bowles: endogenous preferences endogenous preferences: the

Journal of Economic LiteratureVol. XXXVI (March 1998), pp. 75–111

Bowles: Endogenous Preferences

Endogenous Preferences: TheCultural Consequences of Markets and

other Economic Institutions

SAMUEL BOWLESUniversity of Massachusetts at Amherst

Thanks to Eric Verhoogen, James Heintz, Stephanie Eckman, Nicole Huber, and Melissa Os-borne for research assistance, to the University of Siena for providing an unparalleled envi-ronment for research and writing, to Robert Boyd, Colin Camerer, Robert Cialdini, LiliaCostabile, Joshua Cohen, Gerald Cohen, Martin Daly, Peter Dorman, Catherine Eckel, Mar-cus Feldman, Ernst Fehr, Nancy Folbre, Christina Fong, Bruno Frey, Herbert Gintis, KarlaHoff, Daniel Kahneman, Melvin Kohn, David Kreps, Timur Kuran, George Loewenstein, Eli-nor Ostrom, Bentley MacLeod, Paul Malherbe, Karl Ove Moene, Casey Mulligan, RichardNisbett, Ugo Pagano, Jean-Philippe Platteau, John Roemer, Susan Rose-Ackerman, PaulRomer, Paul Rozin, Andrew Schotter, Paul Seabright, Gil Skillman, Peter Skott, Rohini Soma-nathan, Hillel Steiner, Philippe van Parijs, Burton Weisbrod, Elisabeth Wood, Erik OlinWright, Viviana Zelizer, and two anonymous referees for guidance in the literatures coveredhere and comments on an earlier draft, and to the MacArthur Foundation for financial sup-port. Correspondence to [email protected].

1. Hobbes’ Fiction

[Let us] . . . return again to the state of na-ture, and consider men as if but even nowsprung out of the earth, and suddenly (likemushrooms), come to full maturity, withoutany kind of engagement with each other.

Thomas Hobbes ([1651]1949, p. 100)

If friends make gifts, gifts make friends. . . .Thus do primitive people transcend the Hob-besian chaos. Marshall Sahlins (1972, p. 186)

MARKETS AND OTHER economic in-stitutions do more than allocate

goods and services: they also influencethe evolution of values, tastes, and per-sonalities. Economists have long as-sumed otherwise; the axiom of exoge-nous preferences is as old as liberalpolitical philosophy itself. ThomasHobbes’ mushroom metaphor abstracts

from the ways that society shapes the de-velopment of its members in favor of“taking individuals as they are.” Reflect-ing this canon, most economists have notasked how we come to want and valuethe things we do.

Hobbes’ fiction neatly elides the in-fluence of social arrangements on theprocess of human development and thusgreatly simplifies the task of economictheory. But the scope of economic in-quiry is thereby truncated in wayswhich restrict its explanatory power,policy relevance, and ethical coherence.If preferences are affected by the poli-cies or institutional arrangements westudy, we can neither accurately predictnor coherently evaluate the likely con-sequences of new policies or institu-tions without taking account of prefer-ence endogeneity. In the pages which

75

follow I review models and evidenceconcerning the impact of economic in-stitutions on preferences, broadly con-strued, and comment on some implica-tions for economic theory and policyanalysis.1

The production and distribution ofgoods and services in any society is or-ganized by a set of rules, among whichare allocation by fiat in states, firms,and other organizations, patriarchal andother customary allocations based ongender, age, and kinship (as for exampletakes place within families), gift, theft,bargaining, and of course markets. Par-ticular combinations of these rules giveentire societies modifiers such as “capi-talist,” “traditional,” “communist,” “pa-triarchal,” and “corporatist.” These dis-tinct allocation rules along with otherinstitutions dictate what one must do orbe to acquire one’s livelihood. In sodoing they impose characteristic pat-terns of interaction on the people whomake up a society, affecting who meetswhom, on what terms, to perform whichtasks, and with what expectation of re-wards.

One risks banality, not controversy, insuggesting that these allocation rulestherefore influence the process of hu-man development, affecting personality,habits, tastes, identities, and values.One cannot be too far out on a limbwhen in the company of Adam Smith aswell as Edmund Burke, Alexis de Toc-queville and Karl Marx, John StuartMill and Frederick Hayek: all cele-brated or lamented the effects of mar-

kets and other economic institutions onhuman development.2 But consensuseludes any of the grand claims madeconcerning the nature of the effects orhow they might be generated. The rea-son is that most writers have implicitlyinvoked a kind of functionalist corre-spondence between economic struc-tures on the one hand and values, cus-toms, and tastes on the other, withoutexplaining the mechanisms by which theformer might affect the latter. Thought-ful works on the subject—JosephSchumpeter (1950) on “the civilizationof capitalism,” Daniel Bell (1976) on“the cultural contradictions of capital-ism,” David Potter (1954) on the “peo-ple of plenty,” Karl Polanyi (1957) on“the great transformation,” or PeterLaslett (1965) on “the world we havelost”—are surprisingly bereft of causalarguments.

Nonetheless, the argument that eco-nomic institutions influence motivationsand values is plausible, and the amountof evidence consistent with the hy-pothesis is impressive. Many ethno-graphic and historical studies, for exam-ple, recount the impact of moderneconomic institutions on traditional orindigenous cultures.3 The rapid rise offeminist values, the reduction in familysize, and the transformation of sexualpractices coincident with the extensionof women’s labor force participationlikewise suggest that changes in eco-nomic organization may foster dramaticchanges in value orientations.

Drawing on literatures from the other1 I abstract from other forms of preference en-

dogeneity such as the many variants of HarveyLeibenstein’s (1950) “bandwagon” and “snob” ef-fects or James Duesenberry’s (1949) analysis ofkeeping up with the Jones’ or Thorstein Veblen’s(1934) emulation effects, or the interdependentpreferences studied by Robert Pollak (1976).Rather, I here develop the research agenda sug-gested by Herbert Gintis’ early (1971, 1972) inves-tigations of the impact of economic institutions onpreferences.

2 In these pages and elsewhere, AlbertHirschman (1977, 1982) has catalogued earlystatements of the cultural effects of markets, obvi-ating the need for more than passing mentionhere.

3 Among the more instructive not mentionedelsewhere in this essay are Kenelm Burridge(1969), Daniel Lerner (1958), Margaret J. Field(1960), T. Scarlett Epstein (1962), Michael Taus-sig (1980), and Jean Ensminger (1992).

76 Journal of Economic Literature, Vol. XXXVI (March 1998)

social sciences, history, and experimen-tal economics, I have identified five ef-fects of markets and other economic in-stitutions on preferences. Few aresupported by empirical evidence thatwill convince a confirmed skeptic, butmost are plausible and consistent withsubstantial evidence.

Framing and situation construal: eco-nomic institutions are situations in thesocial psychological sense and thus haveframing and other situation construaleffects; people make different choicesdepending on whether the identical fea-sible set they face is generated by amarket-like process or not (I addressthese issues in Section 4).

Intrinsic and extrinsic motivations:the ample scope of market choices andoften extrinsic nature of market rewardsmay induce preference changes drivenby individual desires for feelings ofcompetence and self-determination;other institutions may have related ef-fects (Section 5).

Effects on the evolution of norms:economic institutions influence thestructure of social interactions and thusaffect the evolution of norms by alter-ing the returns to relationship-specificinvestments such as reputation-build-ing, affecting the kinds of sanctions thatmay be applied in interactions, andchanging the likelihood of interactionfor different types of people (Section6).

Task performance effects: economicinstitutions structure the tasks peopleface and hence influence not onlytheir capacities but their values andpsychological functioning as well (Sec-tion 7).

Effects on the process of culturaltransmission: in part for the abovereasons, and in part independently,markets and other institutions affectthe cultural learning process itself,altering the ways we acquire our values

and desires, including child rearingand schooling, as well as informal learn-ing rules such as conformism (Section8).

Until recently, economic theory gavelittle guidance in understanding theseeffects, for it purposefully abstractedfrom what were considered to be the ir-relevant sociological details of the ex-change process. In the complete con-tracting world of Walrasian economics,for example, there is little reason for aneconomic actor to be concerned abouthis exchange partner’s psychologicalmakeup or moral commitments; more-over there is no way that these personaltraits could be affected, if one were soconcerned. Markets of this type, wroteAlbert Hirschman in these pages (1982,p. 1473), are peopled by “large numbersof price taking anonymous buyers andsellers supplied with perfect informa-tion” . . . and “function without any pro-longed human or social contact amongor between the parties.” Grocery mar-kets approximate this ideal (a fact whichmay explain why fruit stands and fishmarkets figure so prominently in eco-nomics textbooks).

By contrast, now-standard micro-economic theories of labor, credit, andother markets as well as the contempo-rary theory of the firm treat economicinteractions as personal, strategic, anddurable connections among peoplewhose identities matter for the out-comes. Aspects of social life oncethought to be the province of psychol-ogy or sociology are thus seen to be es-sential to the explanation of the breadand butter of economics: prices andquantities. The theory of asymmetric in-formation and incomplete contractingshows that markets may not clear incompetitive equilibrium, leading toasymmetries between those on the shortside of the market (able to secure allthe transactions they desire) and those

Bowles: Endogenous Preferences 77

on the long side of the market, some ofwhich may be unable to secure anytransaction at all. An important conse-quence is the reappearance of complex,asymmetrically placed, opportunistic,and (especially) malleable economic ac-tors more reminiscent of the flesh andblood dramatis personae of classicaleconomics than the anemic and one-di-mensional homo economicus of thestandard textbooks.

Two aspects of exchanges with incom-plete contracts account for this. First,where contracts are incompletely speci-fied or costly to enforce, the ex postterms of an exchange may depend onthe normative commitments and psy-chological makeup of the parties to theexchange; where the amount of workdone on the job cannot be secured by acontract it will be influenced by the em-ployee’s work ethic or sense of aliena-tion, for example.

Second, because of the durability ofthe exchange, one or both parties mayhave the capacity to structure the rela-tionship so as to affect the preferencesof their exchange partner (Bowles andGintis 1993; Mulligan 1997). Paternalis-tic policies in lifetime employmentfirms are an example. Incomplete con-tracts thus provide both the motivationand the means for deliberate (as well asunwitting) preference modification inthe exchange process.

Models of incomplete contracts notonly dramatize the shortcomings of theexogenous preferences assumption, theyalso provide a basis for a more nuancedtreatment of the effects of markets andother economic institutions on prefer-ences. Walrasian grocery markets sup-port personal interactions quite distinctfrom the long term relationship charac-teristic of a lifelong employment firm;and the differences in the structure ofthese exchanges appear to have effectson preferences, as we will see presently.

Or, to take another example, there aresignificant differences in the personal-ity effects on participants in marketswhich clear in equilibrium and thosewhich do not, and in those marketswhich do not clear, for people on theshort side of the market (whose advan-tageous positions may allow them tomake take it or leave it offers) andthose on the long side of the market,some of whom are simply excludedfrom the exchange process, while othersfear losing the transaction they have se-cured. Thus the details of market struc-tures—and in particular the ways inwhich social interactions are pat-terned—may be important.

I turn first to methodological issues.In the next section I ask what we meanby preferences and how they might beinfluenced by economic institutions;and in Section 3 I present a model illu-minating the influence of economic in-stitutions on the process of cultural evo-lution.

2. Social Interactions and the Evolutionof Preferences

I do not know the fruit salesman personally;and I have no particular interest in his well-being. He reciprocates this attitude. I do notknow, and I have no need to know whetherhe is in direst poverty, extremely wealthy, orsomewhere in between . . .Yet the two of usare able to . . . transact exchanges efficientlybecause both parties agree on the propertyrights relevant to them.

James Buchanan (1975, p. 17)

Preferences are reasons for behavior,that is, attributes of individuals that(along with their beliefs and capacities)account for the actions they take in agiven situation. To explain why a personchose a point in a budget set, for exam-ple, one might make reference to hercraving for the chosen goods, or to a re-ligious prohibition against the excludedgoods. Conceived this way, preferences


go considerably beyond tastes, as anadequate account of individual actionswould have to include values or whatAmartya Sen (1977) terms commit-ments and John Harsanyi (1982) callsmoral preferences (as distinct from per-sonal preferences). Also included arethe manner in which the individual con-strues the situation in which the choiceis to be made (Lee Ross and Nisbett1991), the way that the decision situ-ation is framed (Amos Tversky andKahneman 1986), compulsions, addic-tions, habits, and more broadly, psycho-logical dispositions. Preferences may bestrongly cognitively mediated—my en-joying ice cream may depend criticallyon my belief that ice cream does notmake me fat—or they may be visceralreactions—like disgust or fear—evokingstrong emotions but having only themost minimal cognitive aspects (RobertB. Zajonc 1980; David Laibson 1996;Loewenstein 1997; Rozin and Carol Ne-meroff 1990). The term “preferences”for these heterogeneous reasons forbehavior is perhaps too narrow, andruns the risk of falsely suggesting thata single model of action is sufficient;Patrick H. Nowell-Smith’s (1984) “proand con attitudes” or “reasons forchoosing” are more descriptive, but un-wieldy.4

For preferences to have explanatory

power they must be sufficiently per-sistent to explain behaviors over timeand across situations.5 If preferencesare endogenous with respect to eco-nomic institutions it will be importantto distinguish between the effects ofthe incentives and constraints of aninstitutional setup (along with givenpreferences) on behaviors, and theeffect of the institution on prefer-ences per se. The key distinction is thatwhere preferences (and not just be-haviors) are endogenous they willhave explanatory power in situationsdistinct from the institutional environ-ments which account for their adoption.Thus, however acquired, preferencesmust be internalized, taking on thestatus of general motives or constraintson behavior. Values which become du-rable attributes of individuals—for ex-ample, the sense of one’s own efficacyintroduced below—may explain behav-iors in novel situations, and hence areincluded in this broad concept of pref-erences.

We acquire preferences through ge-netic inheritance and learning. Verylong lasting economic institutions, suchas the social structures of the simple so-cieties which predominated in the first100,000 years (90 percent) of biologi-cally modern human existence, couldsubstantially affect gene distributionsin a population and hence could pro-vide part of a genetic explanation ofpreferences (Christopher Boehm 1993;Linnda Caporael et al. 1989; Feldmanand Kevin Laland 1996; William Dur-ham 1991). Nonetheless it seems likelythat the more important effects of eco-nomic organization on preferences op-erate through cultural transmission,that is, learning. Drawing on the exten-sive literature on food tastes, Clark

4 In order to account for an individual’s actionspreferences need not coincide with the reasonsgiven by the particular individual, of course. Nordo preferences alone generally give a sufficient ac-count of behaviors: my consumption of aspirin isaccounted for by my aversion to pain plus my be-lief that aspirin will relieve the pain and that thislittle white object is indeed an aspirin, and so on.

5 Benjamin Bloom (1964) documents stabilityover time of a range of measured personalitytraits. For particular psychological dimensions in-troduced below see Herman Witkin and JohnBerry (1975 p. 41; field independence), Paul An-drisani and Gilbert Nestel (1976 p. 161; internal-external locus of control), and Kohn and CarmiSchooler (1983 p. 147; self-directedness). Rossand Nisbett (1991) provide a critical review of the

evidence for intertemporal and cross situationalconsistency of behavior.


McCauley, Rozin, and Barry Schwartz(1994, p. 27) write:

The human being comes into the world withcertain likes and dislikes, such as innate dis-like of pain, bitter tastes, and many types ofstrong stimulation, and an innate liking forcertain types of touch or sweet tastes . . . Al-most the entire adult ensemble of likes anddislikes is acquired, presumably in the pro-cess of enculturation.

For this reason I will treat preferencesas cultural traits, or learned influenceson behavior: liking ice cream, or neverlying, or reciprocating dinner invitationsare cultural traits.6

We know surprisingly little about howwe come to have the preferences we do;the theory of cultural evolution is thussimilar to the theory of natural selectionprior to its integration with Mendeliangenetics. While it is comforting to recallthat Darwin’s contribution was possibleeven though he did not know how traitsare passed on, this lacuna is nonethelessa major impediment to endogenizingpreferences. We know that intentionalmotivations are sometimes involved;one learns to appreciate classical musicbecause one notices that aficionados ap-pear to enjoy it (Gintis 1972; Bowlesand Gintis 1986; Gary Becker 1996).But instrumental motivations may be oflimited importance compared to otherinfluences such as mere exposure(Zajonc 1968), the unintended conse-quences of activities motivated by otherends (such as migration to a different

culture in search of work), and confor-mism (Solomon Asch 1952; MuzaferSherif 1937; Theodore Newcomb et al1967). While the individual benefits ac-cruing to those exhibiting particularcultural traits may affect these learningprocesses and hence the rate at whichthe traits are replicated, most prefer-ences are not chosen in the usual senseof intentional action toward given ends.Rather, preferences are learned as anaccent or a taste for a national cuisine isacquired, that is, by processes whichmay but need not be intentional.

However acquired, preferences areinternalized: there is considerable evi-dence that preferences learned underone set of circumstances become gener-alized reasons for behavior. Thus eco-nomic institutions may induce specificbehaviors—self-regarding, opportunis-tic, or cooperative, say—which then be-come part of the behavioral repertoireof the individual. The effects of mereexposure just mentioned provide a par-ticularly transparent example: “likes” orhabits initially induced by exposure orrepetition become permanent reasonsfor behavior.

Learning by doing is another mecha-nism for the generalization of prefer-ences: behaviors found successful incoping with the tasks defined by onesphere of life are generalized to otherrealms of life. Paul Breer and EdwinLocke (1965, p. 253) present substan-tial experimental evidence to this ef-fect. They asked subjects to performdifferent sets of tasks and investi-gated changes in apparently unrelatedvalues:

In a period of less than four hours and with-out a single verbal reference to family, frater-nity, way of life, or any of the other areasmeasured, we succeeded in changing a widevariety of attitudes ranging from specific be-liefs about the most effective way to organizea work group, to abstract values concerning

6 The pioneering works in the formal theory ofcultural evolution are Luigi Cavalli-Sforza andFeldman (1981) and Boyd and Peter Richerson(1985). Robert LeVine (1973) earlier developedwhat he termed a Darwinian “variation-selectionmodel to culture-personality relations.” SarahOtto, Freddy Christiansen, and Feldman (1995),Robert Plomin and his collaborators (see Plominand Denise Daniels 1987) as well as Rozin (1991).Rozin and T. A. Vollmecke (1986) provide evi-dence that food tastes, psychological functioning,and other traits are far from exhaustively deter-mined by genetic inheritance.


the individual and society. This evidence wastaken to mean that task experience is capableof exerting a very powerful influence on allsorts of beliefs, values, and preferenceswhich, to the casual observer, appear to beonly remotely related to the task itself.

Finally, preferences may become gener-alized through a process which LeonFestinger (1957, p. 260) termed disso-nance reduction:

the human organism tries to establish inter-nal harmony, consistency or congruity amonghis opinions, attitudes, knowledge, and val-ues. . . . there is a drive toward consonanceamong cognitions.

The cognitive elements in dissonancecould be one’s values and a behavior, aswhen one is doing something which is in-consistent with one’s values. Festinger(1957, pp. 271–73) frequently used thisreasoning to explain “specific ideologicalchanges or opinion changes subsequentto the change in a person’s way of life”such as a:

sudden change in the job which a persondoes. A worker in a factory, for example maybe promoted to the job of foreman. Suddenlyhe finds himself giving orders instead of re-ceiving them . . . these new actions will bedissonant in many instances with opinionsand values which he acquired as a worker andstill holds. In pursuit of dissonance reduc-tion, one would expect this person to quiterapidly accept the opinions and values ofother foreman, that is, opinions and valueswhich are consonant with the things he nowdoes.

Dissonance reduction thus provides an-other explanation for how economiccircumstances may induce new prefer-ences, and why the new preferencesmight become general reasons for behav-ior.

In contrast to the social interactionsbased approach taken here, many wouldemphasize the role of religious or politi-cal indoctrination or advertising in pref-erence change. These intentional formsof inculcation are undoubtedly impor-

tant, but where empirical studies areavailable, other influences appear aspowerful if not more.7 If I am right thatacquiring preferences is akin to acquir-ing an accent, studies of languagechange may be illuminating. On the ba-sis of intensive empirical study of lin-guistic change in Philadelphia, for ex-ample, William Labov concluded that

linguistic traits are not transmitted acrossgroup boundaries simply by exposure in themass media or in schools. . . . Our basic lan-guage system is not acquired from schoolteachers or from radio announcers, but fromfriends and competitors: those who we ad-mire, and those who we have to be goodenough to beat. (Labov 1983, p. 23)

The inference is not that institutionssuch as schools and churches are unim-portant, but that understanding theirrole in the acquisition of cultural traitsmay be enhanced by seeing them—alongwith markets, firms, families, and gov-ernments—as distinct patterns of socialinteraction affecting the diffusion of cul-tural traits in a population in ways oftenunrecognized by any of the participants.

3. Economic Institutions and Cultural Evolution

[The 17th century Salem “witches” and theirdefenders were] a group of people who wereon the advancing edge of profound historicalchange. If from one angle they were diverg-ing from an accepted norm of behavior, from

7 Studies of preferences for brands of food,soap, movies, and other consumption items forwhich one would expect an important advertisingeffect indicate that personal contact is consider-ably more important than advertising in motivatingbrand changes (Elihu Katz and Paul Lazarsfeld1955.) Everett Rogers’ (1962) classic study of dif-fusion of innovations found personal communica-tion to be of substantial importance in the diffu-sion of both ideas and practices such as cookingmethods. Some behavioral changes may be in-duced simply by providing information; in thesecases media exposure appears to be effective. Butwhere information alone is insufficient (changes insmoking behavior, e.g.) face to face contact ap-pears to be more effective (June Flora, NathanMaccoby, and John Farquhar 1989).


another angle their values represented the“norm” of the future. In an age about to pass,the assertion of private will posed the direstpossible threat to the stability of the commu-nity; in the age about to arrive it would forma central pillar on which that stability rested.

Paul Boyer and Stephen Nissenbaum (1974, p. 109)

How might allocation rules affect thedifferential replication of culturaltraits? The gist of an answer is givenbest by a concrete example. ErichFromm and Michael Maccoby’s (1970,p. 232) study of social character in aMexican village led them to this conclu-sion:

In a relatively stable society (or class) with itstypical social character there will always bedeviant characters who are unsuccessful oreven misfits under the traditional conditions.However in the process of socioeconomicchange, new economic trends develop forwhich the traditional character is not welladapted, while a certain heretofore deviantcharacter type can make optimal use of thenew conditions. As a result the “ex deviants”become the most successful individuals orleaders of their society or class. They acquirethe power to change laws, educational sys-tems, and institutions in a way that . . . influ-ences the character development of succeed-ing generations. . . . deviant and secondarytrait personalities never fully disappear andhence . . . social changes always find the indi-viduals and groups that can serve as the corefor the new social order.

The traits of the “ex deviants” in this ex-ample enjoy heightened replication pro-pensities both directly (others may wantto emulate the successful) and indirectly,because bearers of the traits becomeprivileged cultural models, such asteachers.

The Fromm-Maccoby view is sup-ported by the field research of LeVinein Nigeria. Using David McClelland’s(1961) measures of achievement moti-vation and other value orientations,LeVine (1966) found significant differ-ences among distinct cultural groups

which were not explicable by religiousor educational influences. However, thepattern of personality differences wereconsistent with the hypothesis thatdistinct motivations had evolved asadaptations to longstanding geographi-cally determined differences in struc-tures of competitive economic opportu-nity among the various cultural groups.Moreover as market-based and othercompetitive systems of advancement be-came more generalized during thecourse of the twentieth century, thoseexhibiting high levels of achievementorientation—some of them presumably“deviants” in the premodern compli-ance-based rather than achievement-oriented subcultures—gained positionsof educational leadership, thus assum-ing roles as privileged cultural models.

These studies of Mexico and Nigeriasuggest that new economic arrange-ments might affect cultural evolution intwo ways: either by influencing the eco-nomic well-being of those exhibitingdistinct traits or by altering the learningrules which make up the process of cul-tural transmission itself. The culturaltransmission process translates eco-nomic well-being, exposure to rolemodels, and other influences into repli-cation of traits, and thus intervenes be-tween payoffs and replication.

Evolutionary game theoretic modelstypically abstract entirely from the pro-cess of cultural transmission, repre-senting payoffs associated with particu-lar traits as if they were the onlyinfluences on the replication of traits.8

By contrast, models of cultural evolu-tion typically address what is knownabout the particulars of the process by

8 One could interpret the payoffs in evolutionarygame theoretic models as the replication propensi-ties themselves, but while thus formally accommo-dating analysis of the process of cultural transmis-sion, this would add no insight to the distinctinfluences of the transmission process per se.


which traits are acquired, distinguishingbetween vertical transmission from par-ents, oblique transmission from non-pa-rental members of the previous genera-tion (for example, teachers), andhorizontal transmission from one’s owncohort (as in the case of languagechange or fashion; Penelope Eckert1988, 1982; Labov 1972, p. 304).9 Thesemodels, as well as the studies of Mexicoand Nigeria above, make it clear that atrait which is advantaged in the trans-mission process may diffuse in a popula-tion even if the economic benefits asso-ciated with the trait are inferior to thepopulation average. Thus the effects ofeconomic institutions on both payoffs todistinct traits and the cultural transmis-sion process must be studied.

A particularly important example ofhow a trait may be advantaged in thetransmission process is termed con-formist transmission: the prevalence ofa trait in a population may enhance thereplication propensity of each repre-sentative of that trait, independently ofthe payoff to those exhibiting the trait.Under quite general conditions wherelearning is costly, conformist transmis-sion may be efficient in the sense thatan individual who sometimes adoptstraits by simply copying what others aredoing rather than on the basis of thepayoffs associated with various actionswill do better than those who always en-gage in costly investigation of the rele-vant payoffs (Boyd and Richerson 1985;Feldman, Kenichi Aoki, and JochenKumm 1996). Conformist transmissionof preferences thus might have evolvedunder the influence of either genetic or

cultural inheritance. Frequency depen-dent replication may also arise wheregroups that are numerically preponder-ant are disproportionately likely to oc-cupy privileged roles as teachers orother cultural models. Persistent ethnicdifferences in food tastes, coupled withvery low vertical (parental) transmissionof these tastes (Rozin 1991) is a pieceof evidence suggesting the importanceof conformist transmission. Another,from empirical cross cultural psycho-logical studies, is the importance ofmembership in particular tribes as apredictor of values orientations, inde-pendently of sources of livelihood, ecol-ogy, and other possible influences(Robert Edgerton 1971).

The relationship between payoff-based and conformism or other fre-quency dependent influences on thereplication of cultural traits and theways that these may be influenced byeconomic institutions may be illustratedby means of a simple model based onBowles (1996). The basic intuition isthat the distribution of cultural traits ina population is determined as the equi-librium of a system whose exogenouselements are subject to long-term influ-ence of markets and other economic in-stitutions. Economic institutions affectthe evolution of preferences by chang-ing these exogenous determinants ofthe cultural equilibrium.

Suppose x and y are mutually exclu-sive cultural traits. Each member of alarge population is a “cultural model”with replication propensities, rx or ry,defined as the number of copies of eachmodel made at end of each period, pos-sibly a generation. Agents implementthe strategy dictated by their trait in agame which assigns benefits to each,following which the traits are replicatedthrough an updating process describedbelow, generating a new population fre-quency (one may think of the popula-

9 Empirical studies in this tradition includeKuang-Ho Chen, Cavalli-Sforza, and Feldman(1982) and Cavalli-Sforza et al. (1982); Boyd andRicherson (1985) survey many empirical studies ofthese three transmission processes. Alberto Bisinand Thierry Verdier (1996) present a model ofpreference evolution integrating the cultural evo-lution and evolutionary game theory approaches.


tion as composed of single parents eachwith a single child, who in the processof growing up may or may not adopt thetraits of the parent). Cultural equilib-rium is defined as a frequency of traitswhich is stationary.

Members of the population arepaired to play a two-person game, thepayoffs of which are denoted π(i, j), thepayoff to playing trait i against a j-play-ing partner. The “game” may be one ofthe familiar interactions of the hawk-dove, prisoners’ dilemma, or coordin-ation game type. It might refer to aninteraction as everyday as eating a mealtogether, or meeting in public, wherethe two traits might dictate matters ofstyle (“wearing a tie”) or taste (“wantinga drink”). Or it could refer to an ex-change of goods or some more conven-tionally economic interaction. The pay-off structure could be degenerate in thesense that my enjoyment of a beer maynot depend at all on what you are eatingor drinking; but the model is designedto address more interesting cases ofemulation, social dilemmas and the like.

For any population frequency of the xtrait, p ∈ [0,1] let µij = µij(p;δ) be theprobability of being paired with a j typeconditional on being an i type, whereδ ∈ [0,1] is a measure of the exog-enously determined extent to whichpairing is nonrandom.10 If pairing israndom δ = 0 and the probability ofmeeting an x type is simply p, irrespec-tive of one’s own type: µx = µyx = p.But where residence is correlated withtype, or where sorting by type takesplace by means of social networks orother groups, the probability of meetingone’s own type may exceed that given

by the population frequency. The ex-pected payoffs are thus

bx(p;δ) = µxxπ(x,x) + µxyπ(x,y)(1)

by(p;δ) = µyxπ(y,x) + µyyπ(y,y).

To take account of frequency depen-dent biases in cultural transmissionsuppose that conformist transmission isdescribed by the conformist bias func-tion σ(p), which we write as σx(p − k)and σy(k − p) where for simplicityσy = σx ≡ σ > 0 and k ∈ [0,1] is the valueof p for which no bias operates. Furtherwe define α ∈ [0,1], the degree of con-formism, as the weight placed on σ (p)as opposed to b(p; δ) in the transmis-sion process. Thus we have the replica-tion propensities:

rx = ασ(p − k) + (1 − α)(bx(p;δ) − by(p;δ)) + 1 (2)

ry = ασ(k − p) + (1 − α)(by(p;δ) − bx(p;δ)) + 1.

Where p = k, or if α = 0 conformisttransmission does not operate so replica-tion depends solely on payoffs, as in con-ventional evolutionary game theoreticmodels. Equilibrium is defined by dp/dt= 0, which for p ∈ (0,1) requires thatthe effects of conformist transmissionoffset the effects of unequal game out-comes so that rx = ry, or

ασ(p − k)⁄(1 − α) = by(p;δ) − bx(p;δ) (3)from which it can be seen that culturalequilibrium does not require equal pay-offs. Figure 1 illustrates this equilibriumcondition for the case of an interior sta-ble equilibrium. This equilibrium can beseen to be stable because for p > p* thepayoff advantage of the y trait (the right-hand side of (3)) more than offsets itsdisadvantage due to conformist transmis-sion (the lefthand side), as a result ofwhich ry > rx giving dp/dt < 0. So dis-turbances of p will be self correcting.

Markets and other economic institu-

10 The explicit relationship between the µ’s andp is this: define δ as the degree of segmentationof the population, then for p ∈ (0,1), µxx = δ +(1−δ)p and µxy = (1−δ)(1−p), from which it is clearthat δ is a non-genetic analogue to the “degree ofrelatedness” in biological models (W. D.Hamilton1975; Alan Grafen 1979).


tions will affect the distribution of cul-tural traits in the population becausethey influence the determination of theexogenous variables in the above model:

i) the rules governing who interactswith whom (as indicated by the func-tions µij(p,δ) measuring the degree ofsegmentation into distinct social net-works and other sources of nonran-dom pairing);ii) the payoffs π(i,j) to any given inter-action (determined by the frequencyof interaction, ease of recognition oftypes, for example);iii) the structure of the transmissionprocess itself (in this case the natureand strength of conformism, σ, k,and α, including the assignment ofsome types as compulsory or other-wise advantaged models, such asteachers).

In more complex models, allowing formovement among population groups,cultural equilibria are influenced by mi-gratory flows, which in turn are subjectto the influence of economic institutions(Bowles and Gintis 1998).

To say that economic institutionshave these effects is, of course, to com-pare markets, say, with some other allo-cation rule or to compare various typesof markets. Allocation rules are differ-ing mechanisms for coordinating thetransfer of goods and services. Econo-mists tend to focus on the relationshipsthereby established among the objectsof exchange, relative prices, for exam-ple. But allocation rules also establishrelationships among people, based onassignment to distinct positions withcorresponding rights, status and obliga-tions and patterns of interaction. Thusmarkets support interpersonal experi-ences distinct from other allocationrules. Robert Lane (1991), whose TheMarket Experience must be the startingpoint for any consideration of the psy-chology of markets, writes:

In spite of the variety of markets over timeand across cultures, I believe that it is possi-ble to conceive of a market experience that istypical, frequent, and paradigmatic for thosewho do market work for pay, use money andbuy—rather than make, inherit or receivefrom government—the commodities with

Figure 1. Cultural Equilibrium

payoffunits by(p;δ) - bx(p;δ)

(payoff effect)

ασ(p-k)1-α

(conformism effect)

k p* 1.0 p


which they adorn their lives. (p. 4). . . Inevi-tably the market shapes how humans flourish,the development of their existences, theirminds, and their dignity. (p. 17)

What is psychologically distinctiveabout markets as opposed to otherallocation mechanisms? Max Weber([1922]1978, p. 636) wrote “A marketmay be said to exist wherever thereis competition for opportunities ofexchange among a plurality of poten-tial parties.”11 Markets structure socialinteractions “each of which is specifi-cally ephemeral insofar as it ceasesto exist with the act of exchangingthe goods.” As a result, according toWeber,

The market community as such is the mostimpersonal relationship of practical life intowhich humans can enter with one another.This is not due to the potentiality of struggleamong the interested parties which is inher-ent in the market relationship. . . . The rea-son for the impersonality of the market is itsmatter-of-factness, its orientation to the com-modity and only to that.

In Weber’s view, then, markets—at leastideally—are characterized by imperson-ality, ephemerality of contact, and easeof entry and exit.12 This might betermed the economics textbook concep-tion of competitive markets.

Contrast these arrangements with

what Parsons (1967, p. 507) calls the“two principal competitors” of the mar-ket: “requisitioning through the directapplication of political power” and“non-political solidarities and communi-ties.” These allocation rules contrastwith markets in at least one of the char-acteristics—impersonality and epheme-rality—stressed by Weber. Centralizedbureaucratic allocations are in some re-spects as impersonal as markets—atleast ideally—but membership in thegroup defining the allocation is gener-ally given, entry and exit costs are high(often involving a change in citizenshipor at least residence), and contacts arefar from ephemeral. In contrast to bothbureaucratic and market allocation,kin-like directly interacting communi-ties with stable membership exhibitneither ephemerality nor impersonalityin their characteristic rules governingallocation.13 Figure 2 presents thesethree ideal types along with a fourth—ephemeral and personal social interac-tion—which I have termed ascriptivelyordered markets. Racially segmentedspot labor markets are an example, asthey are personal (the racial identitiesof the participants matter) but thecontact among participants is not on-going.

The contrast between personalizednon-market transactions and the puta-tive impersonality of market exchange

11 Georg Simmel (1900, p. 297) writes in similarvein that “money . . . is conducive to the removalof the personal element from human relationshipsthrough its indifferent and objective nature.” Tal-cott Parsons (1949, p. 688) describes markets simi-larly: “When a man walks into a store in a strangecity to make a purchase, his only relevant relationto the clerk behind the counter concerns mattersof kind of good, price, etc. All other facts aboutboth persons may be disregarded. Above all it isnot necessary even to know whether the two haveany further interests in common beyond the im-mediate transaction.”

12 On the psychological dimensions of distincteconomic arrangements, see Alan Page Fiske(1991, 1992). Among economists, Yoram Ben-Po-rath (1980, p. 4) provides the fullest description ofthe idealized interactions among market agents as“impersonal.” See also Zelizer (1996).

13 I have in mind allocation systems of the typedescribed by Emile Durkheim’s “organic solidar-ity” ([1933]1964), or Sahlins’ (1972) “generalizedreciprocity,” or Ferdinand Tonnies’ ([1887]1963)Gemeinschaft, or William Ouchi’s “clans” (1980).

Figure 2. Allocation Rules as Learning Environments

Communities

Personal

Ascriptive Markets

Bureaucracies

Anonymous

Ideal Markets

Durable

Ephemeral


is, of course, a matter of degree, par-ticularly in markets characterized bythe asymmetric information, incompletecontracts and hence importance oftrust, ongoing interaction, and sharedunderstandings of the type analyzed bythe theory of social exchange.14 Imper-sonality of contact and permeability ofboundaries, while characteristic of allmarkets by comparison to other alloca-tion rules, describe some markets moreaptly than others. Thus in assessing thecultural effects of markets it will benecessary to distinguish not only be-tween markets and other allocationrules such as bureaucracy and commu-nity, but among differing types of mar-kets as well.

I turn now to evidence and reasoningconcerning the five effects of economicinstitutions on preferences previewedin the introduction.

4. Markets, Situations, and Framing

Money is one of the shatteringly simplifyingideas of all time, and like any other new andcompelling idea, it creates its own revolution.. . . [The] Tiv [of Nigeria] have tried to cate-gorize money with other imported goods . . .to be ranked morally below subsistence. Theyhave, of course, not been successful in so do-ing. Paul Bohannan (1959, pp. 500, 503)

Markets frame choices; we will seethat a choice problem presented in amarket environment may induce behav-iors different from the identical prob-lem in framed in a non-market way.Consider an example of market framing:paying a tax and receiving a governmen-tal service differs relevantly from buy-ing the identical service on a market. Inthe first case one may—as a citizen—

feel entitled to the service (irrespectiveof the taxes paid) and may be unlikelyto compare the value of the service tothat of other goods and serviceswhether traded or not; in the secondcase one may feel that the good is ac-quired by dint of one’s talents as an in-come earner, and may readily compareits price with other traded goods andservices, the value of one’s own labortime, and the like. This framing effectmay thus be part of an account of why aparticular action was taken; if differentinstitutions induce different choicesfrom an identical choice set institutionsmay affect preferences. This is becausechoices made under the influence of in-stitutionally determined framing maylater be repeated even in the absence ofthe framing effect if the effects of expo-sure to the object of choice, or disso-nance reduction effects are strong; how-ever, I am aware of no evidence to thiseffect.

Markets thus affect behavior in waysnot fully captured by the fact that mar-ket-determined prices and endowmentsdefine the budget set: markets providepresumptive reasons why people pos-sess the goods they do, and they promptsome comparisons while inhibiting oth-ers. I call these the construal effects ofmarkets, borrowing the term from so-cial psychology in preference to the fa-miliar but narrower concept of framingeffects. The construal effects of marketsarise in large part because people ap-pear to have what might be termed rela-tional preferences: the terms on whichthey are willing to transact dependsboth on the perceived relationshipsamong the exchanging parties, and onrelated concepts of fairness. Markets af-fect both.

There is considerable experimentalevidence consistent with the impor-tance of the construal effects of mar-kets. Experimental markets and bar-

14 Kenneth Arrow (1971), George Akerlof(1984), Heinz Hollander (1990), Rachel Kranton(1996), Kreps (1990), have recently formalizedsome of Peter Blau’s (1964) and George Homans’(1958) early reasoning concerning social exchange.


gaining environments consistently yielddiscrepant results, with markets quicklyconverging to the competitive equilibriaimplied by self-regarding preferences,and bargaining games often yieldingevidence consistent with other-regard-ing or relational preferences.15 An ex-ample of the later is the comparativestudy of bargaining and market behav-ior in four cultures by Alvin Roth et al.(1991). In their study both marketand bargaining experiments were de-signed to have distributionally extremeequilibria, one player receiving all ofthe benefits. The market experimentsquickly converged to this equilibrium inall four cultures. By contrast, proposersin the ultimatum game (the bargainingsituation) made much higher than equi-librium offers; and substantial positiveoffers were frequently rejected.16 Posi-tive offers are also common in dictatorgames. These results are consistent witha large body of experimental evidenceby others, beginning with Werner Güth,Rolf Schmittenberger, and BerndSchwarze (1982).

A considerable body of research hassought to explain this now well estab-lished aspect of ultimatum and dictatorgame play, with interest centering onthe question (unrelated to my concernhere) of whether the unexpected resultswere motivated by intrinsically gener-ous preferences on the part of the pro-posers. But the subsequent experi-

ments, particularly those by ElizabethHoffman et al. (1994), have yieldedconsiderable insight on the construal ef-fects of markets. Hoffman and her col-laborators varied two aspects of the ex-perimental environment for ultimatumand dictator games: proposers eitherwon their position by doing well on atrivia quiz or were randomly assigned,and their relationship to their gamepartner was described either as an “ex-change” (with prices elicited by the ex-perimenter) or simply as “divide $10.”The combined “earned status” plus “ex-change” experimental condition ap-proximates a market arrangement inthat competitive success is not simply amatter of chance but is based on appar-ent accomplishment; and the exchangeframing of the game structure is tran-sparently more market-like than the “di-viding the pie” framework.17 Despitethe fact that the experimental situationwas otherwise identical, the two marketlike protocols yielded significantlysmaller offers in both the ultimatumgame and the dictator game.

In other experiments, market-likeanonymity generates behaviors differingfrom those induced by more personalsettings. Communication or other con-ditions contributing to group identity ora reduction in social distance among ex-perimental subjects increases contribu-tions in public goods games (JohnLedyard 1995; Robyn Dawes, Alphons

15 Camerer and Richard Thaler (1995) is a re-cent survey of results for ultimatum and dictatorexperiments. Alternating offer bargaining experi-ments yield similar results (Jack Ochs and Roth1989).

16 A high offer in an ultimatum game may notindicate generosity toward the other player, butrather the anticipation that low offers will be re-jected. But the rejection of a positive offer is otherregarding (perhaps motivated by spite or a com-mitment to reciprocal fairness) so I represent highoffers by the proposer as evidence of other regard-ing behavior (either that of the proposer or theproposer’s anticipation of the responder).

17 If, as I have suggested, markets enhance theperception that one’s possessions are acquired bymerit rather than by chance, the extent of endow-ment effects—the unwillingness to part with goodsin one’s possession for prices considerably higherthan the maximum price one would have paid toacquire them—may be subject to market effects.Experimental subjects reported by Loewensteinand Samuel Issacharoff (1994) were particularlyunwilling to part with objects in their possession ifthey came to possess them as a result of winningan inconsequential contest. Thus market-acquiredgoods may be more subject to endowment effectsthan goods acquired as gifts or public transfers.


van de Kragt, and John Orbell 1988)and induces cooperative play in prison-ers’ dilemma interactions (Peter Kol-lock 1997). Simon Gaechter and Fehr(1997) find that in a public goods inter-action, even quite minimal (experimen-tally induced) social familiarity amongsubjects enhances the impact of socialapproval incentives (implemented by expost revelation of subjects’ identitiesand contributions); when familiarity andthe public revelation of one’s contribu-tions is combined, a significant increasein participation results.

On the basis of experimental evi-dence from dictator and ultimatumgames, Schotter, Avi Weiss, and InigoZapater (1996, p. 38) offer this conclu-sion:

The morality of economic agents embeddedin a market context may . . . be quite differ-ent from their morality in isolation. While weare not claiming that people change their na-ture when they function in markets, it may bethat the competition inherent in markets andthe need to survive offers justifications for ac-tions that in isolation would be unjustifiable.

A striking example illustrating thissuggestion is found in Catherine Andreand Platteau’s (1997, p. 32) study ofthe impact of the land market inRwanda:

customary obligations attached to lineagelands, in particular obligations to redistributeland in favor of landscarce kith and kin, ceaseto apply when the lands are acquired througha purchase instead of being handed downwithin the lineage.

The experimental results might besummarized by saying that the more theexperimental situation approximates acompetitive (and complete contracts)market with many anonymous buyersand sellers, the less other-regarding be-havior will be observed.18 Because the

above market and non-market experi-ments were conducted with the samesubject pools, these results are consis-tent with the view that market-like situ-ations induce self-regarding behavior,not by making people intrinsically self-ish, but by evoking the self-regardingbehaviors in their preference reper-toires. Thus, the hypothesis that marketsituations induce self-regarding behav-ior does not imply that those living innon-market societies would be intrinsi-cally less self-regarding.19

Where competitive markets approxi-mate the law of the single price andwhere the extent of markets is such thatfew things do not have a price, marketshave further construal effects: they fa-cilitate comparison among disparate ob-jects. The appropriate comparison is tosettings (in families or in so-called“primitive exchange”) in which goodsmay be transacted at vastly different ex-change ratios depending on the socialrelationships among the parties to theexchange.20 Reporting on a pre-marketsociety in southeastern New Guinea,Raymond Firth (1958, p. 69) writes:“There is . . . no final measure of thevalue of individual things, and no com-mon medium whereby every type ofgood and service can be translated intoterms of every other.” Well workingmarkets, by contrast, favor thinking ofgoods both abstractly (bananas in gen-eral, not this particular banana) and

18 On the differing outcomes of anonymous andface to face bargaining see Roth (1995) and RoyRadner and Schotter (1989).

19 In fact high (“other regarding”) offers in theultimatum games of Roth et al. were made in whatwould appear to be the most and least market-likesocieties in the sample—U.S. and the former Yu-goslavia (by contrast to Israel and Japan). In thepublic goods experiments by Steven Kachelmeierand Mohamed Shehata (1997) subjects in Beijingacted no different than the Canadian subjects un-der conditions of anonymity but proved signifi-cantly less self regarding when the identity of theplayers was public knowledge.

20 Sahlins’ (1972) theory of primitive exchange isdistinct from market exchange precisely in this de-viation from the law of the single price.


comparatively (objects seen as repre-senting more or less market value, di-vorced from their particular uses orproperties). Markets are thus powerfulcognitive simplifiers, allowing radicalreductions in the complexity with whichone typically views an assortment of dis-parate goods.

A dramatic example is provided Bo-hannan’s study (1959) of the extensionof markets in an African subsistenceeconomy, that of the Tiv in Nigeria.

The most distinctive feature about the econ-omy of the Tiv—and it is a feature they sharewith many, perhaps most, of the pre-mone-tary peoples—is what can be called a multi-centric economy . . . in which a society’s ex-changeable goods fall into two or moremutually exclusive spheres, each marked bydifferent institutionalization and differentmoral values. (p. 492)

Among the Tiv, domestic goods, women,and prestige goods were all exchanged,and in the latter there was a monetaryequivalent (brass rods) but “no one, savein the depths of extremity, ever paidbrass rods for domestic goods” (p. 493),while “rights in women had no equiva-lent or ‘price’ in brass rods or in anyother item save, of course, identicalrights in another woman. . . . Exchangeswithin a category . . . excite no moraljudgements. Exchanges between catego-ries, however, do excite a moral reac-tion” (p. 496). The extension of general-ized markets, and with them money,eroded these arrangements:

General purpose money provides a commondenominator among all the spheres, thusmaking the commodities within each express-ible in terms of a single standard and henceimmediately exchangeable. (p. 500)

Among the Tiv, the set of permissible ex-changes has expanded with the advent ofmarkets and basic notions of what itmeans to have a well-ordered life havechanged.

5. Markets and Motivation

In the realm of ends everything has either aprice or dignity. Whatever has a price can bereplaced by something else which is equiva-lent; whatever is above all price, and there-fore has no equivalent, has dignity.

Immanuel Kant ([1785]1949, p. 182)

The reward structures of marketsmay affect motivation independently offraming effects. The impersonality andephemerality of contact which charac-terizes markets (by contrast to other al-location rules) imply that a market“transaction entails a full quid pro quo(with) no left-over business or outstand-ing balance” (Ben-Porath 1980, p. 4) Bydefault, then, the incentives relevant toactivities governed by markets fre-quently center on the quid pro quo andtake the form of what social psycholo-gists term extrinsic rather than intrinsicrewards or sanctions, namely rewardsunrelated to the activities being moti-vated. On the basis of dozens of experi-ments by social psychologists over thepast 30 years one may conclude that thesalience of extrinsic reward in marketactivities will have effects on prefer-ences.

A series of well-designed experimentsshow that the degree to which an activ-ity is liked may be reduced by inducingsubjects to engage in the activity as ameans toward an extrinsic goal, such asbeing paid. The nature of the extrinsicreward is unimportant as long as it isclearly a quid pro quo. Mark Lepper,David Greene, and Nisbett (1973, p.130) write “Contracting explicitly to en-gage in an activity for a reward (will)undermine interest in the activity, evenwhen the reward is insubstantial ormerely symbolic.” Correspondinglywhen people are induced to engage inan activity with little or no extrinsic re-ward, they come to value the activitymore highly, that is, they come to be-


lieve that their actions were intrinsicallymotivated (Lepper and Greene 1978;and Edward Deci and Richard Ryan1985; Deci 1975).

Similar changes in evaluations in-duced by extrinsic rewards have beenshown to affect subsequent behavior innon-experimental situations. Frey andFelix Oberholzer-Gee (1997) foundthat proposing financial compensa-tion reduced Swiss citizens’ willingnessto host a nuclear waste facility. RichardTitmuss’ (1971) claim that elicitingblood donations by monetary incentiveshad perverse effects on preferenceslacked compelling evidence, as waspointed out by Arrow (1972) and Chris-topher Bliss (1972). However a fieldexperiment by William Upton (1974)partially supports Titmuss’ sugges-tion. Among 1,261 prospective blooddonors in Kansas City and Denver,some were offered financial induce-ments, others not. Among those in-itially exhibiting strong motivations tocontribute blood (as indicated by pastdonations), those offered financialinducements were substantially lesslikely to actually donate blood thanthose offered no financial reward.Among those expressing low intrinsicmotivation, however, the prospect offinancial reward had a (not statisticallysignificant) positive effect on eventualdonation.

The underlying psychological mecha-nism appears to be a fundamental de-sire for “feelings of competence andself determination” which are associ-ated with intrinsically motivated behav-iors (Deci 1975). Relatedly, a person’sperceived degree of self-determinationin making a choice influences the evalu-ation of the things over which thechoice is being made. For example, riskimposed by others is weighed morenegatively than risk chosen by the sub-ject. (See Chauncy Starr 1969; see also

Camerer and Howard Kunreuther1989.)

While the evidence for extrinsic re-ward and other self-determination ef-fects on preferences appears quitestrong, the relevant data provide littlesupport for the anti-market normativeinferences sometimes thought to follow.First, the evidence does not implicatemonetary rewards per se, but rather anyextrinsic reward (including negative re-wards such as punishments or admoni-tions). Moreover, distinctly non-marketaspects of governance—close supervi-sion, externally imposed time limits forwork tasks for example—appear to havesimilar effects (Lepper and Greene1978, p. 121). Paying someone to per-form a task which they might willinglyhave done without pay seems likely toundermine motivation; but this says lit-tle about the relative effectiveness ofthe various ways—pay, supervision,threat of job loss, etc.—to induce peo-ple to undertake tasks which they wouldrather not do. Second, while the extrin-sic nature of market rewards may un-dermine motivations, the wide range ofchoices often afforded in market situ-ations may support the sense of self-determination and thus induce positivemotivational effects.

6. Markets, Reputations, and Norms

The real reason why all these economic obli-gations [among the Trobriand Islanders] arenormally kept, and kept very scrupulously, isthat failure to comply places a man in an in-tolerable position . . . The honourable citizenis bound to carry out his duties, though hissubmission is not due to any instinct or intui-tive impulse or mysterious “group senti-ment,” but to the detailed and elaborateworking of a system, in which every act hasits own place and must be performed withoutfail . . . . every one is well aware of its exist-ence and in each concrete case he can fore-see the consequences.

Bronislaw Malinowski (1926, p. 40)


“Market-like arrangements” wroteCharles Schultze (1977, p. 18) “reducethe need for compassion, patriotism,brotherly love, and cultural solidarity.”Minimizing the demand on what mightnow be termed social capital, plus theconviction that the market system is, asHayek (1948, p. 11) wrote “a system un-der which bad men can do least harm,”are among the attractive features ofmarkets. This is not to say, however,that markets make norms redundant;where contracts are incomplete or un-enforceable, trustworthiness and othernorms facilitate exchange. Arrow (1971,p. 22) writes:

In the absence of trust . . . opportunities formutually beneficial cooperation would haveto be foregone . . . norms of social behavior,including ethical and moral codes [may be]. . . reactions of society to compensate formarket failures.

But if, as Schultze and Hayek say, mar-kets make fewer demands on people’selevated motivations, the impersonal andephemeral nature of market interactionsalso affect the benefits and costs of ac-quiring cultural traits affecting sociallyvalued behaviors. Markets thus affect notonly the demand for, but also the supplyof cultural traits. Among these are repu-tations for trustworthiness, generosity,and vengefulness.

Where markets govern the exchangeof well defined (meaning third party en-forceable) property rights, reputationsof any kind will tend to be both costlyfor people to acquire and of little bene-fit to those who do, and for these rea-sons unlikely to be favored by differen-tial replication. A consequence is thatmarkets lack the personal element ofnon-market connections, and as Ben-Porath (1980, p. 18) writes, with “[t]hedevelopment of markets . . . the bene-fits from a connection decline as iden-tity becomes less important.” Thuswhere markets approximate the ideal

complete-contracting assumptions ofthe standard model, the adverse conse-quences of lack of trustworthiness orgenerosity may be attenuated; but atthe same time markets may militateagainst the evolution of these traits.Thus markets may undermine the re-production of traits necessary for effi-cient market transactions in the absenceof complete contracting.

To see how this might be the case, Iwill consider a subset of norms which Icall nice traits; these are behaviorswhich in social interactions conferbenefits on others. Others would like tobe paired with those exhibiting nicetraits in an interaction. Included aresuch strategies as conditional or uncon-ditional cooperation in a prisoners’ di-lemma game, contributing rather thanwithholding in a public goods game, orplaying dove in a hawk-dove game.21 Asmy subsequent examples will confirm, itis not possible to generalize about theeffect of markets on socially valuednorms: “nice traits” may sustain collu-sion where competition would be moresocially beneficial, for example (Rose-Ackerman 1997). Using a model similarto that presented above, I (1996) showthat allocation rules which closely con-form to idealized markets may supportlower equilibrium population frequen-cies of nice traits, by comparison withalternative allocation rules which devi-ate from the market ideal. The intuitionbehind this result is that behaviors de-termined by nice traits affect others innon-contractible ways, and the regula-tion of non-contractible behaviorsthrough market-like interactions gener-

21 Bowles (1996) gives the following definition:in a population with traits x and x′ the latter indi-cating all other traits, x is a nice trait if π(x,x) >π(x,x′), π(x′,x) > π(x′x′), and π(x,x) > π(x′,x′).Other nice traits are “learn” rather than “imitate”in games of conformism and learning of the typestudied by John Conlisk (1980) and Boyd andRicherson (1993).


ates analogues to familiar market failureswhich in many cases may be attenuatedby deviations from the market ideal.

To see why this might be true con-sider the various ways—identified by bi-ologists, students of cultural evolution,and evolutionary game theorists—thatnice traits might flourish in a largepopulation. All, I will suggest, may beweakened by the impersonality andephemerality of contact that charac-terize markets. First, frequently re-peated interaction of a given pair of in-dividuals provides opportunities tosanction violations of norms and to re-ward nice traits. By contrast to other al-location rules, the ephemerality andanonymity of market interactionsclearly militate against repeated pair-ings and hence against this mechanismfor supporting nice traits. Second, fre-quent interaction of a limited numberof people likewise lowers the cost of ac-quiring information about the recentbehaviors of others, thus increasing thevalue of acquiring a reputation for be-ing “nice.”22 The impersonality andephemerality of contact in marketsclearly militate against these mecha-nisms favoring nice traits.

The third mechanism—segmenta-tion—is less familiar to economists,having been introduced by biologists as“games among relatives.”23 Where, as inthe model introduced above, popula-

tions are segmented so that individualsof a given “type” tend to interact dis-proportionately with one another, nicetypes will be favored. For example, ifbecause of geographical or cultural seg-mentation, the probability of interact-ing with one’s own type is greater thanthe population frequency of the trait,the equilibrium level of the nice traitwill exceed the equilibrium distributionof traits in a population under randompairing. The reason is that segmentationpartially internalizes the externality as-sociated with the nice trait: the non-random pairing means that the benefitsof niceness are disproportionately likelyto be conferred on others bearing nicetraits, thereby favoring the replicationof nice traits. To the extent that the im-personality (and hence anonymity) ofmarkets erodes the bases of segmenta-tion, markets inhibit this mechanismthat fosters the proliferation of nicetraits.

Finally, socially beneficial culturallytransmitted traits may evolve if thepressure of cultural group selection issufficiently strong. This occurs whenthe prevalence of nice traits in a sub-group enhances the average perfor-mance of the group sufficiently to allowthe trait to proliferate even if it is dis-advantaged in replication within eachgroup. Group selection pressures varywith the extent of group differences inthe distribution of traits among the sub-groups in a large population, which inturn depends on the level of migrationamong groups and the extent to whichthe formation of new groups contrib-utes to between group differences, forexample by favoring the formation ofgroups which are more homogeneousthan the population as a whole (Boydand Richerson 1990). High entry andexit costs and other supports for popula-tion segmentation sustain the group dif-ferences which render the pressure of

22 Bowles and Gintis (1997a) define a level of“optimal parochialism” based on the latter mecha-nism in a model of endogenous group formation ina large population. In a population paired to playone shot prisoners’ dilemma games with the strat-egy set augmented by an opportunity to pay a costto determine the type of the other agent and tocooperate if the other is either a cooperator or an“inspector,” the fraction of defectors in a (stable,interior) equilibrium population distribution varieslinearly with the cost of inspection.

23 Grafen (1979) and Robert Axelrod and Wil-liam Hamilton (1981). Bowles (1996) and Bowlesand Gintis (1998) study the effect of segmentationon the evolution of nice traits in a large popula-tion.


group selection effective. The low entryand exit costs typical of markets—bycomparison to other allocation rules—weaken the pressures of group selec-tion.

Figure 3 summarizes these fourmechanisms supporting the replicationof nice traits, and the manner in whichmarket allocation rules may underminethem.24 Because these conjectures pre-dict differences between the distri-bution of cultural traits in whole popu-lations, adequate testing would requirecomparative data in which entire popu-lations governed by more or less marketlike arrangements are the units ofanalysis. As the following studies sug-gest, however, less demanding tests us-ing experimental data on a commonpool of subjects under varying institu-tional conditions suggest that deviations

from idealized markets may induce“nice behaviors.”

We have seen already that even inexperimental markets characterizedby complete contracting and distri-butionally extreme (unfair) equilibriumoutcomes, the competitive equilibriumimplied by self-regarding preferencesis rapidly obtained in a wide varietyof subject pools. This does not occurin experimental markets with incom-plete contracts. In a series of experi-ments, Fehr and his co-authors havefound that contractual incompletenessinduces a pattern of reciprocity amongsubjects which has durable effects oncompetitive equilibrium (Fehr andJean-Robert Tyran 1996; Fehr et al.1997; and Fehr, Gaechter, and GeorgKirchsteiger 1997). For example in anexperimental labor market in which ef-fort is selected by the “worker” after awage offer is made by the “firm,” thesubgame perfect equilibrium based onself-regarding preferences in a one time

Figure 3. How Markets May Discourage the Evolution of Nice Traits

Enhanced groupselection pressuresfavor nice traits

Advantageouspairing for thosewith nice traits

Enhanced valueof reputationsfor niceness

Punishment of antisocial behaviors

Effect favoring the replication of nice traits

Limited intergroupmigration, non-randomgroup formation

Non randompairing of agents

Low cost ofinformation about others

Frequent or longlasting interactions

Necessary structural characteristic

Unlikely given lowentry and exit costsand impersonality

Unlikely givenimpersonality, lowentry and exitcosts

Unlikely givenimpersonality, ephemerality

Unlikely givenephemerality

Relationship to idealized markets

Reputation Kreps (1990),Carl Shapiro (1983),Bowles and Gintis (1997a)

Group selectionDavid Wilson and ElliotSober (1994), Boyd andRicherson (1990)

SegmentationGrafen (1979), Hamilton(1975), Axelrod-Hamilton(1981)

RetaliationTaylor (1976), DrewFudenberg-Eric Maskin(1986)

Model

24 The bibliographic references merely give ex-amples of the relevant models; inferences con-cerning the relationship between economic insti-tutions and the evolution of norms are my own.


interaction (offer the lowest wage, pro-vide the lowest effort level) does not oc-cur. Rather, “firms” offer wages higherthan necessary and “workers” recipro-cate by working harder than the mini-mum.

Relatedly, Kollock (1994, p. 341) in-vestigated “the structural origins oftrust in a system of exchange, ratherthan treating trust as an individual per-sonality variable” with similar results.Using an experimental design based onthe exchange of goods of variable qual-ity, Kollock found that trust in and com-mitment to trading partners as well as aconcern for one’s own and others’reputations emerges when productquality is variable and non-contractiblebut not when it is contractible. Theseexperimental results appear to capturesome of the structure of actual ex-changes. Ammar Siamwalla’s (1978)study of marketing structures in Thai-land contrasts the impersonal structureof the wholesale rice market—wherethe quality of the product is readily as-sayed by the buyer—with the personal-ized exchange based on trust in the rawrubber market—where quality is impos-sible to determine at the point of pur-chase.

These experimental results suggestthat trust or reciprocity may depend onthe form of the contract, contractual in-completeness leading to trusting and re-ciprocal behaviors, and conversely.Fehr, Gaechter and Kirchsteiger (1997)present a surprising case of this in theirexperiments with “firms” and “workers.”When they provided more completecontracting of labor effort throughmonitoring and the imposition of fineson workers in cases of verified shirking,worker effort significantly declined.Their interpretation is that explicit in-centives may destroy trust- and reci-procity-based incentives.

Outside the experimenter’s lab, of

course, the degree of contractual in-completeness is not exogenous, and itmay respond to the levels of trust andreciprocity exhibited by the relevantpopulation of traders. For example,lower levels of trust and reciprocitywould plausibly lead those designingcontracts and the relevant enforcementenvironments to be willing to pay morefor more complete contracts. Greif’s(1994) analysis of the divergent culturaland institutional trajectories of theGenovese and Maghribi traders in thelate medieval Mediterranean provides awell documented historical example.The individualism of the Genovese trad-ers precluded the communitarian en-forcement techniques of the Maghribitraders; but it also provided an impe-tus for the development and perfectionof ultimately more successful thirdparty enforcement of claims by theGenovese.

If levels of trust and reciprocity onthe one hand and contractual incom-pleteness on the other are mutually de-termining one may define an equilib-rium set of norms and contracts. If thenature of the mutual influences are as Ihave suggested, there may be any num-ber of these equilibria, some with highlevels of trust and relatively incompletecontracts (like the Maghribi traders)and others with the converse (like theGenovese). If this vastly oversimplifiedview captures something about the dy-namics of cultural change, we might ex-pect rapid shifts in both norms and con-tracts where exogenous events “tip” thesociety from the basin of attraction ofone norm-contract equilibrium to an-other. Thus, it seems reasonable thatsome of the apparently profound cul-tural changes associated with the exten-sion of markets in previously non-mar-ket systems might be explained by thestructural characteristics in Figure 3along with the increasingly contractual


nature of transactions between peopleand the related incentives to reallocatetime and effort away from human in-vestments which are specific to a par-ticular relationship (trust, ethnic orcommunal capital) and toward generalinvestments (schooling).

Florencia Mallon’s (1983) study ofthe growth of markets—particularly la-bor markets—and the erosion of com-munity institutions in the central high-lands of Peru during the early twentiethcentury suggests that some of the abovemechanisms may have been at work.Central to the institutions of local soli-darity among residents was the practiceof contributing labor to road building,irrigation, and other communal proj-ects: “Community membership itself,and access to village resources wasdefined in terms of a quota of labortime that households owed to the com-munity as a whole.” With the extensionof labor markets, many found employ-ment in distant mines for extended pe-riods of time, eventually converting thelabor dues they owed to the communityto cash payments collected and senthome by migrants associations in themining towns. But “migration, by com-modifying relationships and separatingthem out from the intricately wovenfabric of local life, was changing thevery context within which communitycould be defined”(Mallon 1983, pp.264–65).

Traditional institutions were furtherundermined by the sale of commonlands (or charging fees for the use ofthe common lands) and the use of theproceeds to build schools and roads. In-creased access of the richer peasants todistant markets for their produce freedthem of dependence on the locality.The obligation to provide communal la-bor—or even money payments in theirstead—thus became unenforceable, andthe practice declined. The institutions

which had directed community mem-bers’ efforts and imagination towardcommon projects, and the dense net-work of social relationships sustainingthis gave way to investments—schoolingand transportation—whose returns wererelatively independent of the commu-nity social fabric, and contributed littleto it.25

The ethnographic literature on theenvironmental degradation of localcommons provides numerous examplesof similar processes (Jean-Marie Balandand Platteau 1995).

7. Markets, Firms, and Tasks

It is only our Western societies that quite re-cently turned man into an economic animal.But we are not yet all animals of the samespecies. Marcel Mauss ([1925]1967, p. 74)

Learning by doing is a ubiquitousform of personal development; it ap-plies to preferences no less than toskills. The activities we engage in andthe tasks they present to us are not fullydetermined by technology; they dependas well on economic institutions. Thuseconomic institutions may shape prefer-ences by influencing the tasks we per-form.

We know from the experiments ofSherif (1937), Breer and Locke (1965),and others that task performance affectsvalues. Relatedly, a substantial ethno-graphic literature suggests that differ-ing modes of livelihood are associatedwith differing general attitudes and val-ues. Edgerton’s (1971) cross-culturalcomparisons, for example, revealed alarge and statistically significant rela-tionship between the predominance ofpastoral as opposed to farming liveli-hoods and the general cultural valuation

25 An analogous process, occurring in Botswana,is described by T. S. Zufferey (1986). See alsoRaul and Luis Garcia-Barrios (1990).


of independence.26 A plausible mecha-nism for both the Edgerton and theBreer and Locke findings is that strate-gies found successful in coping with thetasks defined by one sphere of life aregeneralized to other realms of life. Be-cause markets and other economic insti-tutions affect the kinds of tasks we con-front and structure rewards andpenalties to various behaviors, we maypresume that they affect learning.

What, then do market tasks teach?Lane (1991, p. 11) reasons that the be-lief that one is effective in influencinghis or her fate (called self-attribution):

is learned from experiences of acting and see-ing the world respond, contingent responses.[Because] a transaction . . . requires mutuallycontingent responses . . . an economy basedon transactions teaches self-attribution.

I know of no evidence for or againstthis plausible conjecture; but if true, theeffects on self-attribution may well de-pend on the structure of the relevantmarkets.

Consider, first, the following counterintuitive example. Market interactionsare particularly likely to contribute tothe sense of personal efficacy underconditions which allow what I term con-sumer sovereignty with teeth—thatwhich occurs when the consumer’s pur-chase confers a rent on the seller be-cause price exceeds marginal cost. Inthis situation the consumer whoswitches to another supplier imposes acost (the loss of the rent) on the seller(Gintis 1989). In this sense monopo-listically competitive markets may pro-

vide at least as fertile a ground for thedevelopment of a sense of personal effi-cacy as do perfectly competitive mar-kets: the latter will exhibit a largernumber of potential suppliers, while theformer will exhibit the stronger versionof consumer sovereignty (because theconsumer confers a rent on the sellerwhenever p > mc) albeit vis à vis amore limited array of suppliers.

If Lane is right that markets teachself-attribution or personal efficacy, theextent to which this is true appears todepend on one’s success in market ac-tivities: income predicts self-attributionbetter than other demographic variablesincluding level of schooling; whites aremore self-attributing than AfricanAmericans; men are more self-attribut-ing than women; and self-attributionrises with age until leveling off in mid-dle age.27

What the market teaches depends notmerely on the degree of success asmeasured by income, but also on thestructural location in a market situation.That the inability to find suitable em-ployment may undermine ones’s senseof efficacy is unsurprising, but having ajob can do the same. The relevant fea-ture of the labor market is that it re-quires employees to relinquish (sub-stantial, but not unlimited) authority

26 Independent minded people may becomeherders rather than farmers, of course, but Edger-ton’s results are robust even when the pastoral-ism/farming measure is not the actual means oflivelihood (e.g., livestock ownership) but ratherthe geographical suitability of the relevant localefor each of these two pursuits. The relevant corre-lations are only slightly diminished when the un-derlying, and presumably exogenous, measure isused.

27 Gerald Gurin and Patricia Gurin (1976);James Birren, Walter Cunningham, and KoichiYamamoto (1983). In the former study the normal-ized regression coefficient of income in a multipleregression predicting a measure of personal effi-cacy is twice as large as that for education andthree times as large as that for race (p. 137). Lon-gitudinal evidence suggests that internality (orself-attribution) and success are mutually deter-mining; while an internal locus of control contrib-utes to success, the reverse is also true (Andrisaniand Nestel 1976). The strong relationship betweenincome and self-attribution may thus arise becausethe more self-attributing people are also more suc-cessful (rather than the other way around); butthis reasoning obviously does not apply to the sub-stantial correlations with exogenous determinantsof economic success such as race, gender, and age.


over their actions to the employer (Her-bert Simon 1951). The employer’sauthority can be effectively wieldedbecause the labor market does notclear, and the employee is thus not in-different to having the job or losing it(Bowles and Gintis 1992, 1993). Ofcourse employees differ greatly in thedegree to which they are subjected tohierarchical authority; and these differ-ences appear to have psychological con-sequences.

Over a period of three decades Kohnand his collaborators have studied therelationship between one’s position inan occupational hierarchy and the indi-vidual’s valuation of self-direction andindependence in their children, intel-lectual flexibility, and personal self-di-rectedness, concluding that “the experi-ence of occupational self direction has aprofound effect on people’s values, ori-entation, and cognitive functioning”(Kohn et al. (1990), p. 967; see alsoKohn 1969, 1990). His collaborativestudy of Japan, the U.S. and Poland(1990) based on sample surveys of maleemployees (from the 1960s and 1970s)yielded cross culturally consistent find-ings: people who exercise self-directionon the job, also value self-directionmore in other realms of their life (in-cluding child-rearing and leisure activi-ties) and are less likely to exhibit thenexus of traits termed the authoritarianpersonality. (See Kohn and Schooler1983, p. 142.)

These results do not arise becauseself-directed people select (or are se-lected into) jobs where occupationalself-direction is substantial. In a seriesof related studies using longitudinaldata, Kohn and his colleagues use twostage least squares estimation to ad-dress the question of reciprocalcausation (personality dimensions ascauses of job position); effects in bothdirections are found, but the job to per-

sonality causal effects are robust.28

Compared to direct measures of occu-pational self-direction, covarying influ-ences such as income, race, ethnicity,family structure, religion, and educationwere considerably less robust and con-sistent predictors of personality andtastes.

Kohn and his co-authors reason that“social structure affects individual psy-chological functioning mainly by affect-ing the conditions of people’s ownlives.” Summarizing his earlier work onchild rearing, Kohn (1969, p. 189)wrote:

Self direction, in short, requires opportuni-ties and experiences that are much moreavailable to people who are more favorablysituated in the hierarchical order of society;conformity is the natural consequence of in-adequate opportunity to be self-directed.

But why should work experiences affectchild-rearing values and leisure timepreferences? Kohn concludes that:

The simple explanation that accounts for vir-tually all that is known about the effects ofjob on personality . . . is that the processesare direct: learning from the job and extend-ing those lessons to off-the job realities.29

Thus, just as the wide range of choicesand contingent reinforcement charac-teristic of consumer goods markets may

28 See Kohn and Carrie Schoenbach (1983). Jey-lan Mortimer, Jon Lorence, and Donald Kumka(1986, p. 113) use similar methods to address theproblem of endogeneity of occupational selection,and report a substantial causal effect of occupa-tionally determined work autonomy on the senseof self-confidence.

29 Kohn (1990, p. 59). Gabriel Almond and Sid-ney Verba (1963, pp. 180ff, 364ff) provide furtherevidence that work experiences are associated withgeneralized subjective orientations. Across all oc-cupational types in five different countries, thosewho were consulted on the job scored significantlyhigher on a measure of subjective civic compe-tence measuring the sense of personal efficacy indealing with local and national government bodies.Differences (between those consulted and thosenot) in subjective competence scores within broadjob types were larger than the differences betweenjob types.


promote personal efficacy, the surrenderof authority to employers which charac-terizes the labor market appears to sup-port far-reaching psychological effects,some of which undermine the sense ofbeing in control of one’s life.

Unlike Kohn’s studies, most researchon the relationship between job struc-tures and personality do not adequatelyaddress the problem of mutual causa-tion mentioned above. Robert Karasek(1978) however, was able to studythe behavioral effects of exogenouschanges in job structure (includingboth expert and self-reports of job char-acteristics) using panel data on theSwedish labor force over the years1968–1974, a period of considerable ex-perimentation with job redesign. Hefound that;

workers whose jobs had become more passivealso became passive in their leisure and po-litical participation and workers with moreactive jobs became more active. These find-ings were significant in eight out of nine sub-populations controlled for education and fam-ily class background. Karasek (1990, pp.53–54)

The effect of economic institutionson task performance and hence on per-sonality may go beyond those stressedby Kohn and Lane. The seemingly de-sirable attribute of markets stressed bySchultze above—that they make few de-mands on our ethical reasoning—mayhave a negative counterpart in a re-duced salience of moral concerns or ca-pacity for moral reasoning. A recentpublic goods experiment suggests thatthese market effects may be important(Norman Frohlich and Joe Oppenheimer1995). Subjects played five-person pub-lic goods games under two conditions:one group played the standard contribu-tion game and the other played a modi-fied game in which a randomized as-signment of payoffs similar to theRawlsian veil of ignorance made it opti-

mal to contribute the maximal amountto the public good. Half of the subjects(in each treatment) were allowed to en-gage in discussion prior to each play (ofcourse the discussion should have hadno effect on the outcome of the stan-dard game, as the dominant strategy isto contribute nothing). After eightrounds of play another seven roundswere conducted, this time with thesame groups but with all playing thestandard game. Among those who hadbeen permitted discussion, those whohad experienced the incentive compat-ible modified game contributed signifi-cantly less in the final seven roundsthan those whose only experience wasthe standard game, and (in subsequentquestionnaires) revealed that their be-havior was less guided by considerationsof fairness.

The authors’ explanation of this strik-ing finding is that the incentive compat-ible mechanism rewarded those con-tributing to the public good, thus mak-ing self interest a good guide to action,while those experiencing the standardgame succeeded only to the extentthat they evoked considerations of fair-ness as a distinct motive. They con-clude

The failure of the . . . (incentive compatible)mechanism to confront subjects with an ethi-cal dilemma appears to lead to little or nolearning in ethical behavior in the subsequentperiod. . . . It is an institution, like other in-centive compatible devices, which can gener-ate near optimal outcomes. . . . Howeverfrom an ethical point of view it is not onlyunsuccessful as pertains to subsequent behav-ior; it appears to be actually pernicious. It un-dermines ethical reasoning and ethically mo-tivated behavior. (Frohlich and Oppenheimer1995, p. 44)

Thus far I have considered the directeffects of markets and other economicinstitutions on the evolution of prefer-ences. But there are indirect effects aswell.


8. Markets and the Process of CulturalTransmission

. . . the modern and the traditional conscious-ness of the [early 19th century] French peas-ant contended for mastery [in]. . .the formof an incessant struggle between the school-masters and the priests.

Karl Marx ([1852]1963, p. 125)

Here I consider the influence of eco-nomic institutions on the structure ofsocial interactions which make up theprocess of cultural transmission, that ison child-rearing practices, childhoodand adolescent socialization, and theavailability of and sometimes compul-sory exposure to entirely new culturalmodels such as teachers and media fig-ures. Some of these effects are the in-tentional result of people’s attempt toacquire and to teach their childrenthose traits required for adequate func-tioning in the social system; other ef-fects are entirely unintended.

One avenue for the effects of eco-nomic organization on cultural trans-mission, the existence of a connectionbetween forms of livelihood and pat-terns of child rearing, has been widelydocumented. Herbert Barry, IrvinChild, and Margaret Bacon (1959) cate-gorized 79 mostly non-literate societiesaccording to the prevalent form of live-lihood (animal husbandry, agricultural,hunting, and fishing) and the relatedease of food storage or other forms ofwealth accumulation, the latter being awell documented correlate of dimen-sions of social structure such as stratifi-cation. They combined these with evi-dence on the dominant forms of childrearing including obedience training,self-reliance, independence, and re-sponsibility. They found large differ-ences in the recorded child-rearingpractices, concluding: “knowledge ofthe economy alone would enable one topredict with considerable accuracy

whether a society’s socialization pres-sures were primarily toward complianceor assertion.”30 Other studies haveconfirmed consistent relationships be-tween these group differences in child-rearing practices and group differencesin various measures of psychologicalfunctioning (Witkin and Berry, 1975).For example, hunter gatherer societiesstress in their child rearing (and achievein their adults) greater independence,while more stratified agricultural socie-ties stress (and achieve) greater obedi-ence.

These results suggest economic struc-tural effects on child rearing andthereby on personality, but do not shedlight on the effects of modern economicinstitutions; indeed markets played alimited role in most of the societiesstudied by Barry, Bacon, and Child andWitkin and Berry. The expansion ofmarkets may have had its largest impactin rendering inadequate the previouslydominant family-based and heterogene-ous forms of socialization studied bythese authors. Ernest Gellner’s (1983)account of the rise of nationalism isbased on the transformation of sociali-zation required by the spatial extensionof the division of labor made possibleby markets:

In the closed local communities of the agrar-ian or tribal worlds, when it came to commu-nication, context, tone, gesture, personalityand situation were everything . . . Among in-timates of a close community, explicitnesswould have been pedantic and offensive [p.33] . . . [but] the requirements of a modern

30 The statistical relationships observed werenot explainable by the covariation of child-rearingpractices and type of livelihood with other mea-sures of social structure such as unilinearity of de-scent, extent of polygyny, levels of participation ofwomen in the predominant subsistence activity,size of population units, and the like. A society-level rather than individual approach has beenadopted in much of the cross cultural literature onchild rearing. See the work of Beatrice and JohnWhiting (Whiting 1963; J. and B. Whiting 1975).


economy obliges [us] to be able to communi-cate contextlessly and with precision with allcomers in face to face ephemeral contacts. (p.140)

This requires

sustained frequent and precise communica-tion between strangers involving a sharing ofexplicit meaning, transmitted in a standardidiom and in writing when required. For anumber of converging reasons this societymust be exo-educational: each individual istrained by specialists, not just by his own lo-cal group, if indeed he has one. (p. 34)

As a result there emerged “a schooltransmitted culture not a folk transmit-ted one” (p. 36) in which children were“handed over by their kin groups to aneducational machine” (p. 37). Universalschooling may be represented as a par-ticular assignment of cultural models tochildren, one unprecedented in its di-vorce from family and degree of centrali-zation.31 As a result the cultural trans-mission processes became markedlymore conformist as cultural models wereselected from (or by) dominant groupsand a society-wide socialization systemintruded into what was once an entirelylocal learning process.

We know strikingly little about thecultural impact of these historicallynovel forms of socialization; most stud-ies of the impact of schooling and itsrelationship to the economy havestressed the contribution of schoolingto cognitive functioning, not to valuesor personality. But the evidence thatpersonality effects of schooling are im-portant is substantial, if indirect. Thesubstantial and apparently causal rela-tionship between years of schooling at-tained by an individual and subsequentlabor market earnings presents a puz-

zle, for available data suggests that alarge part of the schooling-earnings re-lationship is not mediated by the effectof schooling on the level of cognitivefunctioning. Schools make peoplesmarter, and richer, but the latter ef-fect—at least in the U.S.—is surpris-ingly independent of the former. Therelevant evidence is this: the estimatedeffect of schooling on earnings is onlymodestly reduced if the individual levelof cognitive skill is econometrically“held constant” by inclusion in an earn-ings function.32

Gintis and I (1997b) suggest thatschooling may raise earnings through itscontribution to the acquisition of suchpersonality traits as a lower rate of timepreference, a lower disutility of effort,or a cooperative relationship to author-ity figures, which are relevant to thework situation but which are not meas-ured on the existing cognitive measures.We motivate this hypothesis using astandard principal agent model of theproblem of labor discipline in an em-ployment relationship characterized byincomplete contracts. If we are right,the structure of schooling would con-tribute to preparation for adult roles ina manner not dissimilar to that sug-gested by the Bacon, Child, and Barrystudy. But do schools produce thesenon-cognitive employment relatedtraits?

To the best of my knowledge onlyone study has attempted to provide ananswer; it does not provide a satisfac-tory basis for generalization, but it isnonetheless worth reviewing. The strat-egy of the study was to see if schoolsrewarded (and thus inferentially fos-tered the development of) people withthe same personality traits that are val-ued by employers. In parallel investiga-

31 Robert Dreeben’s (1968) book on the social-ization tasks of schooling develops a similar argu-ment based on “the liberating effect” of “the sepa-ration of the workplace from the household” (p.129). See also Gintis (1971) and Bowles and Gintis(1997b).

32 Gintis (1971) first demonstrated this. Bowlesand Gintis (1997b) reviews the large number ofrelevant estimates over a 40 year period.


tions in distinct populations conductedduring the early 1970s, Richard Ed-wards (1977) used peer-rated personal-ity measures of employees in bothprivate and public employment topredict supervisor ratings of theseworkers. Peter Meyer (1972) used thesame peer rated personality variables topredict grade point averages of studentsin a high school, controlling forSAT(verbal and math) and IQ. Edwardsfound that employees judged by theirworkplace peers to be “perseverant,”“dependable,” “consistent,” “punctual,”“tactful,” “identifies with work,” and“empathizes” had significantly highersupervisor ratings, while those judgedby their peers to be “creative” and “in-dependent” were ranked poorly by su-pervisors. Meyer found virtually identi-cal results for the high school studentsin his grading study: the exact traitspredicting favorable supervisor ratingsin the Edwards study, predicted goodgrades (holding constant cognitivescores). Teachers and employers in thesesamples reward the same personalitytraits.33

9. Conclusion

Political writers have established it as amaxim, that, in contriving any system of gov-ernment . . . every man ought to be supposedto be a knave and to have no other end, in allhis actions, than his private interest.

David Hume ([1754]1898 p. 117)

Lawgivers make the citizen good by inculcat-ing habits in them, and this is the aim ofevery lawgiver; if he does not succeed in do-ing that, his legislation is a failure. It is in thisthat a good constitution differs from a badone. Aristotle (1962, p. 1103)

Economists have followed Hume,rather than Aristotle, in positing a givenand self-regarding individual as the ap-propriate behavioral foundation for con-siderations of governance and policy.The implicit premise that policies andconstitutions do not affect preferenceshas much to recommend it: the premiseprovides a common if minimal analyti-cal framework applicable to a widerange of issues of public concern, itexpresses a prudent antipathy towardpaternalistic attempts at social engi-neering of the psyche, it modestlyacknowledges how little we know aboutthe effects of economic structure andpolicy on preferences, and it erects abarrier both to ad hoc explanation andto the utopian thinking of those who in-voke the mutability of human disposi-tions in order to sidestep difficult ques-tions of scarcity and social choice.Realism, however, cannot be among thevirtues invoked on behalf of the exoge-nous preferences premise.34 Economicinstitutions, we have seen, may affectpreferences through their direct influ-ences on situational construal, forms ofreward, the evolution of norms, andtask related learning as well as their in-direct effects on the process of culturaltransmission itself.

One hopes that the active researchagenda now being pursued by econo-mists, other social scientists and biolo-gists in this area may soon allow more

33 We would like to know (but do not) if schoolsproduce the traits they reward, and if traits valuedby supervisors are rewarded by enhanced pay. Theunderlying studies are reported and compared inBowles and Gintis (1976).

34 Indeed Hume, immediately following the pas-sage just quoted, muses that it is “strange that amaxim should be true in politics which is false infact.” While in academic settings most economistsstill adhere to the exogenous preferences canonand its “de gustibus non est disputandum”(George Stigler and Becker 1977) implication,many appear aware of its limitations when itcomes to evaluating institutions and policies. ThusBecker (1995, p. 26) refers to “the effects of afree-market system on self-reliance, initiative, andother virtues” and referring to government trans-fers to the poor, claims that “the present systemcorrupts the values transmitted to children.”


confidence in assessing the empiricalmagnitude and generality of these ef-fects.35 The following research priori-ties seem particularly important.

First, we know very little about theprocess of cultural transmission—whoacquires what trait from whom, underwhat conditions, and why. Yet this in-formation is critical to understandinghow economic institutions may impacton preferences. Empirical studies of therelative importance of parents, otherfamily members, friends, teachers, andothers in cultural learning, and the in-terplay of cultural and genetic transmis-sion would be very valuable.

Second, while we have evidence thattraits acquired in one environment arethen generalized to others (recallKohn’s studies of child rearing) we donot know how this takes place or howpersistent the traits may be once the in-itiating environment is withdrawn.

Third, because imitation of prevalenttraits, or enforced conformism may playan important role in the transmission ofcultural traits, comparative studies ofwhole societies may provide insights notavailable in individual-based studies. Anexample may make this clear. Recallthat Edgerton (1971, p. 195) found thatpastoralists valued independent actionmore than farmers. But farmers in apredominantly pastoral tribe valued in-dependence more (almost twice asmuch by his measure) than farmers inpredominantly farming tribes, whilepastoralists in predominantly farmingtribes valued independence consider-ably less than did pastoralists in the pas-toral tribe. Thus it appears that the pre-dominant livelihood in a tribe may havecultural effects beyond the effects ofthe livelihood of the individual. Analysisof individual data within a single cul-

tural group this may miss important ef-fects of economic institutions operatingon group differences. Comparativeanalysis of economic experiments im-plemented in the differing economicenvironments of distinct societies, in-cluding those with premodern economicinstitutions, would be illuminating.

I emphasize empirical studies be-cause the proliferation of relevant theo-retical models in economics has not beenmatched by empirical investigation. Butimportant contributions could be madeby two types of conceptual work.

Fourth, experiments in economics,sociology, and psychology have raisedserious doubts about the behavioral ac-curacy of the minimalist conception ofhomo economicus: the individual actorwith self-regarding and outcome-basedpreferences. Much of the impact of eco-nomic institutions on behavior may oc-cur through the ways that particular in-stitutional settings prompt individualsto draw one or another response—whether self-regarding, spiteful, gener-ous, or other—from their varied behav-ioral repertoires. A concept ofpreferences more adequately groundedin the empirical study of behaviorwould assist in analyzing these pro-cesses.

Finally, an integration of the insightsof the theory of cultural evolution withthose of evolutionary game theoryseems likely to be insightful, especiallyin view of the apparent importance ofconformism in cultural transmission(and hence the needed modification ofthe concept of cultural equilibrium assuggested in Section 3).

Shortcomings of the existing empiri-cal studies and the unsatisfactory “blackbox” nature of extant knowledge ofsocial learning notwithstanding, theweight of both reason and evidencepoint strongly to the endogeneity ofpreferences. If preferences are indeed

35 Avner Ben-Ner and Louis Putterman (1997)is a valuable collection of relevant work by econo-mists.


endogenous in the senses suggestedhere, four implications follow.

First, economics pays a heavy pricefor its self-imposed isolation from theother behavioral sciences. At its sim-plest, the conception underlying con-temporary disciplinary boundaries isone of society marked by an implausibledegree of specialization among institu-tions: families and religious institutionsshape culture, governments govern, andeconomic institutions allocate re-sources. These disciplinary boundarieshave favored the development of paro-chial, incompatible, and inadequatemodels of human behavior in the vari-ous disciplines, ranging from the over-socialized homo sociologicus to the un-dersocialized homo economicus (MarkGranovetter 1985). Recognition of thecultural effects of markets (and othereconomic institutions) may foster amore unified approach to the behavioralsciences, a benefit of which might bethe more successful resolution of out-standing puzzles in economics.36

Second, the effectiveness of policiesand their political viability may dependon the preferences they induce orevoke.37 Hirschman (1985, p. 10) points

out that economists typically assumeotherwise and for this reason propose

to deal with unethical or antisocial behaviorby raising the cost of that behavior ratherthan proclaiming standards and imposing pro-hibitions and sanctions. The reason is prob-ably that they think of citizens as consumerswith unchanging or arbitrarily changing tastesin matters civic as well as commodity-relatedbehavior. . . . A principal purpose of publiclyproclaimed laws and regulations is to stigma-tize antisocial behavior and thereby to influ-ence citizens’ values and behavioral codes.

Frohlich and Oppenheimer’s and Fehrand Gaechter’s experiments above sug-gest that raising the cost of an antisocialbehavior and other incentive compatibledevices may actually do harm. Moreover,the analysis in Section 6 of the evolutionof nice traits suggests that approximatingthe market ideal by perfecting propertyrights may weaken non-market solutionsto problems of social coordination.There is thus a norm-related analogue tothe second best theorem of welfare eco-nomics: where contracts are incomplete(and hence norms may be important inattenuating market failures), moreclosely approximating idealized completecontracting markets may exacerbate theunderlying market failure (by undermin-ing the reproduction of socially valuablenorms such as trust or reciprocity) andresult in a less efficient equilibrium allo-cation. An analogous caution applies togovernmental, family based, or other so-lutions: for example, numerous experi-ments (as we have seen) suggest that“earning” a claim on a resource differs inpsychologically important ways from sim-ply receiving one, and an adequate un-derstanding of public transfers wouldseem to require attention to these ef-fects.

Third, preference endogeneity givesrise to a kind of market failure and sug-gests a reconsideration of some aspectsof normative economics. The influence

36 For example, given the poor empirical show-ing of most theories of wages (Truman Bewley1995) an adequate understanding of wage settinginstitutions—including why employers do not gen-erally charge job fees (H. Lorne Carmichael1985)—would seem to require an account embrac-ing effects of wages on such preferences as thedisutility of labor and perceptions of just treat-ment, along lines suggested by Akerlof’s (1984)analysis of gift exchange and Robert Solow’s(1990) treatment of “labor markets as social insti-tutions,” as well as the work of Fehr and his co-author mentioned above.

37 Romer’s (1996) account of the origins andevolution of the social security system addressesthe ways that income transfer programs shapepreferences; and Frey’s (1997) econometric studyof tax compliance in Switzerland explores the waythat different constitutional arrangements affect apredisposition to tax avoidance. On the impor-tance of considering the impact of environmentalpolicy on environmental preferences see CassSunstein (1993).


of our preferences on others is not evenapproximately captured by contracts:norms of generosity, non-aggression, orpunishment of antisocial behaviors con-fer external benefits for example, whilea taste (or addiction) for smoking con-fers external costs. Because our prefer-ences have non-contractual effects onothers, how we acquire them is a matterof public concern.

Just as the process of natural selec-tion does not generally maximize aver-age fitness, there is no reason to expectthat the process of cultural transmissiondetermining the equilibrium distri-bution of traits in the population willsupport a socially optimal outcome. Thecultural equivalent of a market failurethus results; indeed the long-term per-sistence of socially and even individu-ally disadvantageous norms is hardlyopen to question, extreme forms ofblood revenge representing a particu-larly well documented example (Jon El-ster 1989; Edgerton 1992; Boehm1984). Because states, communities,and markets may influence the processof cultural evolution, any normativeevaluation of the role and scope ofthese institutions must attempt to taketheir cultural effects into account.

Fourth, there thus may be a novelpublic interest in some types of eco-nomic arrangements which are com-monly considered private. Uncoercedexchange among informed adults isoften considered a private realm inwhich there is no public interest the ab-sence of non-contractual effects onthird parties. The philosopher DavidGauthier (1986, pp. 95–96) writes “Theoperation of the market cannot in itselfraise any evaluative issues . . . The pre-sumption of free activity ensures thatno one is subject to any form of com-pulsion or any type of limitation not al-ready affecting her own actions as asolitary individual.” But if preferences

are shaped by markets and other eco-nomic institutions, both evaluative is-sues and a public interest may arise, foran individual’s preferences induce ac-tions imposing non-contractible costsand benefits on others. Thus part of thereasoning which conventionally estab-lishes a public interest in the natureand amount of schooling —the sociali-zation of children is to some extent apublic good—would seemingly apply tothe effects of economic institutions onpreferences as well.

A broader concept of market failureis thus required, one encompassing theeffects of economic policies and institu-tions on preferences and for this reasonmore adequate for the consideration ofan appropriate mix of markets, commu-nities, families, and states in economicgovernance.38 Such a new welfare eco-nomics would of course have to con-front the longstanding liberal philo-sophical reluctance to privilege someends over others; that is, it would haveto address the problem that Hobbes’mushroom fiction ellides.

REFERENCES

AKERLOF, GEORGE A. “A Theory of Social Cus-tom, of Which Unemployment May Be OneConsequence,” in An economic theorist’s bookof tales: Essays that entertain the consequencesof new assumptions in economic theory.GEORGE A. AKERLOF. Cambridge: CambridgeU. Press, 1984, pp. 69–99.

ALMOND, GABRIEL A. AND VERBA, SIDNEY. Thecivic culture: Political attitudes and democracyin five nations. Princeton: Princeton U. Press,1963.

ANDRE, CATHERINE AND PLATTEAU, JEAN-PHILIPPE. “Land Relations Under UnbearableStress: Rwanda Caught in the Malthusian Trap,”J. Econ. Behav. Organ., 1997, 34(1), pp. 1–55.

38 An example of the reasoning I am recom-mending is Michael Taylor’s (1987) suggestionthat the kinds of opportunism which the Hobbe-sian state is said to curb might be the consequenceof living under a centralized authority, or moresuccinctly that the Hobbesian state produces Hob-besian man (and then more or less successfullycurbs him). Analogous reasoning may apply tomarkets and homo economicus.


ANDRISANI, PAUL J. AND NESTEL, GILBERT. “In-ternal-External Control as Contributor andOutcome of Work Experience,” J. Appl. Psych.,1976, 61(2), pp. 156–65.

ARISTOTLE. Nicomachean ethics. Indianapolis:Bobbs-Merrill, 1962.

ARROW, KENNETH J. “Political and EconomicEvaluation of Social Effects and Externalities,”in MICHAL D. INTRILIGATOR 1971, pp. 3–24.

––––––. “Gifts and Exchanges,” Philosophy andPublic Affairs, 1972, 1(4), pp. 343–62.

ASCH, SOLOMON E. Social psychology. Engle-wood Cliffs, NJ: Prentice-Hall, Inc., 1952.

AXELROD, ROBERT AND HAMILTON, WILLIAM D.“The Evolution of Cooperation,” Science, Mar.1981, 211(27), pp. 1390–96.

BALAND, JEAN-MARIE AND PLATTEAU, JEAN-PHILIPPE. Halting degradation of natural re-sources—Is there a role for rural communities?Oxford: Clarendon Press, 1995.

BARRY, HERBERT, III; CHILD, IRVIN L. AND BA-CON, MARGARET K. “Relation of Child Trainingto Subsistence Economy,” American Anthro-pologist, Feb. 1959, 61(1), pp. 51–63.

BECKER, GARY S. “The Best Reason to Get Peopleoff the Dole,” Business Week, May 1, 1995, p.26.

––––––. Accounting for tastes. Cambridge, MA:Harvard U. Press, 1996.

BELL, DANIEL. The cultural contradictions ofcapitalism. New York: Basic Books, 1976.

BEN-NER, AVNER AND PUTTERMAN, LOUIS, eds.Economics, values, and organizations. Cam-bridge: Cambridge U. Press, 1997.

BEN-PORATH, YORAM. “The F-Connection: Fami-lies, Friends, and Firms and the Organization ofExchange,” Population Devel. Rev., Mar. 1980,6(1), pp. 1–30.

BEWLEY, TRUMAN F. “A Depressed Labor Marketas Explained by Participants,” Amer. Econ.Rev., 1995, 85(2), pp. 250–54.

BIRREN, JAMES E.; CUNNINGHAM, WALTER R.AND YAMAMOTO, KOICHI. “Psychology of AdultDevelopment and Aging,” Annual Review ofPsychology, 1983, 34, pp. 543–75.

BISIN, ALBERTO AND VERDIER, THIERRY. “TheEconomics of Cultural Transmission and theDynamics of Preferences.” Unpublished paper,Massachusetts Institute of Technology andCERAS, Mar. 1996.

BLAU, PETER M. Exchange and power in sociallife. New York: John Wiley and Sons, 1964.

BLISS, CHRISTOPHER J. “Review of R. M. Tit-muss, The gift relationship: From human bloodto social policy,” J. Public Econ., 1972, 1, pp.162–65.

BLOOM, BENJAMIN S. Stability and change in hu-man characteristics. New York: John Wiley &Sons, Inc., 1964.

BOEHM, CHRISTOPHER. Blood revenge: Theanthropology of feuding in Montenegro andother tribal societies. U. of Pennsylvania Press,1984.

––––––. “Egalitarian Behavior and Reverse Domi-

nance Hierarchy,” Current Anthropology, June1993, 34(3), pp. 227–54.

BOHANNAN, PAUL. “The Impact of Money on anAfrican Subsistence Economy,” J. Econ.Hist.,Dec. 1959, 19(4), pp. 491–503.

BOWLES, SAMUEL. “Markets as Cultural Institu-tions: Equilibrium Norms in CompetitiveEconomies.” Working paper 1996–5. Dept. ofEconomics, U. of Massachusetts, June 25, 1996.

BOWLES, SAMUEL AND GINTIS, HERBERT. Schooling in capitalist America: Educational re-form and the contradictions of economic life .New York: Basic Books, Inc., 1976.

––––––. Democracy and capitalism: Property,community, and the contradictions of modernsocial thought. London: Routledge and KeganPaul, 1986.

———. “Power and Wealth in a CompetitiveCapitalist Economy,” Philosophy & Public Af-fairs, Fall 1992, 21(4), pp. 324–53.

———. “The Revenge of Homo Economicus:Contested Exchange and the Revival of PoliticalEconomy,” J. Econ. Perspectives, Winter 1993,7(1), pp. 83–102.

———. “Optimal Parochialism: The dynamics oftrust and exclusion in communities.” Unpub-lished paper, U. of Massachusetts, 1997a.

———. “Labor Discipline and the Returns toSchooling.” Unpublished paper, Department ofEconomics, U. of Massachusetts, July 1997b.

———.“The Moral Economy of Communities:Structured Populations and the Evolution ofPro-Social Norms,” Evolution and Human Be-havior, Jan. 1998, 19(1).

BOYD, ROBERT AND RICHERSON, PETER J. Cul-ture and the evolutionary process. Chicago andLondon: U. of Chicago Press, 1985.

———. “Group Selection among Alternative Evo-lutionarily Stable Strategies,” Journal of Theore-tical Biology, 1990, 145, pp. 331–42.

———. “Rationality, Imitation, and Tradition,” inNonlinear dynamics and evolutionary econom-ics. Eds.: RICHARD H. DAY AND PING CHEN.New York: Oxford U. Press, 1993, pp. 131–51.

BOYER, PAUL AND NISSENBAUM, STEPHEN. Salem possessed: The social origins of witch-craft. Cambridge, MA: Harvard U. Press, 1974.

BREER, PAUL E. AND LOCKE, EDWIN A. Taskexperience as a source of attitudes. Homewood,IL: The Dorsey Press, 1965.

BUCHANAN, JAMES M. The limits of liberty. Chi-cago: U. of Chicago Press, 1975.

BURRIDGE, KENELM. New heaven, new earth: Astudy of millenarian activities. New York:Schocken Books, 1969.

CAMERER, COLIN F. AND KUNREUTHER,HOWARD. “Decision Processes for Low Prob-ability Events: Policy Implications,” J. PolicyAnal. Manage., 1989, 8(4), pp. 565–92.

CAMERER, COLIN AND THALER, RICHARD H.“Anomalies: Ultimatums, Dictators and Man-ners,” J. Econ. Perspectives, Spring 1995, 9(2),pp. 209–19.

CAPORAEL, LINNDA R. ET AL. “Selfishness Exam-


ined: Cooperation in the Absence of EgoisticIncentives,” Behavioral and Brain Sciences,1989, 12, pp. 683–99.

CARMICHAEL, H. LORNE. “Can UnemploymentBe Involuntary? Comment,” Amer. Econ. Rev.,Dec. 1985, 75(5), pp. 1213–14.

CAVALLI-SFORZA, LUIGI L. AND FELDMAN, M.W. Cultural transmission and evolution: Aquantitative approach. Princeton: Princeton U.Press, 1981.

CAVALLI-SFORZA, L. L. ET. AL. “Theory and Ob-servation in Cultural Transmission,” Science,Oct. 1, 1982, 218, pp. 19–27.

CHEN, KUANG-HO; CAVALLI-SFORZA, L. L. ANDFELDMAN, M. W. “A Study of Cultural Trans-mission in Taiwan,” Human Ecology, Sept.1982, 10(3), pp. 365–82.

CONLISK, JOHN. “Costly Optimizers versus CheapImitators,” J. Econ. Behav. Organ., Sept. 1980,1(3), pp. 275–93.

DAWES, ROBYN M.; VAN DE KRAGT, ALPHONS J.C. AND ORBELL, JOHN M. “Not Me or Theebut We: The Importance of Group Identity inEliciting Cooperation in Dilemma Situations:Experimental Manipulations,” Acta Psycholo-gica, 1988, 68, pp. 83–97.

DECI, EDWARD L. Intrinsic motivation. NewYork: Plenum Press, 1975.

DECI, EDWARD L. AND RYAN, RICHARD M. In-trinsic motivation and self-determination in hu-man behavior. New York and London: PlenumPress, 1985.

DREEBEN, ROBERT. On what is learned inschool. Reading: Addison-Wesley, 1968.

DUESENBERRY, JAMES STEMBLE. Income, savingand the theory of consumer behavior. Cam-bridge, MA: Harvard U. Press, 1949.

DURHAM, WILLIAM H. Coevolution: genes, cul-ture, and human diversity. Stanford: StanfordU. Press, 1991.

DURKHEIM, EMILE. The division of labor in soci-ety. New York: Free Press, [1933] 1964.

ECKERT, PENELOPE. “Clothing and Geography ina Suburban High School,” in ResearchingAmerican culture. Ed.: CONRAD PHILLIP KOT-TAK. Ann Arbor: The U. of Michigan Press,1982, pp. 139–44.

———. “Adolescent Social Structure and theSpread of Linguistic Change,” Language in So-ciety, June 1988, 17(2), pp. 183–207.

EDGERTON, ROBERT B. The individual in cul-tural adaptation. Berkeley: U. of CaliforniaPress, 1971.

———. Sick societies: Challenging the myth ofprimitive harmony. New York: Free Press,1992.

EDWARDS, RICHARD C. “Personal Traits and ‘Suc-cess’ in Schooling and Work,” Educational andPsychological Measurement, 1977, 37, pp. 125–38.

ELSTER, JON. “Social Norms and Economic The-ory,” J. Econ. Perspectives, Fall 1989, 3(4), pp.99–117.

ENSMINGER, JEAN. Making a market: The institu-

tional transformation of an African society.Cambridge: Cambridge U. Press, 1992.

EPSTEIN, T. SCARLETT. Economic developmentand social change in south India. Manchester:U. Press, 1962.

FEHR, ERNST; GAECHTER, SIMON AND KIRCH-STEIGER, GEORG. “Reciprocity as a ContractEnforcement Device. Experimental Evidence,”Econometrica, July 1997, 65(4) pp. 833–60.

FEHR, ERNST ET AL. “When Social Norms Over-power Competition—Social Exchange in Ex-perimental Labor Markets,” J. Labor Econ,forthcoming.

FEHR, ERNST AND TYRAN, JEAN-ROBERT. “Insti-tutions and Reciprocal Fairness,” Nordic J. Pol.Economy, 1996, 23(2), pp. 133–44.

FELDMAN, MARCUS W.; AOKI, KENICHI ANDKUMM, JOCHEN. “Individual Versus SocialLearning: Evolutionary Analysis in a Fluctuat-ing Environment.” Working Paper No. 96–05–030, Santa Fe Institute, Mar. 14, 1996.

FELDMAN, MARCUS W. AND LALAND, KEVIN N.“Gene-Culture Coevolutionary Theory.” Work-ing Paper No. 96–05–033, Santa Fe Institute,Mar. 14, 1996.

FIELD, MARGARET J. Search for security: Anethno-psychiatric study of rural Ghana. Evan-ston, IL: Northwestern U. Press, 1960.

FIRTH, RAYMOND W. “Work and Wealth of Primi-tive Communities,” in Human types: An intro-duction to social anthropology. By RAYMONDW. FIRTH. New York: Mentor Books, 1958, pp.62–81.

FISKE, ALAN PAGE. Structures of social life: Thefour elementary forms of human relations . NewYork: Free Press, 1991.

———. “The Four Elementary Forms of Sociality:Framework for a Unified Theory of Social Rela-tions,” Psychological Review, Oct. 1992, 99(4),pp. 689–723.

FLORA, JANE; MACCOBY, NATHAN AND FAR-QUHAR, JOHN W. “Communication Campaignsto Prevent Cardiovascular Disease: The Stan-ford Community Studies,” in Public communi-cation campaigns. 2nd ed. Eds.: RONALD E.RICE AND CHARLES K. ATKIN. Newbury Park:Sage, 1989, pp. 233–52.

FREY, BRUNO S. “A Constitution for KnavesCrowds Out Civic Virtue,” Econ. J., July 1997,107(443), pp. 1043–53.

FREY, BRUNO S. AND OBERHOLZER-GEE, FELIX.“The Cost of Price Incentives: An EmpiricalAnalysis of Motivation Crowding-out,” Amer.Econ. Rev., 1997, 87(4), pp. 746–55.

FROLICH, NORMAN AND OPPENHEIMER, JOE A.“The Incompatibility of Incentive CompatibleDevices and Ethical Behavior: Some Experi-mental Results and Insights,” Public ChoiceStudies, 1995, 25, pp. 24–51.

FROMM, ERICH AND MACCOBY, MICHAEL. Socialcharacter in a Mexican village: A sociopsycho-analytic study. Englewood Cliffs, NJ: Prentice-Hall, 1970.

FUNDENBERG, DREW AND MASKIN, ERIC. “The


Folk Theorem in Repeated Games with Dis-counting or with Incomplete Information,”Econometrica, May 1986, 54(3), pp. 533–54.

GAECHTER, SIMON AND FEHR, ERNST. “Collec-tive Action as a Partial Social Exchange.” Un-published paper, Institute for Empirical Re-search in Economics, U. of Zurich, May 1997.

GARCIA-BARRIOS, RAUL AND GARCIA-BARRIOS,LUIS. “Environmental and Technological Deg-radation in Peasant Agriculture: A Consequenceof Development in Mexico,” World Devel., Nov.1990, 18(11), pp. 1569–85.

GAUTHIER, DAVID P. Morals by agreement. Ox-ford: Clarendon Press, 1986.

GELLNER, ERNEST. Nations and nationalism. Ith-aca: Cornell U. Press, 1983.

GINTIS, HERBERT. “Education, Technology, andthe Characteristics of Worker Productivity,”Amer. Econ. Rev., May 1971, 61(2), pp. 266–79.

———. “A Radical Analysis of Welfare Economicsand Individual Development,” Quart. J. Econ.,Nov 1972, 86(4), pp. 572–99.

———. “The Power to Switch: On the PoliticalEconomy of Consumer Sovereignty,” in Uncon-ventional wisdom: Essays on economics inhonor of John Kenneth Galbraith. Eds.: SA-MUEL BOWLES, RICHARD EDWARDS, ANDWILLIAM G. SHEPHERD. Boston: HoughtonMifflin, 1989.

GRAFEN, ALAN. “The Hawk-Dove Game PlayedBetween Relatives,” Animal Behavior, 1979,27(3), pp. 905–07.

GRANOVETTER, MARK. “Economic Action and So-cial Structure: The Problem of Embeddedness,”Amer. J. Sociology, Nov. 1985, 91(3), pp. 481–510.

GREIF, AVNER. “Cultural Beliefs and the Organi-zation of Society: A Historical and TheoreticalReflection on Collectivist and Individualist So-cieties,” J. Polit. Econ., Oct. 1994, 102(5), pp.912–50.

GURIN, GERALD AND GURIN, PATRICIA. “Per-sonal Efficacy and the Ideology of IndividualResponsibility,” in Economic means for humanneeds. Ed.: BURKHARD STRUMPEL. Ann Arbor:Institute for Social Research, 1976. pp. 131–57.

GÜTH, WERNER; SCHMITTBERGER, ROLF ANDSCHWARZE, BERND. “An Experimental Analysisof Ultimatum Bargaining,” J. Econ. Behav. Or-gan., Dec. 1982, 3(4), pp. 367–88.

HAMILTON, W. D. “Innate Social Aptitudes ofMan: an Approach from Evolutionary Genet-ics,” in Biosocial anthropology. Ed.: ROBINFOX. New York: John Wiley, 1975, pp. 115–32.

HARSANYI, JOHN C. “Morality and the Theory ofRational Behavior,” in Utilitarianism and be-yond. Eds.: AMARTYA SEN AND BERNARD WIL-LIAMS. Cambridge: Cambridge U. Press, 1982,pp. 39–62.

HAYEK, FRIEDRICH A. VON. Individualism andeconomic order. Chicago: U. of Chicago Press,1948.

HIRSCHMAN, ALBERT O. The passions and theinterests: Political arguments for capitalism be-

fore its triumph. Princeton: Princeton U. Press,1977.

———. “Rival Interpretations of Market Society:Civilizing, Destructive, or Feeble?” J. Econ.Lit., Dec. 1982, 20(4), pp. 1463–84.

———. “Against Parsimony: Three Easy Ways ofComplicating Some Categories of EconomicDiscourse,” Econ. Philos., Apr. 1985, 1(1), pp.7–21.

HOBBES, THOMAS. De cive or The citizen. NewYork: Appleton-Century-Crofts, Inc., 1949.

HOFFMAN, ELIZABETH ET AL. “Preferences,Property Rights, and Anonymity in BargainingGames,” Games Econ. Behav., Nov. 1994, 7(3),pp. 346–80.

HOLLANDER, HEINZ. “A Social Exchange Ap-proach to Voluntary Cooperation,” Amer. Econ.Rev., Dec. 1990, 80(5), pp. 1157–67.

HOMANS, GEORGE C. “Social Behavior as Ex-change,” Amer. J. Sociology, May 1958, 63(6),pp. 597–606.

HUME, DAVID. Essays: Moral, political and liter-ary. London: Longmans, Green and Co., [1754]1898.

INTRILIGATOR, MICHAEL D. Frontiers of quanti-tative economics. Amsterdam: North-HollandPub. Co., 1971.

KACHELMEIER, STEVEN J. AND SHEHATA, MO-HAMED. “Internal Auditing and Voluntary Co-operation in Firms: A Cross-cultural Experi-ment,” Accounting Review, 1997, 72(3), pp.407–31.

KANT, IMMANUEL. The philosophy of Kant: Im-manuel Kant’s moral and political writings. Ed.:CARL J. FRIEDRICH. New York: Modern Li-brary, [1785] 1949.

KARASEK, ROBERT. “Job Socialization: A Longitu-dinal Study of Work, Political and Leisure Ac-tivity.” Working Paper 59, Stockholm: Institutefor Social Research, 1978.

KARASEK, ROBERT AND THEORELL, TORES.Healthy work: Stress, productivity, and the re-construction of working life. New York: BasicBooks, 1990.

KATZ, ELIHU AND LAZARSFELD, PAUL F. Per-sonal influence: The part played by people inthe flow of mass communications. Glencoe, IL:Free Press, 1955.

KOHN, MELVIN L. Class and conformity: A studyin values. Homewood, IL: Dorsey Press, 1969.

———. “Unresolved Issues in the RelationshipBetween Work and Personality,” in The natureof work: Sociological perspectives. Eds.: KAIERIKSON AND STEVEN PETER VALLAS. NewHaven, CT: Yale U. Press, 1990, pp. 36–68.

KOHN, MELVIN L. AND SCHOENBACH, CARRIE.“Class, Stratification, and Psychological Func-tioning,” in MELVIN L. KOHN AND CARMISCHOOLER 1983, pp. 154–89.

KOHN, MELVIN L. AND SCHOOLER, CARMI. Workand personality: An inquiry into the impact ofsocial stratification. Norwood: Ablex Pub.Corp., 1983.

KOHN, MELVIN ET AL. “Position in the Class


Structure and Psychological Functioning in theUnited States, Japan, and Poland,” Amer. J. So-ciology, Jan. 1990, 95(4), pp. 964–1008.

KOLLOCK, PETER. “The Emergence of ExchangeStructures: An Experimental Study of Uncer-tainty, Commitment, and Trust,” Amer. J. Soci-ology, Sept. 1994, 100(2), pp. 313–45.

———. “Transforming Social Dilemmas: GroupIdentity and Cooperation,” in Modeling rationaland moral agents. Ed.: PETER DANIELSON. Ox-ford: Oxford U. Press, 1997.

KRANTON, RACHEL E. “Reciprocal Exchange: ASelf-Sustaining System,” Amer. Econ. Rev.,Sept. 1996, 86(4), pp. 830–51.

KREPS, DAVID M. “Corporate Culture and Eco-nomic Theory,” in Perspectives on positive po-litical economy. Eds.: JAMES E. ALT AND KEN-NETH A. SHEPSLE. Cambridge: Cambridge U.Press, 1990, pp. 90–143.

LABOV, WILLIAM. Sociolinguistic patterns. Phila-delphia: U. of Pennsylvania Press, 1972.

———. “De Facto Segregation of Black and WhiteVernaculars.” Project on Linguistic Change andVariation, Lingiustics Laboratory, U. of Penn-sylvania, 1983.

LAIBSON, DAVID. “A Cue-Theory of Consump-tion.” Unpublished Paper, Harvard U. Depart-ment of Economics, 1996.

LANE, ROBERT E. The market experience. Cam-bridge: Cambridge U. Press, 1991.

LASLETT, PETER. The world we have lost. Lon-don: Methuen, 1965.

LEDYARD, JOHN O. “Public Goods: A Survey ofExperimental Research,” in The handbook ofexperimental economics. Eds.: JOHN H. KAGELAND ALVIN E. ROTH. Princeton, NJ: PrincetonU. Press, 1995, pp. 111–94.

LEIBENSTEIN, HARVEY. “Bandwagon, Snob, andVeblen Effects in the Theory of Consumers’Demand,” Quart. J. Econ., May 1950, 64, pp.183–207.

LEPPER, MARK R. AND GREENE, DAVID. Thehidden costs of reward: New perspectives on thepsychology of human motivation. Hillsdale:Lawrence Erlbaum Associates, 1978.

LEPPER, MARK R.; GREENE, DAVID AND NIS-BETT, RICHARD E. “Undermining Children’sIntrinsic Interest With Extrinsic Reward: ATest of the ‘Overjustification’ Hypothesis,” J.Personality and Social Psychology, 1973, 28(1),pp. 129–37.

LERNER, DANIEL. The passing of traditional soci-ety: Modernizing the Middle East. New York:Free Press, 1958.

LEVINE, ROBERT A. Dreams and deeds: Achieve-ment motivation in Nigeria. Chicago: U. of Chi-cago Press, 1966.

———. Culture, behavior, and personality. Lon-don: Hutchinson & Co., 1973.

LOEWENSTEIN, GEORGE. “Out of Control: Vis-ceral Influences on Behavior,” OrganizationalBehavior and Human Decision Processes, 1996,65, pp. 272–92.

LOEWENSTEIN, GEORGE AND ISSACHAROFF,

SAMUEL. “Source Dependence in the Valuationof Objects,” J. Behavioral Decision Making,1994, 7, pp 157–68

MALINOWSKI, BRONISLAW. Crime in savage soci-ety. London: Routledge & Kegan Paul, 1926.

MALLON, FLORENCIA E. The defense of commu-nity in Peru’s central highlands: Peasant strug-gle and capitalist transition 1860–1940. Prince-ton, NJ: Princeton U. Press, 1983.

MARX, KARL. The eighteenth brumaire of LouisBonaparte. New York: International Publishers,[1852] 1963.

MAUSS, MARCEL. The gift: forms and functions ofexchange in archaic societies. New York: W. W.Norton, [1925] 1967.

MCCAULEY, CLARK; ROZIN, PAUL ANDSCHWARTZ, BARRY. “On the Origin and Natureof Preference and Values.” Unpublished paper,Oct. 1994.

MCCLELLAND, DAVID. The achieving society.Princeton, NJ: Van Nostrand, 1961.

MEYER, PETER J. “Schooling and the Reproduc-tion of the Social Division of Labor.” Un-published honors thesis, Harvard U., Mar.1972.

MORTIMER, JEYLAN T.; LORENCE, JON ANDKUMKA, DONALD S. Work, family, and person-ality: Transition to adulthood. Norwood: AblexPub., 1986.

MULLIGAN, CASEY B. Parental priorities and eco-nomic inequality. Chicago and London: U. ofChicago Press, 1997.

NEWCOMB, THEODORE M. ET AL. Persistence andchange: Bennington College and its students af-ter twenty-five years. New York: John Wiley &Sons, Inc., 1967.

NOWELL-SMITH, PATRICK H. Ethics. London:Penguin, 1954.

OCHS, JACK AND ROTH, ALVIN E. “An Experi-mental Study of Sequential Bargaining,” Amer.Econ. Rev., June 1989, 79(3), pp. 355–84.

OTTO, SARAH P.; CHRISTIANSEN, FREDDY B. ANDFELDMAN, MARCUS W. “Genetic and CulturalInheritance of Continuous Traits.” Working Pa-per No. 0064. Morrison Institute for Populationand Resource Studies, Stanford U., 1995.

OUCHI, WILLIAM G. “Markets, Bureaucracies,and Clans,” Administrative Science Quarterly,Mar. 1980, 25(1), pp. 129–41.

PARSONS, TALCOTT. The structure of social action:A study in social theory with special referenceto a group of recent European writers. 2nd ed.Glencoe, IL: Free Press, 1949.

———. “On the Concept of Political Power,”1967a in TALCOTT PARSONS, ed. 1967, pp. 297–354.

––––––, ed. Sociological theory and modern soci-ety. New York: Free Press, 1967.

POLANYI, KARL. The great transformation: Thepolitical and economic origins of our time. Bea-con Hill: Beacon Press, 1957.

POLLAK, ROBERT A. “Interdependent Prefer-ences,” Amer. Econ. Rev., June 1976, 66(3), pp.309–20.


POTTER, DAVID MORRIS. People of plenty: Eco-nomic abundance and the American character.Chicago: U. of Chicago Press, 1954.

PLOMIN, ROBERT AND DANIELS, DENISE. “WhyAre Children in the Same Family So Differentfrom One Another?” Behavioral and Brain Sci-ences, 1987, 10, pp. 1–60.

RADNER, ROY AND SCHOTTER, ANDREW. “TheSealed Bid Mechanism: An ExperimentalStudy” J. Econ. Theory, June 1989, 48(1), pp.179–220.

ROGERS, EVERETT M. Diffusion of innovations.New York: Free Press, 1962.

ROMER, PAUL. “Preferences, Promises and thePolitics of Entitlement,” in Individual and so-cial responsibility: Child care, education, medi-cal care and long-term care in America. Ed.:VICTOR FUCHS. Chicago: U. of Chicago Press,1996, pp. 195–228.

ROSE-ACKERMAN, SUSAN. “Gifts and Bribes” inEconomics, values and organization. Eds.:AVNER BEN-NER AND LOUIS PUTTERMAN.Cambridge: Cambridge U. Press, 1997.

ROSS, LEE AND NISBETT, RICHARD E. The per-son and the situation. Philadelphia, PA: TempleU. Press, 1991.

ROTH, ALVIN E. “Bargaining Experiments,” inThe handbook of experimental economics.JOHN H. KAGEL AND ALVIN E. ROTH. Prince-ton, NJ: Princeton U. Press, 1995, pp 253–346.

ROTH, ALVIN E. ET AL. “Bargaining and Mar-ket Behavior in Jerusalem, Ljubljana, Pitts-burgh, and Tokyo: An Experimental Study,”Amer. Econ. Rev., Dec. 1991, 81(5), pp. 1068–95.

ROZIN, PAUL. “Family Resemblance in Food andOther Domains: The Family Paradox and theRole of Parental Congruence,” Appetite, Apr.1991, 16(2), pp. 93–102.

ROZIN, PAUL AND NEMEROFF, CAROL. “TheLaws of Sympathetic Magic: a PsychologicalAnalysis of Similarity and Contagion,” in Cul-tural psychology: Essays on comparative humandevelopment. Eds.: JAMES W. STIGLER, RICH-ARD A. SHWEDER AND GILBERT HERDT.Cambridge: Cambridge U. Press, 1990, pp.205–32.

ROZIN, PAUL AND VOLLMECKE, T. A. “FoodLikes and Dislikes,” Ann. Rev. Nutrition, 1986,6, pp. 433–56.

SAHLINS, MARSHALL P. Stone age economics.Chicago: Aldine Pub. Co., 1972.

SCHOTTER, ANDREW; WEISS, AVI AND ZAPATER,INIGO. “Fairness and Survival in Ultimatum andDictatorship Games,” J. Econ. Behav. Organ.,Oct. 1996, 31(1), pp. 37–56.

SCHULTZE, CHARLES L. The public use of privateinterest. Washington, DC: Brookings Institu-tion, 1977.

SCHUMPETER, JOSEPH A.. Capitalism, socialismand democracy. 3rd ed. New York: Harper &Row, 1950.

SEN, AMARTYA. “Rational Fools: A Critique of the

Behavioral Foundations of Economic Theory,”Philosophy and Public Affairs, Summer 1977,6(4), pp. 317–44.

SHAPIRO, CARL. “Premiums for High QualityProducts as Returns to Reputations,” Quart. J.Econ., Nov. 1983, 98(4), pp. 659–79.

SHERIF, MUZAFER. “An Experimental Approachto the Study of Attitudes,” Sociometry, July-Oct. 1937, I, pp. 90–98.

SIAMWALLA, AMMAR. “Farmers and Middlemen:Aspects of Agricultural Marketing in Thailand,”Economic Bulletin for Asia and the Pacific, June1978, 29(1), pp. 38–50.

SIMMEL, GEORG. The philosophy of money. Lon-don and New York: Routledge, 1990.

SIMON, HERBERT A. “A Formal Theory of theEmployment Relationship,” Econometrica, July1951, 19, pp. 293–305.

SOLOW, ROBERT M. The labor market as a socialinstitution. Cambridge, MA: Basil Blackwell,1990.

STARR, CHAUNCEY. “Social Benefits versus. Tech-nological Risk,” Science, Sept. 19, 1969, 165,pp. 1232–38.

STIGLER, GEORGE J. AND BECKER, GARY S. “DeGustibus Non Est Disputandum,” Amer. Econ.Rev., Mar. 1977, 67(2), pp. 76–90.

SUNSTEIN, CASS R. “Endogenous Preferences,Environmental Law,” J. Legal Studies, June1993, 22(2), pp. 217–54.

TAUSSIG, MICHAEL T. The devil and commodityfetishism in South America. Chapel Hill: U. ofNorth Carolina Press, 1980.

TAYLOR, MICHAEL. Anarchy and cooperation.Cambridge: Cambridge U. Press, 1976.

––––––. The possibility of cooperation. Cam-bridge: Cambridge U. Press, 1987.

TITMUSS, RICHARD M. The gift relationship:From human blood to social policy. New York:Pantheon Books, 1971.

TONNIES, FERDINAND. Community & society.New York: Harper & Row, 1963.

TVERSKY, AMOS AND KAHNEMAN, DANIEL. “Ra-tional Choice and the Framing of Decisions,” J.Bus., Oct. 1986, 59(4), pp. S251–78.

UPTON, WILLIAM EDWARD III. “Altruism, Attri-bution, and Intrinsic Motivation in the Recruit-ment of Blood Donors.” Doctoral dissertation,Cornell U., 1973, pp. 1–89.

VEBLEN, THORSTEIN. The theory of the leisureclass. New York: Modern Library, 1934.

WEBER, MAX. Economy and society: An outline ofinterpretive sociology, Vols. I and II. Eds.:GUENTHER ROTH AND CLAUS WITTICH.Berkeley: U. of California Press, [1922]1978.

WHITING, BEATRICE B., ed. Six cultures: Studiesof child rearing. New York: John Wiley andSons, 1963.

WHITING, BEATRICE B. AND WHITING, JOHN W.M. Children of six cultures: A psycho-culturalanalysis. Cambridge, MA: Harvard U. Press,1975.

WILSON, DAVID SLOAN AND SOBER, ELLIOT.


“Reintroducing Group Selection to the HumanBehavioral Sciences,” Behavioral and Brain Sci-ences, 1994, 17, pp. 585–654.

WITKIN, HERMAN A. AND BERRY, JOHN W. “Psy-chological Differentiation in Cross-CulturalPerspective,” J. Cross-Cultural Psychology,Mar. 1975, 6(1), pp. 4–87.

ZAJONC, ROBERT B. “Attitudinal Effects of MereExposure,” J. Personality and Social Psychol-ogy . Monograph Supplement, June 1968, 9(2,Part 2), pp. 1–27.

———. “Feeling and Thinking: Preferences NeedNo Inferences,” American Psychologist, Feb.1980, 35(2), pp. 151–75.

ZELIZER, VIVIANA A. “Payments and Social Ties,”Sociological Forum, Sept. 1996, 11(3), pp.481–95.

ZUFFERY, F. S. “A Study of Local Institutions andResource Management Inquiry in Eastern Cen-tral District.” Research paper no. 88, Land Ten-ure Center, U. of Wisconsin-Madison, Mar.1986.


bowles: endogenous preferences endogenous preferences: the

Documents