an! t#e &o$ern%entGs financial assistance out ei&# t#e lon&(run costs of fore&one earnin&s an! !i%inis#e! %arria&e prospects* Basic %icro is a one(si>e( fits(all t#eory of c#oice: ,ra!e(offs "et een t o i%%e!iate "enefits are of one clot# it# tra!e(offs "et een i%%e!iate an! %ore !istant "enefits* .it#in t#e confines of stan!ar! %icroecono%ics, t#en, %any clai%s a"out t#e elfare stateGs per$erse effects cannot "e sustaine!* /at#er t#an re=ect t#ese clai%s as inco#erent, #o e$er, t#e current paper ar&ues t#at e s#oul! ret#in' t#e% usin& t#e tools of "e#a$ioral econo%ics* ?Ka#ne%an, Slo$ic, an! ,$ers'y, 1@F:E ,#aler, 1@@:E /a"in, 1@@FE Sunstein, :;;;A ,#is lar&e an! &ro in& literature casts a !e&ree of e%pirical !ou"t on e$en t#e %ost ele%entary principles of %icroecono%ics* It stron&ly un!er%ines, for e+a%ple, t#e rational e+pectations account of "elief for%ation, unco$erin& an array of syste%atic errors in t#e t#in'in& of t#e a$era&e su"=ect* ?/a"in, 1@@FA It also unco$ers a $ariety of interte%poral ano%alies, suc# as self(control pro"le%s* ?BLDono&#ue an! /a"in, 1@@@a, 1@@@"E Kir"y an! Herrnstein, 1@@6E 4'erlof, 1@@1E ,#aler an! S#efrin, 1@F1A* ,#e per$erse effects freCuently attri"ute! to t#e elfare state are easy to interpret fro% a "e#a$ioral perspecti$e* If people o$eresti%ate t#e %a&nitu!e of i%%e!iate "enefits relati$e to %ore !istant ones, you can actually < on net < #ar% t#e% "y offerin& t#e% a!!itional i%%e!iate "enefits* ,#ey alrea!y ten! to un!er(in$est* -a'in& t#eir present %ore li$a"le it# cas# &ifts only a%plifies t#is ten!ency* Si%ilarly, if in!i$i!uals syste%atically o$er(esti%ate t#eir o n a"ilities, you coul! easily #ar% a stu!ent "y a!%ittin& #i% to a pro&ra% for #ic# #e is un!er(Cualifie!* Blin!e! "y o$er(confi!ence, #e oul! "e li'ely to select t#e "est sc#ool t#at accepte! #i%, scarcely consi!erin& t#e possi"ility t#at #e ill "e out of #is lea&ue*
6 Loo'in& at t#e elfare state fro% a "e#a$ioral stan!point lays t#e &roun! or' for a stron&er clai%: 5otential elfare recipientsG !e$iations fro% neoclassical assu%ptions ten! to "e especially pronounce!* If t#e a$era&e 4%erican falls s#ort of t#e neoclassical i!eal, t#e a$era&e recipient of &o$ern%ent assistance !oes not e$en co%e close* ,o =ustify t#is &enerali>ation, e !ra on t#e lar&e literature on )pat#olo&ical) "e#a$ior a%on& t#e poor* -any pat#olo&ies can "e rea!ily un!erstoo! as e+tre%e $ersions of t#e ano%alies e%p#asi>e! in t#e "e#a$ioral literature* ,#e e+peri%ental literature fin!s, for e+a%ple, t#at t#e a$era&e e+peri%ental su"=ect un!eresti%ates future costs of i%%e!iate satisfactionE an! e$en #en t#ey reco&ni>e t#ese costs, self(control pro"le%s %ay 'eep t#e% fro% c#an&in& t#eir "e#a$ior ?,#aler, 1@@:A* But t#e a$era&e e+peri%ental su"=ectGs !e$iations see% %il! co%pare! to t#ose typical of t#e poor* -any acti$ities < fro% o$er(eatin&, !rin'in&, s%o'in&, an! !ru& a"use to cri%e an! unprotecte! se+ < co%"ine i%%e!iate &ratification it# !elaye! costs* .e ar&ue t#at it is no coinci!ence t#at t#e poor are %uc# %ore prone to en&a&e in suc# acti$ities t#an t#e rest of t#e population* ,#us, "e#a$ioral econo%ics can rein$i&orate ar&u%ents a"out t#e per$erse effects of t#e elfare state in t o !istinct ays* 8or starters, e nee! "e#a$ioral econo%ics to %a'e t#is critiCue co#erent* Bnce t#ey #a$e "een e+plicitly place! on a "e#a$ioral foun!ation, clai%s a"out t#e per$erse effects of $arious pro&ra%s "eco%e #ar!er to i&nore or !is%iss* -ore i%portantly, t#ere are &oo! e%pirical reasons to t#in' t#at "e#a$ioral econo%ics "etter !escri"es t#e poor t#an it !oes t#e rest of t#e population* Be#a$ioral econo%ics is t#erefore e+ceptionally rele$ant to po$erty policy* ,#e paper is structure! as follo s* Section t o !iscusses pre$ious literature on t#e per$erse effects of t#e elfare state, #i&#li&#tin& its t#eoretical ea'nesses* Section t#ree s#o s #o a $ariety of fin!in&s fro% "e#a$ioral econo%ics put #at Co en calls t#e Gtra!itional conser$ati$e critiCueG of t#e elfare state in a ne an! fa$ora"le li&#t* Section four presents e$i!ence t#at "e#a$ioral econo%ics offers consi!era"ly more insi&#t into potential elfare recipients t#an it !oes for t#e rest of t#e population* Section
6 fi$e e+plores t#e counter(intuiti$e policy i%plications of our analysis* Section si+ conclu!es*
II* 5E/9E/SE E88EC,S B8 ,HE .EL84/E S,4,EM
-any clai%s a"out t#e ne&ati$e i%pact of &o$ern%ent policies on t#e poor can "e rea!ily parse! in neoclassical ter%s* ,o ta'e a stan!ar! e+a%ple, t#e ric# li$e lon&er t#an t#e poor, an! start or' later in life, so t#e &ross "enefit t#e poor recei$e fro% Social Security an! -e!icare is relati$ely s%all* Since payroll ta+es are re&ressi$e, t#e net "enefit of Social Security for t#e poor %i&#t turn out to "e ne&ati$e* Si%ilarly, e$en if t#e poor recei$e t ice as %any !ollars in &o$ern%ent %e!ical care as t#ey pay in ta+es to fun! suc# pro&ra%s, t#ey are still orse off if t#ey $alue a !ollarGs ort# of %e!ical care at less t#an fifty cents* 4r&u%ents a&ainst t#e %ini%u% a&e an! rent control fit t#e sa%e pattern: t#e "enefits of #i&#er a&es or lo er rents coul! easily "e less t#an t#e e+pecte! costs of rationin&* ,#is is strai&#t out of t#e te+t"oo'* Less !irect co%plaints a"out #o policy #urts t#e poor can "e rationali>e! in neoclassical ter%s as e+ternalities* Bne poor fa%ily ill "e "etter off if it ta'es &o$ern%ent assistance* But its %e%"ers can "e orse off on "alance "ecause ot#er fa%ilies in t#eir area participate in t#e sa%e pro&ra%* ,#eir nei&#"or#oo! "eco%es a orse place to li$e as t#e %i+ of local acti$ity s#ifts fro% pro!ucti$e or' to !ru& use or cri%e* ,#ere coul! also "e intra(fa%ily e+ternalities* 4 fat#er illin& to support #is fa%ily %i&#t a"an!on it if t#e &o$ern%ent ill assu%e #is responsi"ilities* ,#e fat#er is "etter off as a result of t#e &o$ern%ent pro&ra%, "ut #is ife an! fa%ily are orse off if t#ey prefer t#e presence of a or'in& fat#er to a strea% of &o$ern%ent c#ec's* ,#is paper focuses on a t#ir!, Cuite !ifferent, cate&ory of ar&u%ents* ,#ese %a'e t#e !ra%atic clai% t#at &o$ern%ent &ifts by themselves %a'e t#eir recipients orse off* ,#e si%plest $ersion is t#at "y &i$in& t#e poor %aterial support, e !iscoura&e t#e% fro% &ettin& =o"s, acCuirin& e+perience, an! e$entually pullin& t#e%sel$es up "y t#eir "ootstraps* In So ellGs or!s, G8or youn& or'ers especially, t#e t#in&s you can learn on H)%enial)I =o"s < responsi"ility, cooperation, punctuality < can "e lifelon& assets in
J %any ot#er occupations* Insulatin& people fro% suc# realities is one of t#e %any cruelties perpetuate! un!er t#e "anner of )co%passion*)G ?1@FJE p* 36A .#at appears to "e altruistic assistance %a'es t#e recipient orse off "y %a'in& #i% une%ploya"le in t#e future* In t#e si%plest %o!el of c#oice un!er certainty, t#is ar&u%ent is in !irect conflict it# re$eale! preference* ?8i&ure 1A 2o one is force! to ta'e t#e %oney* If t#e future sacrifices out ei&# t#e present "enefit, t#e prospecti$e elfare recipient can si%ply turn t#e %oney !o n* 4n e+tra option %ust ea'ly increase #is utility* 4!!in& uncertainty a!%itte!ly opens up t#e possi"ility t#at you select a ne ly a$aila"le option t#at or's out "a!ly for you* But #ere t#e fa%iliar !istinction "et een e+ post an! e+ ante utility co%es into play* .e can reconceptuali>e a&ents as c#oosin& gambles instea! of outcomes* ,#en puttin& an e+tra c#oice on your %enu of &a%"les %ust ea'ly increase your expected utility* Bf course, if your e+pectations are co%pletely at o!!s it# t#e facts, t#is result is far ea'er t#an it soun!s* Stan!ar! neoclassical practice, t#ou&#, is to assu%e rational e+pectations ?S#effrin, 1@@6E 5esaran, 1@FJA* By i%plication, t#e pro"a"ilities t#at a&ents assi&n to outco%es are, on a$era&e, correct* So #en a&ents expect to "enefit "y selectin& a ne option, on a$era&e t#ey really do "enefit* In su%, t#ere are t#ree %ain fa%ilies of criticis% of t#e elfare state* Bne is t#at t#e poor so%eti%es pay %ore in ta+es t#an t#ey &et in "enefitsE anot#er appeals to inter( or intra(fa%ily e+ternalities* ,#ese are not t#e su"=ect of our paper* .e focus our attention on a t#ir! fa%ily of criticis%, #ic# "la%es &o$ern%ent for so%e#o lea!in& its )"eneficiaries) astray* 4r&u%ents of t#is for% are r#etorically persuasi$e in spite of t#eir inco%pati"ility it# "asic %icro* Is t#ere any ay to %a'e sense of t#e%M
III* 4 BEH49IB/4L /E,HI2KI20 B8 ,HE ,/4DI,IB24L CB2SE/94,I9E C/I,IN1E
In!ee! t#ere is* ,#e $olu%inous "e#a$ioral literature su"=ects t#e fun!a%ental propositions of te+t"oo' %icroecono%ics to e+tensi$e e%pirical tests, an! !ocu%ents a
F i!e $ariety of e+ceptions* It also offers alternati$e t#eories of #u%an "e#a$ior to acco%%o!ate t#e e%pirical fin!in&s ?Ka#ne%an an! ,$ers'y, 1@J@E Ka#ne%an et al*, 1@@;E Lai"son, 1@@JE Ca%erer an! ,#aler, 1@@6A* 3et in spite of t#e %aturity of t#e "e#a$ioral econo%ics literature, "e#a$ioral econo%ics #as al%ost ne$er "een use! to ret#in' elfare policy* : ,#is is unfortunate "ecause "e#a$ioral econo%ics offers a no$el perspecti$e in t#is %uc#(!e"ate! area* ,#ere are t o %ain a$enues e oul! #ere li'e to e+plore*3 ,#e first is judgmental biases* ,#e secon! is self-control problems*
1. Judgmental Biases
Ju!&%ental "iases are ten!encies to #ol! syste%atically %ista'en "eliefs, "eliefs pre!icta"ly at o!!s it# esta"lis#e! e%pirical facts or it# principles of lo&ic, %at#e%atics, or statistics* ?Ka#ne%an an! ,$ers'y 1@F:, p* D@3A* But of t#e lar&e literature on suc# "iases, t o stan! out for our purposes: self(ser$in& "ias an! "iase! ris' esti%ates* 4 recurrin& fin!in& in "e#a$ioral econo%ics is t#at people Gten! to "e "ot# unrealistically opti%istic an! o$erconfi!ent a"out t#eir =u!&%entsG ?Sunstein :;;;, p* FA* ,#ey suffer, in s#ort, fro% )self(ser$in& "ias*) -ore t#an #alf of sur$ey respon!ents rate t#e%sel$es in t#e top 6; percent of #ealt#, et#ics, !ri$in& a"ility, life e+pectancy, an! la"or pro!ucti$ity ?,aylor, 1@F@E .einstein, 1@F@A < a %at#e%atical i%possi"ility* .#at is particularly interestin&, t#ou&#, is t#at %ista'es tilt in t#e !irection of positi$e self(i%a&e* ,#is &oes !irectly a&ainst t#e stan!ar! rational e+pectations assu%ption*
Bertran!, -ullainat#an, an! S#afir ?:;;DA is t#e only e+ception of #ic# e are a are* 4'erlofGs ?:;;:, p*D1:A !iscussion of t#e Gstu""orn persistence of a self(!estructi$e un!erclassG is "roa!ly in t#e spirit of t#e current paper, "ut #e focuses on a no$el e+planation ?)t#e econo%ics of i!entity)A instea! of turnin& to t#e pre(e+istin& "e#a$ioral literature* 0laeser ?:;;6A, "uil!in& on 8rey an! Eic#en"er&er ?1@@D, 1@F@a, 1@F@"A uses "e#a$ioral political econo%y to Cuestion t#e $alue of paternalist policies, "ut !oes not specifically !iscuss elfare policy*
3
.e !o not inten! t#is to "e a co%pre#ensi$e e+a%ination of t#e lin's "et een "e#a$ioral econo%ics an! elfare policy* Bur &oal is to #i&#li&#t t#e connections e e+pect to "e %ost fruitful*
@ Si%ilarly, in!i$i!uals %a'e "iase! ris' esti%ates* 9iscusi ?1@@6, 1@@:, 1@FJA fin!s t#at in!i$i!uals put too %uc# e%p#asis on s%all ris's an! not enou&# on lar&e ones* In!i$i!uals are o$erly concerne! it# #i&#ly pu"lici>e! e$ents ?suc# as poisons in t#eir foo! an! ater or ant#ra+ conta%inationA "ut ta'e %uc# &reater #a>ar!s for &rante!*D ,#aler an! Jo#nson ?1@@;A si%ilarly fin! t#at racetrac' "ettors suffer fro% )lon&s#ot "ias): "ettors o$eresti%ate t#e c#ance t#at t#e lon&s#ot #orse ill in* .#at !o t#ese "e#a$ioral fin!in&s #a$e to !o it# t#e poorM ,a'e t#e case of sin&le %ot#ers* Bn t#e roa! to sin&le %ot#er#oo!, t#ere are %any points #ere =u!&%ental "iases plausi"ly play a role* 4t t#e outset, o%en %ay un!eresti%ate t#eir pro"a"ility of pre&nancy fro% unprotecte! se+* 4fter "eco%in& pre&nant, t#ey %i&#t un!eresti%ate t#e !ifficulty of raisin& a c#il! on oneGs o n, or o$eresti%ate t#e ease of =u&&lin& fa%ily an! career* 5olicies t#at %a'e it easier to "eco%e a sin&le %ot#er %ay per$ersely lea! %ore o%en to %a'e a c#oice t#ey are &oin& to re&ret* 4 si%ple nu%erical e+a%ple can illustrate t#e lin' "et een #elpin& t#e poor an! #ar%in& t#e%* Suppose t#at in t#e a"sence of &o$ern%ent assistance, t#e true net "enefit of #a$in& a c#il! out(of( e!loc' is (O:6,;;;, "ut a teena&e &irl it# self(ser$in& "ias "elie$es it is only (O6;;;* Since s#e still sees t#e net "enefits as ne&ati$e s#e c#ooses to ait* But suppose t#e &o$ern%ent offers O1;,;;; in assistance to un e! %ot#ers* ,#en t#e perceived "enefits rise to O6;;;, t#e teena&e &irl opts to #a$e t#e "a"y, an! e+ post e+periences a net "enefit of (O:6,;;; P O1;,;;; Q (O16,;;;*6 Br ta'e anot#er scenario: affir%ati$e action in #i&#er e!ucation* Coul! &i$in& %inority stu!ents %ore c#oices %a'e t#e% orse offM It coul! if t#ey are unrealistically opti%istic a"out t#eir pro"a"ility of success, lea!in& t#e% to c#oose an opportunity "eyon! t#eir capa"ilities* Self(ser$in& "ias %i&#t also incline eac# stu!ent to assu%e t#at
D
In t#e !iscrete case, of course, you can e$entually %a'e t#e recipient "etter off it# a lar&e enou&# &ift ?in t#is e+a%ple, a &ift &reater t#an O:6,;;;A* .it# continuous c#oices, t#ou&#, e$en an unli%ite! "u!&et %ay not suffice* 8or e+a%ple, if a #eroin a!!ict spen!s 1;;R of #is inco%e on #eroin, an! eac# !ose of #eroin on net %a'es #i% orse off, t#en #is utility is strictly decreasin& in inco%e*
1; #e as a!%itte! on #is o n %erits: )If I ere "ein& a!%itte! "ecause of affir%ati$e action, I s#oul! "e orrie!* But unli'e %any ot#er stu!ents, I as accepte! on %y %erits*)
. !elf-"ontrol #roblems
Et#no&rap#ic riters often e%p#asi>e t#at t#e poor, to put it "luntly, are la>y an! s#ort( si&#te! ?Haley an! S 1@@@ H1@6DI, p* :J6, 31FE .ilson 1@@6, p* 11FA* But t#ere is no nee! to turn to "e#a$ioral econo%ics to un!erstan! t#is* 2eoclassical t#eory certainly allo s for a&ents to #a$e #i&# !isutility of or' ?i*e*, to "e )la>y)A an! #i&# !iscount rates ?i*e*, to "e )s#ort(si&#te!)A* ,#e lo&ic of neoclassical elfare econo%ics still applies to t#ose at t#e e+tre%e tails of t#e preference !istri"ution: e+pan!in& t#e opportunity set of t#e la>y an! s#ort(si&#te! %a'es t#e% su"=ecti$ely "etter off* If a person #as a &enuine #atre! for or', a cas# &ift %ay ena"le #i% to lea$e t#e la"or force as #e al ays !rea%e! of !oin&* If a person is e+tre%ely i%patient, si%ilarly, #e is "etter off o$erall if a &o$ern%ent pro&ra% su"si!i>es front(loa!e! consu%ption* 1n!er t#ese circu%stances, li%itin& t#e safety net can "e a "enefit for ta+payers, "ut not recipients* Be#a$ioral econo%ists #a$e stu!ie! a superficially si%ilar set of c#oice ano%alies !istinct fro% la>iness an! s#ort(si&#te!ness: self(control pro"le%s* 5eople often spea' a"out failin& to lose ei&#t, stru&&lin& to Cuit s%o'in&, or procrastinatin& on a ter% paper* ,#ere appears to "e a conflict it# tra!itional c#oice t#eory: 4n a&ent s#oul! eit#er s%o'e ?if t#e "enefits e+cee! t#e costsA or not s%o'e ?if t#e costs e+cee! t#e "enefitsA* .#y oul! t#e !ecision "e torture! an! $acillatin&M ?BLDono&#ue an! /a"in, :;;:A 8or so%e, apparently, it is a "attle < an! often a losin& one < to %a+i%i>e e+pecte! lifeti%e utility* Ho precisely !o self(control pro"le%s !iffer fro% !isutility of or' or #i&# !iscount ratesM 4 person lac'in& in self(control predictably regrets #is c#oices* 4fter #e %a'es t#e%, #e is#es #e #a! !one so%et#in& else* In!ee!, e$en as #e %a'es one c#oice, in so%e sense #e %i&#t prefer to !o so%et#in& else*6 Lai"son et al* ?p* @6A point to t#e
6
,#us, pro"le%s of self(control are !istinct fro% =u!&%ental "iases* 4 person can #a$e rational e+pectations a"out t#e actual conseCuences of #is actions, "ut still internally stru&&le to ta'e t#e action #e !ee%s opti%al*
11 conflict "et een Gattitu!es, intentions, an! "e#a$ior*G .#en a neoclassical actor it# a #i&# !iscount rate i&nores lon&(run conseCuences, it is not a !ifficult !ecision* 1nli'e an a&ent it# a self(control pro"le%, #e !oes not re&ret #is c#oice or repeate!ly ple!&e to c#an&e #is life* 3ou %i&#t say t#at a self(control pro"le% is a'in to #a$in& intransiti$e preferences < or %ore accurately, irreflexive preferences* In so%e sense, t#e procrastinatin& stu!ent prefers partyin& to stu!yin&, "ut t#e opposite preference is also present* Br to ta'e a topic of far &reater policy si&nificance, "e#a$ioral e$i!ence on self( control #as also often "een use! to critici>e t#e opti%ality of retire%ent plannin&* 4s Lai"son et al* ?1@@F, p* @:A put it:
HCIonsu%ers face t o c#allen&es: %a'in& &oo! !ecisions an! stic'in& to t#e%* Econo%ists #a$e a!opte! opti%istic assu%ptions on "ot# counts* ,#e consu%ers in %ainstrea% econo%ic %o!els are assu%e! "ot# to "e e+ceptionally &oo! !ecision%a'ers an! to "e a"le to carry out t#eir plans* ,#ese econo%ic assu%ptions are !u"ious, particularly in re&ar! to sa$in& for retire%ent* ?1@@F, p* @:A
.#en an a&ent #as self(control pro"le%s, tra!itional elfare analysis is pro"le%atic* Is a s%o'er #o ants to Cuit "etter off if #e t#ro s all of #is to"acco a ayM .#at if a fello s%o'er te%pts #i% "y offerin& #i% a ci&aretteM 4t any &i$en %o%ent, #e prefers to li&#t up an! in#ale* But at t#e "e&innin& of eac# !ay, #e prefers to "e a non(s%o'er* 0i$en self(control pro"le%s, policies t#at #elp an a&ent reali>e #is %o%entary &oals can !erail #is effort to ac#ie$e #is %eta(&oals* It is possi"le, t#en, t#at e+pan!in& an a&entGs opportunity set %a'es #i% orse off*J ,#is is particularly clear if e affir% t#at t#e %eta(&oals are t#e )true) ar&u%ents in a&entsG utility functions* Be#a$ioral econo%ists often i%plicitly ta'e t#is stance* E$en if e re%ain a&nostic a"out t#e priority of %eta( &oals o$er %o%entary ones, t#ou&#, t#e e+istence of self(control pro"le%s un!er%ines t#e certainty of neoclassical elfare econo%ics* 0i$in& your o"ese neoclassical frien! a "i& "o+ of c#ocolates cannot %a'e #i% orse offE "ut it is at least possi"le your o"ese "e#a$ioral frien! oul! "e "etter off it#out your !elicious present*
See for e+a%ple Cutler, 0laeser, an! S#apiro ?:;;3A, #ic# consi!ers t#e possi"ility t#at !ue to self( control pro"le%s, lo er costs of foo! preparation #a$e on net #ar%e! 4%ericans "y increasin& o"esity*
1: .#at !o self(control pro"le%s #a$e to !o it# t#e pro"le%s of t#e poorM /eturnin& to an earlier e+a%ple, t#e c#oice to #a$e unprotecte! se+ is a classic case #ere a&ents e+perience inner conflict* ,#e i%%e!iate "enefits are so enticin&, "ut #at a"out t#e enor%ous lon&(run costsM 4 neoclassical a&ent it# #i&# ti%e preference ill #appily c#oose t#e i%%e!iate "enefit* But real( orl! a&ents %i&#t not only pre!icta"ly re&ret #a$in& unprotecte! se+, "ut re&ret it e$en at t#e ti%e* Self(control pro"le%s coul! also easily a%plify t#e per$erse effects of affir%ati$e action* 4fter !oin& poorly in an o$erly co%petiti$e uni$ersity, a %inority stu!ent #as to !eci!e #et#er to re!ou"le #is efforts or !rop out* 8ollo in& a strin& of set"ac's, !roppin& out #as i%%e!iate "enefits < escapin& aca!e%ic stress an! personal !e=ection < "ut t#e costs in ter%s of fore&one earnin&s ta'e years to %ateriali>e* Stu!ents it# self(control pro"le%s ill ten! to ta'e t#e easy ay out < an! re&ret it* ,#e literature on self(control also su&&ests t#at )la>iness) %ay not "e t#e "est ay to e+plain t#e !ifficulty t#e poor #a$e #ol!in& !o n a =o"* Instea!, t#ey %ay si%ply lac' self(control* If oneGs current =o" is so%e#o unpleasant, Cuittin& is an easy ay out* But t#ere are lon&(ter% costs of runnin& a ay fro% your pro"le%s: Eac# Cuit %a'es you less attracti$e to future e%ployers* 4 neoclassical a&ent oul! opti%ally "alance t#ese concerns* 4n a&ent it# self(control issues %i&#t repeate!ly &i$e in to t#e te%ptation to Cuit until #e "eco%es al%ost une%ploya"le*F
I9* ,HE 5BB/ DE9I4,E -B/E
Be#a$ioral econo%ics tells us t#at in!i$i!uals #a$e =u!&%ental "iases an! suffer fro% self(control pro"le%s* 0i$in& a person %ore c#oices t#erefore #as t#e potential, contrary to "asic %icroecono%ics, of pre!icta"ly %a'in& t#at person orse off* 8ro% t#is &eneral principle, e can !e!uce t#at #en &o$ern%ent assistance e+pan!s t#e c#oice sets of t#e !isa!$anta&e!, it can %a'e recipients orse off* In ot#er or!s, "e#a$ioral econo%ics
F
.ilsonGs ?1@@6, p* 1:;A inter$ie s it# inner(city e%ployers a"out e%ployee turno$er nicely capture t#is !an&er: GTt#eyLll, on t#e application itself, =ust say so%et#in& li'e )!i!nLt &et alon& it# super$isor) an! t#en t#e ne+t =o", reason for lea$in&, )!i!nLt &et alon& it# super$isor,) ne+t =o" reason for lea$in&, )!i!nLt li'e it,) an! t#eyLll #a$e &one t#rou&# t#ree or four =o"s in a %atter of si+ or ei&#t %ont#s an! t#en t#ey !onLt un!erstan! #y t#ey !onLt &et #ire! #ere*G
13 pro$i!es a )possi"ility t#eore%) for elfare policy: E+pan!in& opportunity sets may "ac'fire* 4t %ini%u%, t#en, t#is counsels a&ainst su%%arily !is%issin& t#eories a"out t#e elfare stateGs per$erse effects* But e can &o furt#er* E+istin& literature pro$i!es &oo! reasons to t#in' t#at t#e !e$iations of t#e poor fro% t#e stan!ar! neoclassical %o!el are especially pronounce!* ,#eir =u!&%ental "iases are %ore e+tre%e, an! t#eir self(control pro"le%s %ore se$ere, t#an t#ose of t#e rest of t#e population* Stan!ar! neoclassical t#eory is unusually li'ely to %islea! us #en e analy>e policies inten!e! to assist t#e !isa!$anta&e!*
1. $uantitative %vidence
Be#a$ioral econo%ists freCuently #i&#li&#t certain 'in!s of "e#a$ior < li'e o$er(eatin&, !rin'in&, s%o'in&, !ru& use, an! retire%ent plannin& < to %a'e t#eir case* It is not #ar! to len&t#en t#e list to inclu!e cri%e, unprotecte! se+, an! ot#er tra!itional )social pat#olo&ies*) .#at is rarely e%p#asi>e! in t#e "e#a$ioral literature, t#ou&#, is t#at !e$iant "e#a$ior is %uc# %ore pronounce! a%on& t#e poor* ,a'e !rin'in&* ,#e Su"stance 4"use an! -ental Healt# Ser$ices 4!%inistration ?1@@JA fin!s t#at G@*D percent of or'ers in t#e O@,;;;(O1@,@@@ inco%e &roup reporte! #ea$y alco#ol use, #ile only :*1 percent of or'ers in t#e #i&#est inco%e "rac'et ?OJ6,;;; an! %oreA reporte! #ea$y alco#ol use*G Br consi!er o"esity* Cutler, 0laeser, an! S#apiro ?:;;3, p*@JA report t#at o"esity &enerally !eclines %ar'e!ly as e!ucation rises* Durin& 1@J1(J6 perio!, t#is as true for "ot# &en!ers: :DR of o%en an! 16R of %en it#out #i&# sc#ool !iplo%as ere o"ese, $ersus JR of o%en an! FR of %en it# colle&e !e&rees* In t#e 1@FF(@D perio!, 3FR of o%en it# less t#an #i&# sc#ool e!ucations ere o"ese, $ersus :;R it# colle&e !e&rees* ?-ale o"esity rates #o e$er no lon&er s#o a su"stantial e!ucational patternA* 0i$en t#e stron& correlation "et een e!ucation an! inco%e, t#ere is little !ou"t t#at t#e poor #a$e %ore trou"le %aintainin& a #ealt#y "o!y ei&#t*
1D
5arallel results #ol!s #en e e+a%ine s%o'in& an! !ru& use* 4ccor!in& to t#e 2ational House#ol! Sur$ey on Dru& 4"use ?:;;:A:
T36 percent of persons fa%ilies it# total co%"ine! fa%ily inco%es of less t#an O@,;;; reporte! it# :@ percent of t#ose fro%
s%o'in& ci&arettes !urin& t#e past %ont# co%pare! fa%ilies it# inco%es of OJ6,;;; or %ore*
it# inco%es "et een O:;,;;; an! O3@,@@@ an! 1@ percent of t#ose fro%
Si%ilarly:
an esti%ate! 16*D percent of une%ploye! a!ults ?a&e! 1F an! ol!erA of part(ti%e e%ploye! a!ults usin& illicit !ru&s* ere current illicit
!ru& users in :;;;, co%pare! to 6*3 percent of full(ti%e e%ploye! a!ults an! J*F percent
/ecent sociolo&ical or' confir%s t#at t#e poor also !e$iate %ore in t#eir se+ual "e#a$ior* -anlo$e et al* ?:;;;A report t#at ris' of "irt# t#rou&#out a!olescence is !ecreasin& in inco%e* Si%ilarly, controllin& for fa%ily "ac'&roun!, Sucoff an! 1pc#urc# ?1@@FA foun! t#at "lac' teena&ers li$in& in )un!erclass) an! ) or'in& class) nei&#"or#oo!s ere %ore li'ely to #a$e se+ at earlier a&es an! %ore li'ely to &i$e "irt# as teena&ers* ,#e 2ational Sur$ey of 8a%ily 0ro t# ?1@@6A reports t#at out(of( e!loc' "irt#s occur far %ore freCuently a%on& t#e poor < only one in se$en un%arrie! %ot#ers #a$e inco%es a"o$e O:6,;;;E four out of e$ery ten un%arrie! %ot#ers #a$e inco%es "elo O1;,;;;* ,o ta'e a final social pat#olo&y, t#e poor are also prone to cri%inal "e#a$ior U a )career option) notorious for its fleetin& "enefits an! lon&(run costs* 4 Special /eport of t#e Bureau of Justice Statistics ?Harlo 1@@F, p*DA, to ta'e only one e+a%ple, e+plains t#at GJail in%ates reporte! lo inco%es* 4l%ost #alf reporte! inco%es of less t#an O6;; a %ont# < at %ost OJ,:;; annually < in t#e %ont# "efore t#eir %ost recent arrest*G It is te%ptin& to assu%e t#at cri%inals turn to cri%e "ecause of t#eir lo earnin&s, "ut only 16R of t#e in%ates reporte! recei$in& ille&al income* So%e of t#e %ost co%%on offenses < inclu!in& assault, !ru& possession, an! !run' !ri$in& < are rarely lucrati$e* In!ee!, as Le$itt an! Du"ner ?:;;6, pp*J@(1;DA !ocu%ent, e$en cri%e t#at pays rarely pays ellE !espite occupational ris's, !ru& !ealers often earn a"out t#e %ini%u% a&e*
16 B"=ecti$ely spea'in&, cri%e is an ineffecti$e ay to escape po$erty* But it is easy to see t#e appeal of cri%e to t#ose #o o$eresti%ate t#eir c#ance of "eco%in& a &an& lea!er, or #o si%ply #a$e poor i%pulse control* Bne coul! try to %ini%i>e t#is e$i!ence: 3es, t#e poor #a$e a #i&#er a$era&e inci!ence of pat#olo&ical "e#a$ior, "ut t#e !e$iation is only in t#e tails, not t#e %e!ian* ,#e a"solute le$el typically re%ains far "elo 6;R* In t#e "e#a$ioral literature, #o e$er, serious pro"le%s are nor%ally seen as extreme manifestations of continuous tendencies* 8ro% t#is stan!point, if you see a su"(population it# #i&#er le$el of e+tre%e "e#a$ior at t#e tails, you s#oul! also infer a #i&#er %e!ian* E+tre%e "e#a$ior is only t#e tip of t#e ice"er&E t#e %ore you see a"o$e t#e ater line, t#e %ore you can infer lies "eneat# it*@
. &ccounting for the 'eviation
.#y oul! t#e poor "e %ore prone to $iolate neoclassical assu%ptionsM 5er#aps t#e si%plest %ec#anis% is &eneral intelli&ence* ,#e a$era&e IN of t#e poor is ell "elo nor%al, #ic# accounts for %uc# of t#eir earnin&s &ap* ?Herrnstein an! -urray 1@@DE Jensen 1@@FA -easure! intelli&ence also correlates ne&ati$ely it# ille&iti%acy, c#il! a"use, s%o'in&, an! acci!ents* ?Herrnstein an! -urray 1@@D, pp* 161(3, 1J@(F1, :;J(1;, :1DA 4ll of t#ese patterns can "e plausi"ly interprete! as a%plifie! !epartures fro% neoclassical assu%ptions* ,a'e ille&iti%acy* Herrnstein an! -urray su&&est t#e follo in& causal %o!el:
,#e s%arter t#e o%an, t#e %ore li'ely t#at s#e !eli"erately !eci!es to #a$e a c#il! an! o%an is, t#e %ore li'ely t#at
s#e !oes not t#in' a#ea! fro% se+ to procreation, !oes not re%e%"er "irt# control, !oes not carefully consi!er #en an! un!er #at circu%stances s#e s#oul! #a$e a c#il!* Ho intelli&ent a o%an is %ay interact it# #er i%pulsi$eness, an! #ence #er a"ility to e+ert self(!iscipline an! restraint on #er partner in or!er to a$oi! pre&nancy* ?1@@D, p* 1J@A
In ot#er or!s, lo er intelli&ence a%plifies =u!&%ental "iases an! lac' of self(control* Less intelli&ent o%en are %ore li'ely to syste%atically un!eresti%ate t#e pro"a"ility of
@
16 pre&nancy fro% unprotecte! se+* ,#is lea!s t#e% to ta'e ris's t#ey oul! a$oi! if t#ey un!erstoo! t#e true nu%"ers* Herrnstein an! -urray also appear to say t#at it is %ore !ifficult for t#e% to control t#eir o n "e#a$iorE t#e less intelli&ent #a$e a stron&er ten!ency to %a'e c#oices t#at t#ey ill co%e to re&ret* Cri%e is anot#er interestin& case* Herrnstein an! -urray ac'no le!&e t#e stan!ar! #u%an capital account of t#e lin' "et een lo intelli&ence an! cri%e: GIf, for e+a%ple, people of lo intelli&ence #a$e a #ar! ti%e fin!in& a =o", t#ey %i&#t #a$e %ore reason to co%%it cri%es as a ay of %a'in& a li$in&*G ?1@@D, p* :D;A But t#ey also offer an alternate account "ase! on =u!&%ental "ias:
4 lac' of foresi&#t, #ic# is often associate! it# lo IN, raises t#e attractions of t#e #ic# co%e later
i%%e!iate &ains fro% cri%e an! lo ers t#e stren&t# of t#e !eterrents, ?if t#ey co%e at allA* ,o a person of lo uncertain* ?1@@D, p* :D;A
prison %ay fa!e to %eanin&lessness* ,#ey are too a"stract, too far in t#e future, too
.arner an! 5leeter ?:;;1A pro$i!es !irect e$i!ence t#at lo er IN a%plifies "e#a$ioral ano%alies*1; Durin& t#e 1@@;Gs, t#e 1*S* %ilitary offere! "enefit pac'a&es to in!uce $oluntary separation* ,#e "enefit pac'a&e coul! "e ta'en as an annuity or a lu%p(su% pay%ent* In line it# %uc# of t#e "e#a$ioral literature, .arner an! 5leeter fin! t#at actors use !iscount rates far in e+cess of t#e %ar'et interest rate < usually ta'en as a si&n of =u!&%ental "ias, lac' of self(control, or "ot#*11 ,#e present $alue of t#e lu%p su% as far less t#an t#e annuity < it# a "rea'(e$en !iscount rate &reater t#an 1JR < "ut %ost separatees c#ose t#e lu%p su%* .#at is note ort#y fro% t#e stan!point of t#e current paper is t#at in!i$i!uals it# lo er %easure! intelli&ence ere si&nificantly %ore li'ely to select t#e lu%p su%* ,#is re%ains true after controllin& for e!ucation, e+perience, fa%ily si>e, a&e, fiel! of speciali>ation, an! %ore* ,#us, separateesG ano%alous "e#a$ior "eco%es $isi"ly %ore pronounce! as t#eir intelli&ence falls*
1;
See also t#e %ore recent contri"utions "y 8re!eric' ?:;;6A, an! Ben=a%in an! S#apiro ?:;;6A* 8urt#er%ore, %any stu!ies !ocu%ent t#at e!ucation < a stron& correlate of "ot# inco%e an! intelli&ence < re!uces t#e %a&nitu!e of =u!&%ental "ias in a $ariety of su"=ects, inclu!in& econo%ics ?Caplan :;;1A, politics ?Delli Carpini an! Keeter 1@@6A, an! to+icolo&y ?Kraus, -al%fors, an! Slo$ic 1@@:A*
11
,#ere is of course a si%ple neoclassical e+planation: cre!it %ar'et i%perfections* ,#e pro"le% it# t#is account is t#at t#e ano%alies persist e$en if actors are !efinitely not liCui!ity constraine! ?e*&*, t#ey #a$e si&nificant #o%e eCuity or ot#er liCui! assetsA ?,#aler 1@@:, pp* 11@(:;A*
1J 4n intri&uin& stu!y "y Kru&er an! Dunnin& ?1@@@A reac#es t#e %uc# %ore &eneral conclusion t#at self(ser$in& "ias is !ecreasin& in o"=ecti$e a"ility* In ot#er or!s, t#e least co%petent in!i$i!uals o$eresti%ate t#e%sel$es t#e %ost* 4s Kru&er an! Dunnin& put it:
To$eresti%ation occurs, in part, "ecause people #o are uns'ille! in t#ese Hsocial an!
intellectualI !o%ains suffer a !ual "ur!en: 2ot only !o t#ese people reac# erroneous conclusions an! %a'e unfortunate c#oices, "ut t#eir inco%petence ro"s t#e% of t#e %etaco&niti$e a"ility to reali>e it* ?p* 11:1A
Since t#e poor are "elo (a$era&e on %ost of t#e stan!ar! %easures of a"ility < inclu!in& e!ucation an! intelli&ence < e s#oul! e+pect t#eir self(ser$in& "ias to "e especially se$ere* Bt#er researc# "ears out t#is pre!iction* -cClen!on an! .i&fiel! ?1@@FA for e+a%ple fin!s t#at in t#e face of poor o"=ecti$e perfor%ance, un!erac#ie$in& "lac' c#il!ren re%ain con$ince! t#at t#ey are &oo! in "ot# %at# an! science* ,#e )culture of po$erty) literature offers a ?potentially co%ple%entaryA alternati$e to Herrnstein an! -urrayGs &eneral intelli&ence story* Banfiel! ?1@6FA re%ains an insi&#tful intro!uction* 0enerali>in& fro% a lar&e et#no&rap#ic literature, Banfiel! i!entifies contrastin& sets of upper, %i!!le, or'in&, an! lo er class $alues ?1@6F, pp* DF(6DA* He ar&ues t#at lo er class $alues, trans%itte! fro% &eneration to &eneration, are an i%portant cause of po$erty* 4s Banfiel! star'ly puts it:
H,I#e lo er(class in!i$i!ual li$es %o%ent to %o%ent* If #e #as any a areness of t#e future, it is of so%et#in& fi+e!, fate!, "eyon! #is control: t#in&s #appen to #i%, #e !oes not %a'e t#e% #appen* I%pulse &o$erns #is "e#a$ior, eit#er "ecause #e cannot #ate$er #e cannot consu%e !iscipline #i%self to sacrifice a present for a future satisfaction or "ecause #e #as no sense of t#e future* He is t#erefore ra!ically i%pro$i!ent: i%%e!iately #e consi!ers $alueless* His "o!ily nee!s ?especially for se+A an! #is taste for )action) ta'e prece!ence o$er e$eryt#in& else < an! certainly o$er any or' routine* ?1@6F, p* 63A
4 #i&# !iscount rate %ay ell "e part of t#e story, "ut re!ucin& it to t#at alone see%s force!* In!ee!, #en Banfiel! ela"orates, #is account is al%ost e+plicitly "e#a$ioral: Gt#e in!i$i!ualGs orientation to ar! t#e future ill "e re&ar!e! as a function of t o factors: ?1A a"ility to i%a&ine a future, an! ?:A a"ility to !iscipline oneself to sacrifice present for
1F future satisfaction*G ?1@6F, p* DJA ,#e for%er stron&ly su&&ests =u!&%ental "iasE t#e latter, lac' of self(control* -ayer ?1@@JA is a %ore recent contri"ution to t#e culture of po$erty literature* S#e
conclu!es t#at e$en !ou"lin& t#e inco%e of t#e poor oul! #a$e little effect on lon&(ter% po$erty* Instea!, t#e pri%ary !eter%inants of success are parental c#aracteristics an! attitu!es, suc# as #onesty, !ili&ence, an! relia"ility < t#e $ery attitu!es Banfiel! %aintains t#at t#e poor are lac'in&* 8inally, 0laeser ?:;;6A su&&ests a t#ir! e+planation for #y t#e poor !e$iate %ore: ,#e ric# #a$e %ore %oti$ation an! resources to ma(e t#e%sel$es rational:
But t#ere are %any reasons to t#in' t#at incenti$e effects real ill "e %uc# stron&er in t#e tools it# #ic#
to i%pro$e t#eir reasonin&s, an! t#eir only real %et#o! of respon!in& to incenti$es is to t#in' #ar!er* Butsi!e of t#e la", people #a$e access to a!$isers, "oo's, t#e Internet, an! %ore ti%e* ,#eir illin&ness to spen! ti%e an! %oney to use t#ese resources ill surely !epen! on t#e sta'es in$ol$e! in t#e !ecision* ?:;;6, p*1D;A
Since t#e ric# #a$e "ot# #i&#er sta'es an! %ore resources, 0laeserGs analysis pre!icts t#at t#eir c#oices ill "e %ore rational ?t#ou&# a!%itte!ly, t#eir #i&#er opportunity cost of ti%e cuts t#e ot#er ayA*
9* 5BLIC3 I-5LIC4,IB2S
-oral #a>ar! is t#e tra!itional neoclassical ar&u%ent for li%itin& &o$ern%ent assistance* 4s pro&ra%s "eco%e %ore &enerous, "e#a$ioral !istortions increase* 1lti%ately, t#e %ar&inal !ollar transferre! costs !onors %uc# %ore t#an a !ollar, so if you put any ei&#t on !onorsG elfare, t#ere co%es a point #en you say )enou&#*) ,#e "e#a$ioral perspecti$e a!$ises us to superi%pose furt#er costs onto t#is calculus: .e %ust consi!er t#e cost to t#e recipients as ell as t#e cost to t#e !onors* Bnce e ac'no le!&e t#at t#e !isa!$anta&e! not only !e$iate fro% neoclassical assu%ptions, "ut !e$iate to an unusually stron& !e&ree, t#ere are at least ar&ua"ly e+pansi$e i%plications for policy* Be#a$ioral issues !o not tru%p all ot#er consi!erations, "ut t#ey !o pus# us in certain !irections*
1@
-ost o"$iously, if &o$ern%ent assistance to t#e !isa!$anta&e! a%plifies t#e ill effects of t#eir =u!&%ental "iases an! self(control pro"le%s, it stren&t#ens t#e case for re!ucin& t#e si>e of elfare "enefits, li%itin& t#eir !uration, restrictin& eli&i"ility, an! e$en a"olition* ,#e %ore )&enerous) pro&ra%s are, t#e %ore li'ely t#ey are to #ar% t#eir ostensi"le "eneficiaries furt#er !o n t#e roa!* But !ue to self(control pro"le%s, %any people ill accept t#e "enefits any ay* ,#ey are =ust too te%ptin& to turn !o n* -oreo$er, recipients ill &enerally o$eresti%ate t#eir o n a"ility to pursue t#eir lon&(run interests #ile on t#e !ole: )Bt#er people %ay li$e in t#e present an! &ro increasin&ly une%ploya"le, "ut not me*) )Bt#er sin&le %o%s %ay "e una"le to care for t#eir c#il!ren an! pursue a career, "ut not me*) Be#a$ioral econo%ics also puts affir%ati$e action in a ne li&#t* If stu!ents o$eresti%ate t#eir o n aca!e%ic a"ilities, you coul! #elp t#e )"eneficiaries) of affir%ati$e action "y curtailin& or en!in& preferential a!%issions* E$en in t#e a"sence of affir%ati$e action, self(ser$in& "ias lea!s stu!ents to pursue o$erly ris'y aca!e%ic strate&ies: )Bt#er 'i!s %ay &et in o$er t#eir #ea!s, "ut not me*) E+pan!in& t#eir set of ris'y c#oices it# preferential a!%issions %a'es t#e pro"le% orse* .#ile t#is is not an ironcla! ar&u%ent a&ainst affir%ati$e action, it is an a!!itional cost of t#e policy t#at %ost analyses ne&lect* 2eoclassical econo%ics oul! #a$e to la"el our reasonin& )paternalistic*) 4s Sunstein ?:;;;, p* D6A o"ser$es, "e#a$ioral econo%ics is at least )anti(anti(paternalist*) It opens up t#e possi"ility of #urtin& people "y e+pan!in& t#eir c#oice set, an! #elpin& t#e% "y s#rin'in& t#eir c#oice set* ,#is t#eoretical result is plainly sy%%etric* If e+pan!in& la"or %ar'et c#oices a"o$e t#e laisse>(faire le$el %a'es t#e poor orse off, t#en re!ucin& t#e% belo) t#e laisse>(faire le$el %i&#t %a'e t#e poor "etter off still* Bne %i&#t interpret la s a&ainst $a&rancy or truancy in t#is li&#t < as an atte%pt to correct for a&entsG self(control pro"le%s #en
:; t#ey ei&# t#e "enefits of leisure to!ay a&ainst t#e future "enefits of #u%an capital acCuisition* I%a&ine &rap#in& socially opti%al policy as a function of t#e se$erity of a&entsG !e$iation fro% neoclassical assu%ptions* ?8i&ure :A 8or %il! !e$iations, t#e opti%al a!=ust%ent oul! presu%a"ly only "e to sli&#tly curtail t#e &enerosity of &o$ern%ent assistance* 4s !e$iations fro% neoclassical assu%ptions "eco%es &reater, t#ou&#, t#e lon&(run #ar% of a &i$en le$el of assistance rises, an! t#e opti%al le$el of &enerosity accor!in&ly falls* 8or se$ere "iases, si%ple refusal to #elp ill not suffice, an! opti%al policy oul! #a$e to restrict e$en t#e c#oice set re%ainin& to t#e poor un!er laisse>(faire* So t#e "e#a$ioral perspecti$e !efinitely ar&ues for different &o$ern%ent policies, "ut not necessarily for less &o$ern%ent* ,#ere is #o e$er a contin&ent factor t#at pus#es in fa$or of laisse>(faire: t#e fi+e! costs of &o$ern%ent pro&ra%s* 4s lon& as any for% of inter$ention < #et#er positi$e ?e*&* &i$in& t#e poor %oneyA or ne&ati$e ?e*&* for"i!!in& $a&rancyA < #as fi+e! costs, t#ere e+ists a !iscrete ran&e o$er #ic# laisse>(faire is opti%al* ?8i&ure 3A If, i&norin& fi+e! costs, t#e opti%al policy in$ol$es only %il! &o$ern%ent action, t#en ta'in& fi+e! costs into account, t#e opti%al policy is no &o$ern%ent action at all*1: ,#us, t#ere is a contin&ent connection "et een "e#a$ioral econo%ics an! conser$ati$e an! li"ertarian pleas for laisse>(faire* 4t t#e sa%e ti%e, t#e "e#a$ioral approac# pus#es a&ainst laisse>(faire alon& ot#er policy !i%ensions* Consi!er re&ulation of !ru&s, alco#ol, an! to"acco* 8ro% a stan!ar! neoclassical perspecti$e, consu%ption of t#ese su"stances is an una%"i&uous social "enefit* ,#e te+t"oo' case for restriction or pro#i"ition t#erefore #in&es on e+ternalities < e+ternalities #ic# %any econo%ists #a$e calle! into Cuestion* ,#e stan!ar! rationale for ta+es on ci&arettes, for instance, is t#at s%o'ers !o not pay t#e full social cost of t#eir #a"it "ecause &o$ern%ent pic's up a s#are of t#eir %e!ical e+penses ?0ross%an et al*, 1@@3E Hay, 1@@1E -annin& et al*, 1@F@,
1:
,o t#is e+tent, e s#are 0laeserGs ?:;;6A conclusion t#at Greco&ni>in& t#e li%its of #u%an co&nition %ay stren&t#en t#e case for li%ite! &o$ern%ent*G ?p*133A But our arguments are al%ost ort#o&onal* .e use "e#a$ioral econo%ics to un!er%ine neoclassical econo%icsG analysis of t#e elfare state* 0laeser uses "e#a$ioral political econo%y to un!er%ine "e#a$ioral econo%icsG analysis of paternalis%*
:1 1@@1A* ,#is ar&u%ent #as "een !e"un'e! "y econo%ists #o point out t#at s%o'ers also !ie youn&er, an! t#at t#e sa$in&s fro% earlier !eat# out ei&# t#e costs of s%o'in&( relate! %e!ical treat%ents ?9iscusi, 1@@DA* 8ro% a "e#a$ioral perspecti$e, one can construct a Cuite !ifferent case for ta+es on to"acco* .e %ust consi!er not =ust t#e costs s%o'ers i%pose on non(s%o'ers, "ut also t#e costs t#at s%o'ers i%pose on t#e%sel$es* S%o'ers often la%ent t#at lac' of self( control pre$ents t#e% fro% 'ic'in& t#eir #a"it ?Sur$ano$ic et al*, 1@@@E .erten"roc#, 1@@FE 4'erlof, 1@@1E .inston, 1@F;A* B$erconfi!ence ar&ua"ly e+acer"ates t#eir pli&#t: 3oun& s%o'ers %ay for% a plan to Cuit #en t#ey turn forty, t#in'in& )1nli'e ot#ers, I can stop anyti%e I ant*) 1sin& ta+es to raise t#e price of ci&arettes coul! t#erefore actually #elp s%o'ers* By increasin& t#e upfront cost of t#e #a"it, you coul! %a'e t#e% )act as if) t#ey accurately a!=uste! for its lon&(ter% conseCuences* -oreo$er, insofar as t#e poor suffer fro% %ore pronounce! "iases t#an ot#er se&%ents of t#e population, #at is usually seen as a !efect of ci&arette ta+es < t#eir re&ressi$e i%pact < is actually a "enefitK ,#e %ore se$ere peopleGs !e$iation fro% neoclassical assu%ptions, t#e lar&er t#e &ap "et een t#eir c#oices an! t#eir elfare* 4 constant per( pac' ta+ presu%a"ly %atters %ore to lo (inco%e s%o'ers* ,#us, ci&arette ta+es pro"a"ly #a$e t#e lar&est effect on t#e to"acco consu%ption of t#ose #o, left to t#eir o n !e$ices, ten! to !o t#e%sel$es t#e %ost #ar%* ,o repeat, "e#a$ioral consi!erations !o not tru%p con$entional neoclassical cost("enefit analysis* Bne coul! a!%it t#at consu%ers of #eroin oul! &enerally "e "etter off consu%in& no #eroin, "ut still re=ect t#e efficiency case for pro#i"ition "y pointin& to in!irect effects of street $iolence an! a!ulteration* .#at t#e "e#a$ioral perspecti$e teac#es us is t#at t#e stan!ar! neoclassical fra%e or' ne&lects so%e of t#e "enefits of pro#i"ition* .#et#er or not t#ese for&otten "enefits tip t#e scales !epen!s on #o close t#e contest as to "e&in it#*
:: It is ise to pursue paternalistic reasonin& cautiously* ?0laeser :;;6A ,#ere is a ris' of re!efinin& all "e#a$iors you !isappro$e of as )self(control pro"le%s,) an! all "eliefs you !isa&ree it# as )=u!&%ental "iases*) ,#e !an&er you pose to yourself is pro"a"ly tri$ial co%pare! to t#e !an&er of li$in& un!er t#e $eto of a ran!o%ly selecte! "e#a$ioral econo%ist* 4 ris'(a$erse policy(%a'er s#oul! t#in' lon& an! #ar! "efore !irectly or!erin& t#e poor a"out for t#eir o n &oo!* But curtailin& t#e &enerosity of e+istin& pro&ra%s is a lo (ris', strai&#tfor ar! ay to %o!ify policy to ta'e account of "e#a$ioral concerns* Inci!entally, our analysis is eCually rele$ant for pri$ate c#arity* Concerne! altruists #a$e to consi!er t#e possi"ility t#at less fortunate oul! "e "etter off if t#ey #elpe! t#e% less* ,#is su&&ests a ne spin on t#e lon&(stan!in& !istinction "et een t#e )!eser$in&) an! t#e )un!eser$in&) poor ?Hi%%elfar", 1@@1A* 3ou %i&#t concei$e of t#e !eser$in& poor as t#ose #o are clear(#ea!e! an! !iscipline! enou&# to "enefit fro% financial assistance* ,#e un!eser$in& poor, con$ersely, can "e seen as t#ose too far fro% t#e rational actor %o!el to %ec#anically "enefit fro% #i&#er inco%e* 5#ilant#ropists it# )#ar! #ea!s an! soft #earts) < to use 4lan Blin!erGs ?1@FJA p#rase < %ust face t#e #ars# reality t#at t#ey %ay "e una"le to #elp t#ose #o ill not #elp t#e%sel$es*
9I* CB2CL1SIB2
,o t#e "est of our 'no le!&e, t#is is t#e first paper to analy>e t#e connection "et een "e#a$ioral econo%ics an! #at ,yler Co en calls t#e Gtra!itional conser$ati$e critiCue of t#e elfare state*G -ost !etractors of t#e elfare state #a$e turne! to neoclassical econo%ics for intellectual support* 8e pro%oters of "e#a$ioral econo%ics #a$e stoppe! to consi!er its i%plications for po$erty policy* Bur paper ai%s to re$erse "ot# of t#ese tren!s* So%e of t#e %ost co%%on co%plaints a"out t#e elfare state are, fro% a strict neoclassical perspecti$e, senseless* But fro% t#e stan!point of "e#a$ioral econo%ics, t#ey are Cuite co#erent* -oreo$er, e$en t#ou&# "e#a$ioral econo%ists #a$e &i$en policy to ar!s t#e !isa!$anta&e! s#ort s#rift, t#is turns out to "e a topic #ere "e#a$ioral fin!in&s are especially rele$ant* 4 $ariety of sources
:3 in!icate t#at )t#e poor !e$iate %ore*) If t#e a$era&e person $iolates neoclassical assu%ptions, t#e a$era&e elfare recipient $iolates t#e% to a %ar'e!ly &reater !e&ree* So%e of t#e policy i%plications are strai&#tfor ar!: Be#a$ioral econo%ics pro$i!es a!!itional reasons for less &enerous &o$ern%ent assistance alon& a nu%"er of %ar&ins* If t#e people t#e &o$ern%ent ants to #elp !o not fully account for t#e ne&ati$e lon&(run effects of acceptin& #elp, t#ey are "etter off if t#e &o$ern%ent !oes t#e accountin& for t#e%* Bt#er policy i%plications are less o"$ious* Specifically, it is t#eoretically possi"le for &o$ern%ent to #elp t#e !isa!$anta&e! "y re!ucin& t#eir c#oice set "elo t#e laisse>( faire le$el* ,#e tra!itional conser$ati$e critiCue of t#e elfare state is fun!a%entally paternalist* Bnce you accept t#e i!ea t#at you can #urt people "y &i$in& t#e% %ore c#oices, you cannot !is%iss t#e i!ea t#at you can #elp t#e% "y ta'in& so%e of t#eir c#oices a ay* In practice, of course, t#e latter is %uc# %ore costly an! intrusi$e t#an t#e for%er* ?0laeser :;;6A ,#ere are i%portant i%plications for future researc#* ,#e "e#a$ioral literature #as !ocu%ente! t#at t#e a$era&e person freCuently $iolates neoclassical assu%ptions* But it rarely in$esti&ates variation in t#e ten!ency to $iolate neoclassical assu%ptions*13 Casual e%piricis% an! li%ite! for%al e$i!ence su&&est t#at t#e poor !o !e$iate %ore* 4 &reat !eal %ore coul! "e learne! at lo cost if ne "e#a$ioral stu!ies collecte! infor%ation on participantsG inco%e an! e!ucation to test for #etero&eneity* Bne of t#e %ain contri"utions of "e#a$ioral econo%ics #as "een to su"=ect t#e fun!a%ental assu%ptions of econo%ic t#eory to e%pirical scrutiny* ,e+t"oo' %icro is useful in a &reat %any conte+ts, "ut "e#a$ioral econo%ists #a$e foun! so%e "lin! spots* In t#is paper, e #a$e ar&ue! t#at co%%on sense !ou"ts a"out t#e efficacy of t#e elfare state are often iso%orp#ic to so%e of t#e %ain "lin! spots "e#a$ioral econo%ists #a$e !isco$ere!* Bn a #o%e or' pro"le% in inter%e!iate %icro, you can eli%inate po$erty it# lu%p(su% transfers* In t#e real orl!, %atters are not so si%ple*
13
8or nota"le e+ceptions, see 8re!eric' ?:;;6A, Ben=a%in an! S#apiro ?:;;6A, an! .arner an! 5leeter ?:;;1A*
:D REFERENCES 4'erlof, 0eor&e 4* ?:;;:A* Be#a$ioral -acroecono%ics an! -acroecono%ic Be#a$ior, &merican %conomic *evie). @:: D11:33* 4'erlof, 0eor&e 4* ?1@@1A* 5rocrastination an! B"e!ience, &merican %conomic *evie). F1?:A: 1(1@* Banfiel!, E! ar! C* ?1@6FA* +he ,nheavenly "ity* Boston: Little, Bro n an! Co* Ben=a%in, Daniel an! .illia% Dou&an ?:;;1A* In!i$i!ualsL Esti%ates of t#e /is's of Deat#: 5art II<2e E$i!ence, Journal of *is( and ,ncertainty. 33(6J* Ben=a%in, Daniel an! .illia% Dou&an ?1@@JA* In!i$i!ualsL Esti%ates of t#e /is's of Deat#: 5art I<4 /eassess%ent of 5re$ious E$i!ence, Journal of *is( and ,ncertainty. 116(33* Ben=a%in, Daniel an! Jesse S#apiro ?:;;6A* .#o is GBe#a$ioralGM Co&niti$e 4"ility an! 4no%alous 5references* unpu"* %s* #ttp:VV#o%e*uc#ica&o*e!uVW=%s#apirViC;J;D;6*p!f* Bertran!, -arianne, Sen!#il -ullainat#an an! El!ar S#afir ?:;;DA* 4 Be#a$ioral( Econo%ics 9ie of 5o$erty* &merican %conomic *evie). @D: D1@(D:3* Blin!er, 4lan ?1@FJA* -ard -eads. !oft -earts / +ough-0inded %conomics for a Just !ociety* /ea!in&, -4: 4!!ison(.esley* Ca%erer, Colin an! /ic#ar! H* ,#aler ?1@@6A* 4no%alies: 1lti%atu%s, Dictators, an! -anners, Journal of %conomic #erspectives. @: :;@(1@* Caplan, Bryan ?:;;1A* .#at -a'es 5eople ,#in' Li'e Econo%istsM E$i!ence on Econo%ic Co&nition fro% t#e Sur$ey of 4%ericans an! Econo%ists on t#e Econo%y, Journal of La) and %conomics. DD: 3@6(D:6* C#a$e>, Lin!a ?:;;:A* &n ,nli(ely "onservative* 2e 3or': Basic Boo's* C#a$e>, Lin!a ?1@@:A* 1ut of the Barrio/ +o)ard a 2e) #olitics of -ispanic &ssimilation* 2e 3or': Basic Boo's* Co en, ,yler ?:;;:A* Does t#e .elfare State Help t#e 5oorM !ocial #hilosophy and #olicy* 1@: 36(6D* Cutler, Da$i!, E! ar! 0laeser an! Jesse S#apiro ?:;;3A* .#y Ha$e 4%ericans Beco%e -ore B"eseM, Journal of %conomic #erspectives. 1J: @3(11F* Delli Carpini, -ic#ael S*, an! Scott Keeter ?1@@6A* 3hat &mericans 4no) &bout #olitics and 3hy It 0atters. 2e Ha$en, C,: 3ale 1ni$ersity 5ress* 8re!eric', S#ane :;;6* Bn t#e Ball: Co&niti$e /eflection an! Decision -a'in&* unpu"* %s* #ttp:VV *%it*e!uVpeopleVs#anefreVBnR:;t#eR:;Ball*p!f*
:6 8rey, Bruno an! Eic#en"er&er, /einer ?1@@DA* Econo%ic Incenti$es ,ransfor% 5syc#olo&ical 4no%alies, Journal of %conomic Behavior and 1rgani5ation* :3: :16(3D* 8rey, Bruno an! Eic#en"er&er, /einer ?1@F@aA* 4no%alies an! Institutions, Journal of Institutional and +heoretical %conomics. 1D6: D:3(D3J* 8rey, Bruno an! Eic#en"er&er, /einer ?1@F@"A* S#oul! Social Scientists Care a"out C#oice 4no%aliesM, *ationality and !ociety* 1: 1;1(1::* 8rie!%an, -ilton ?1@F: H1@6:IA* "apitalism and 6reedom* C#ica&o: 1ni$ersity of C#ica&o 5ress* 0laeser, E! ar! ?:;;6A* 5aternalis% an! 5syc#olo&y, ,niversity of "hicago La) *evie)* J3: 133(166* 0ross%an, -ic#ael, Jo!y L* Sin!elar, Jo#n -ulla#y an! /ic#ar! 4n!erson ?1@@3A* 5olicy .atc#: 4lco#ol an! Ci&arette ,a+es, Journal of %conomic #erspectives. J: :11( ::* Haley, 4le+ an! -alcol% S ?1@@@ H1@6DIA* +he &utobiography of 0alcolm 7* 2e 3or': /an!o% House* Harlo , Caroline .* 1@@F* Bureau of Justice !tatistics !pecial *eport/ #rofile of Inmates 1889. 1*S* Depart%ent of Justice* 1/L #ttp:VV *o=p*us!o=*&o$V"=sVpu"Vp!fVp=i@6*p!f* Hay, Joel .* ?1@@1A* ,#e Har% ,#ey Do to Bt#ers: 4 5ri%er on t#e E+ternal Costs of Dru& 4"use, in: -el$yn B* Krauss an! E! ar! 5* La>ear ?e!s*A, !earching for <ernatives/ 'rug "ontrol #olicy in the ,nited !tates* Stanfor!, C4: Hoo$er Institution 5ress* Herrnstein, /ic#ar! J* an! C#arles -urray ?1@@DA* +he Bell "urve/ Intelligence and "lass !tructure in &merican Life* 2e 3or': 8ree 5ress* Hi%%elfar", 0ertru!e ?1@@1A* #overty and "ompassion/ +he 0oral Imagination of the Late :ictorians* 2e 3or': Knopf* Ka#ne%an, Daniel, Jac' L* Knetsc# an! /ic#ar! H* ,#aler ?1@@;A* E+peri%ental ,ests of t#e En!o %ent Effect an! t#e Coase ,#eore%, Journal of #olitical %conomy. @F: 13:6( DF* Ka#ne%an, Daniel, 5aul Slo$ic an! 4%os ,$ers'y, e!s* ?1@F:A* Judgment ,nder ,ncertainty/ -euristics and Biases* 2e 3or': Ca%"ri!&e 1ni$ersity 5ress* Ka#ne%an, Daniel an! 4%os ,$ers'y ?1@F:A* Bn t#e Stu!y of Statistical Intuitions, in: Daniel Ka#ne%an, 5aul Slo$ic an! 4%os ,$ers'y ?e!s*A Judgment ,nder ,ncertainty/ -euristics and Biases* 2e 3or': Ca%"ri!&e 1ni$ersity 5ress*
:6 Ka#ne%an, Daniel an! 4%os ,$ers'y ?1@J@A* 5rospect ,#eory: 4n 4nalysis of Decision 1n!er /is', %conometrica. DJ: :63(@1* Kir"y, Kris an! Herrnstein, /ic#ar! J* ?1@@6A* 5reference /e$ersals Due to -yopic Discountin& of Delaye! /e ar!, #sychological !cience. 6: F3(@* Kraus, 2ancy, ,or"=Xrn -al%fors an! 5aul Slo$ic ?1@@:A* Intuiti$e ,o+icolo&y: E+pert an! Lay Ju!&%ents of C#e%ical /is's, *is( &nalysis. 1:: :16(3:* Kru&er, Justin an! Da$i! Dunnin& ?1@@@A* 1ns'ille! an! 1na are of It: Ho Difficulties in /eco&ni>in& BneLs B n Inco%petence Lea! to Inflate! Self(4ssess%ents, Journal of #ersonality and !ocial #sychology. JJ: 11:1(3D* Lai"son, Da$i! ?1@@FA* Life(Cycle Consu%ption an! Hyper"olic Discount 8unctions, %uropean %conomic *evie) #apers and #roceedings. D:: F61(J1* Lai"son, Da$i! ?1@@JA* 0ol!en E&&s an! Hyper"olic Discountin&, $uarterly Journal of %conomics. 11: ?:A: DD3(JF* Lai"son, Da$i!, 4n!rea /epetto an! Jere%y ,o"ac%an ?1@@FA* Self(Control an! Sa$in& for /etire%ent, Broo(ings #apers on %conomic &ctivity. 1: @1(1@6* Le$itt, Ste$en, an! Step#en Du"ner ?:;;6A* 6rea(onomics/ & *ogue %conomist %xplores the -idden !ide of %verything. 2e 3or': HarperCollins* -anlo$e, Jennifer, Eli>a"et# ,erry, Laura 0itelson, 4n&ela /o%ano 5apillo, an! Step#en ,* /ussell ?:;;;A* E+plainin& De%o&rap#ic ,ren!s in ,een 8ertility, 1@F;(1@@6: ,#e C#an&in& Conte+t of 4!olescence, 6amily #lanning #erspective. 3:: 166(J6* -annin&, .illar!, E%%ett Keeler, Josep# 5* 2e #ouse, Eli>a"et# Sloss, an! Jeffrey .asser%an ?1@@1A* +he "osts of #oor -ealth -abits* Ca%"ri!&e: Har$ar! 1ni$ersity 5ress* -annin&, .illar!, E%%ett Keeler, Josep# 5* 2e #ouse, Eli>a"et# Sloss, an! Jeffrey .asser%an ?1@F@A* ,#e ,a+es of Sin: Do S%o'ers an! Drin'ers 5ay ,#eir .ayM Journal of &merican 0edical &ssociation. :61: 16;D(16;@* -ayer, Susan E* ?1@@JA* 3hat 0oney "an;t Buy/ 6amily Income and "hildren;s Life "hances* Ca%"ri!&e, -4: Har$ar! 1ni$ersity 5ress* -cClen!on, Crystal an! 4llan .i&fiel! ?1@@FA* 0roup Differences in 4frican 4%erican 4!olescentsL 4c#ie$e%ent(/elate! Beliefs a"out -at# an! Science: 4n Initial Stu!y, Journal of Blac( #sychology. :D: :F(D3* -urray, C#arles 4* ?1@FDA* Losing Ground/ &merican !ocial #olicy. 18<=-18>=* 2e 3or': Basic Boo's*
:J 2ational House#ol! Sur$ey on Dru& 4"use ?:;;:A* #ttp:VV 4pplie! Stu!ies, .as#in&ton, DC* *sa%#sa*&o$V* Bffice of
2ational Sur$ey of 8a%ily 0ro t# ?1@@6A* #ttp:VVnic#!*ni#*&o$* 2ational Institute of C#il! Healt# an! Hu%an De$elop%ent, Bet#es!a, -D* BLDono&#ue, ,e! an! -att#e /a"in ?:;;3A* Self 4 areness an! Self Control, in: /oy 8* Bau%eister, 0eor&e Lo enstein, an! Daniel /ea! ?e!s*A +ime and 'ecision/ %conomic and #sychological #erspectives on Intertemporal "hoice* /ussell Sa&e 8oun!ation: :1J(:DD* BLDono&#ue, ,e! an! -att#e /a"in ?1@@@aA* Doin& It 2o or Later, &merican %conomic *evie). F@: 1;3(:D* BLDono&#ue, ,e! an! -att#e /a"in ?1@@@"A* Incenti$es for 5rocrastinators, $uarterly Journal of %conomics. 11D: J6@(F16* 5esaran, Has#e% ?1@FJA* +he Limits to *ational %xpectations* B+for!: Blac' ell* /a"in, -att#e ?1@@FA* 5syc#olo&y an! Econo%ics, Journal of %conomic Literature. 36: 11(D6* /ot#"ar!, -urray 2* ?1@JFA* 6or a 2e) Liberty/ +he Libertarian 0anifesto. 2e 3or': Li"ertarian /e$ie 8oun!ation* S#effrin, Ste$en -* ?1@@6A* *ational %xpectations* Ca%"ri!&e: Ca%"ri!&e 1ni$ersity 5ress* So ell, ,#o%as ?1@@6 H1@F;IA* 4no)ledge and 'ecisions* 2e 3or': Basic Boo's* So ell, ,#o%as ?1@@3A* Inside &merican %ducation/ +he 'ecline. the 'eception. the 'ogmas* 2e 3or': 8ree 5ress* So ell, ,#o%as ?1@@;A* #referential #olicies/ &n International #erspective* 2e 3or': .* -orro * So ell, ,#o%as ?1@FJA* "ompassion :ersus Guilt. and 1ther %ssays* 2e 3or': .* -orro * ,#e Su"stance 4"use an! -ental Healt# Ser$ice 4!%inistration* ?1@@JA* #ttp:VV *sa%#sa*&o$ Bffice of 4pplie! Stu!ies, .as#in&ton, DC* Sucoff, Clea 4*, an! Da n -* 1pc#urc# ?1@@FA* 2ei&#"or#oo! Conte+t an! t#e /is' of C#il! Bearin& 4%on& -etropolitan 4rea Blac' 4!olescents, &merican !ociological *evie). 63: 6J1(F6* Sunstein, Cass /* ?1@@6A* Social 2or%s an! Social /oles, "olumbia La) *evie). @6: @;3(6F*
:F Sunstein, Cass /*, e!* ?:;;;aA* Behavioral La) and %conomics* 2e 3or': Ca%"ri!&e 1ni$ersity 5ress Sur$ano$ic, Ste$en -*, /o"ert S* 0ol!far" an! ,#o%as C* Leonar! ?1@@@A* 4n Econo%ic ,#eory of Ci&arette 4!!iction, Journal of -ealth %conomics. 1F: 1(:@* ,#aler, /ic#ar! H* ?1@@:A* +he 3inner;s "urse/ #aradoxes and &nomalies of %conomic Life* 5rinceton: 5rinceton 1ni$ersity 5ress* ,#aler, /ic#ar! H* an! Eric J* Jo#nson ?1@@;A* 0a%"lin& it# t#e House -oney an! ,ryin& to Brea' E$en: Effects of 5rior Butco%es on /is'y C#oice, 0anagement !cience. 36: 6D3(6;* ,#aler, /ic#ar! H* an! Hers# -* S#efrin ?1@F1A* 4n Econo%ic ,#eory of Self(Control, Journal of #olitical %conomy. F@: 3@:(D;6* 9iscusi, .* Kip ?1@@6A* *ational *is( #olicy* 2e 3or': 2e 3or' 1ni$ersity 5ress* 9iscusi, .* Kip ?1@@DA* Ci&arette ,a+ation an! t#e Social ConseCuences of S%o'in&, 2ational Bureau of %conomic *esearch 3or(ing #aper. 2o* DF@1* 9iscusi, .* Kip ?1@@:A* 6atal +radeoffs/ #ublic and #rivate *esponsibilities for *is(* 2e 3or': B+for! 1ni$ersity 5ress* 9iscusi, .* Kip ?1@FJA* Learning about *is(/ "onsumer and 3or(er *esponses to -a5ard Information* Ca%"ri!&e, -4: Har$ar! 1ni$ersity 5ress* .arner, Jo#n ,* an! Saul 5leeter ?:;;1A* ,#e 5ersonal Discount /ate: E$i!ence fro% -ilitary Do nsi>in& 5ro&ra%s, &merican %conomic *evie). @1: 33(63* .einstein, 2eil D* ?1@F@A* Bpti%istic Biases a"out 5ersonal /is's, !cience. 1:3:* .erten"roc#, Klaus ?1@@FA* Consu%ption Self(Control "y /ationin& 5urc#ase Nuantities of 9irtue an! 9ice, 0ar(eting !cience. 1J: 31J(3J* .ilson, .illia% J* ?1@@6A* 3hen 3or( 'isappears* 2e 3or': /an!o% House* .inston, 0or!on C* ?1@F;A* 4!!iction an! Bac'sli!in&: 4 ,#eory of Co%pulsi$e Consu%ption, Journal of %conomic Behavior and 1rgani5ation. 1: :@6(3:D*
:@
8uture Inco%e
In!ifference Cur$e 4
In!ifference Cur$e B
3;
-ore C#oice
Laisse>(faire
De$iation
Less C#oice
31
Figure %: !eviations from Neoclassical Assum"tions and #"timal Polic$ &ith Fi'ed Costs
-ore C#oice
Laisse>(faire
De$iation
Less C#oice