r the protestant work ethic: a lay theory with dual intergroup … · 2007. 2. 9. · group...

21
Group Processes & Intergroup Relations 2006 Vol 9(1) 95–115 G P I R The Protestant Work Ethic: A Lay Theory with Dual Intergroup Implications Sheri R. Levy, Tara L. West and Luisa Ramirez State University of New York at Stony Brook Dina M. Karafantis Borough of Manhattan Community College The authors propose that, in the US, the Protestant work ethic (PWE) relates both to social tolerance and intolerance. PWE is proposed to have a surface meaning that relates to social tolerance, and also an associated meaning that relates to intolerance, which is acquired in part through social and cultural experience (e.g. PWE being used as a justifier of inequality). In correlational and experimental studies, PWE was related to greater egalitarianism and desired social closeness to African Americans among younger participants (9- to 12- and 14- to 16-year-olds) relative to older participants (college students). Subsequent experiments directly manipulated college students’ interpretations of PWE, showing that those experimentally led to focus on others’ use of PWE in support of their arguments (associated meaning condition) endorsed egalitarianism to a lesser extent (Study 3) and donated less money to a homeless shelter (Study 4) than did those simply focusing on the definition of PWE (definition condition). In contrast to these findings, the authors showed that social dominance orientation has a unitary relation to social intolerance across the three age groups studied (Study 1). The implications of these findings and future work on the duality of lay theories are discussed. KEYWORDS duality, egalitarianism, implicit theories, justification, lay theories, prejudice, Protestant work ethic, racial attitudes, social dominance orientation I N THE UNITED STATES , a pervasive lay theory is the Protestant work ethic (PWE), which is often captured by sayings such as ‘anyone can pull themselves up by their bootstraps’ or ‘the early bird gets the worm’. Essentially, this is the lay theory that ‘people who work hard succeed’ (e.g. Crandall, 1994; Katz & Hass, 1988; Quinn & Crocker, 1999; Somerman, 1993). The PWE has long been discussed as an ingredient in contemporary US racism toward African Americans at the hands of European Americans; African Americans are seen as not Copyright © 2006 SAGE Publications (London, Thousand Oaks, CA and New Delhi) 9:1; 95–115; DOI: 10.1177/1368430206059874 Author’s note Address correspondence to Sheri R. Levy, Department of Psychology, SUNY Stony Brook, Stony Brook, NY 11794-2500, USA [email: [email protected]]

Upload: others

Post on 20-Oct-2020

0 views

Category:

Documents


0 download

TRANSCRIPT

  • Group Processes &Intergroup Relations

    2006 Vol 9(1) 95–115

    GPIR

    The Protestant WorkEthic: A Lay Theory withDual IntergroupImplications

    Sheri R. Levy, Tara L. West and Luisa RamirezState University of New York at Stony Brook

    Dina M. KarafantisBorough of Manhattan Community College

    The authors propose that, in the US, the Protestant work ethic (PWE) relates both to socialtolerance and intolerance. PWE is proposed to have a surface meaning that relates to socialtolerance, and also an associated meaning that relates to intolerance, which is acquired in partthrough social and cultural experience (e.g. PWE being used as a justifier of inequality). Incorrelational and experimental studies, PWE was related to greater egalitarianism and desiredsocial closeness to African Americans among younger participants (9- to 12- and 14- to16-year-olds) relative to older participants (college students). Subsequent experiments directlymanipulated college students’ interpretations of PWE, showing that those experimentally led tofocus on others’ use of PWE in support of their arguments (associated meaning condition)endorsed egalitarianism to a lesser extent (Study 3) and donated less money to a homelessshelter (Study 4) than did those simply focusing on the definition of PWE (definitioncondition). In contrast to these findings, the authors showed that social dominance orientationhas a unitary relation to social intolerance across the three age groups studied (Study 1). Theimplications of these findings and future work on the duality of lay theories are discussed.

    KEYWORDS duality, egalitarianism, implicit theories, justification, lay theories,prejudice, Protestant work ethic, racial attitudes, social dominance orientation

    IN THE UNITED STATES, a pervasive lay theoryis the Protestant work ethic (PWE), which isoften captured by sayings such as ‘anyone canpull themselves up by their bootstraps’ or ‘theearly bird gets the worm’. Essentially, this is thelay theory that ‘people who work hard succeed’(e.g. Crandall, 1994; Katz & Hass, 1988;Quinn & Crocker, 1999; Somerman, 1993). ThePWE has long been discussed as an ingredient

    in contemporary US racism toward AfricanAmericans at the hands of EuropeanAmericans; African Americans are seen as not

    Copyright © 2006 SAGE Publications(London, Thousand Oaks, CA and New Delhi)9:1; 95–115; DOI: 10.1177/1368430206059874

    Author’s noteAddress correspondence to Sheri R. Levy,Department of Psychology, SUNY Stony Brook,Stony Brook, NY 11794-2500, USA[email: [email protected]]

  • conforming to the work ethic (not workinghard enough) and thus deserving disadvantage(e.g. Kinder & Sears, 1981; McConahay &Hough, 1976). More recently, PWE has beencast more broadly as a justifier of one’s ownprejudice and society’s differential treatment ofa wide variety of less successful or stigmatizedpersons including homeless persons, over-weight persons, and women (e.g. Crandall,1994, 2000; Levy, Freitas, & Salovey, 2002;Pratto, Sidanius, Stallworth, & Malle, 1994;Quinn & Crocker, 1999). Consistent with thistheorizing, in the US, PWE has been shown torelate to stronger negative attitudes towardAfrican Americans (e.g. Katz & Hass, 1988; alsosee Biernat, Vescio, & Theno, 1996), dislike ofoverweight persons (e.g. Crandall, 1994),negative attitudes toward homeless persons(e.g. Levy et al., 2002; Somerman, 1993),negative affect toward people facing AIDS (e.g.Levy et al., 2002), and opposition to a com-munity facility for homeless families (e.g.Somerman, 1993).

    Notwithstanding the commonly observedrelations between endorsing PWE and socialintolerance, we suggest that PWE may alsorelate to greater social tolerance among somepeople (or in some contexts). In this paper, weaim to demonstrate that the PWE is not simplyabout intolerance but rather has dual inter-group implications—one supporting socialintolerance (as previous work has shown) andone supporting tolerance. We further proposethat PWE has a surface meaning that supportssocial tolerance through the proposition thatanyone who works hard can succeed, implyingthat differences in abilities or talents areminimal and inconsequential. We also proposethat PWE’s intolerance meaning develops inpart with experience in particular (e.g. US)cultures.

    The proposition that the same construct canbe construed differently across people orcontexts has precedence in other lines of socialpsychological research (e.g. Bruner, 1957; Ross& Nisbett, 1991). However, research on laytheories has tended to focus on the unitaryintergroup meaning of lay theories, document-ing which lay theories support intergroup bias

    and which lay theories instead support inter-group tolerance (e.g. Altemeyer, 1998; Biernat,et al., 1996; Crandall, 1994; Haslam, Roth-schild, & Ernst, 2002; Hong, Chiu, Young, &Tong, 1999; Levy, Stroessner, & Dweck, 1998;Plaks, Stroessner, Dweck, & Sherman, 2000;Pratto et al., 1994; Whitley, 1999; Yzerbyt,Leyens, & Corneille, 1998; see Hong, Levy, &Chiu, 2001; Wegener & Petty, 1998). Forexample, greater endorsement of the lay viewthat group hierarchies are natural and necess-ary (‘social dominance orientation’; Pratto etal., 1994) relates to greater negative attitudestoward policies that promote equality acrossgender, social class, ethnic or racial groups, andsexual orientation; and toward the groups thatwould benefit from such policies in numerouscountries including Canada, China, Israel,Lebanon, Mexico, Taiwan, and the US (e.g.Altemeyer, 1998; Whitley, 1999; also seeSidanius & Pratto, 1999).

    Dual intergroup meanings

    Despite substantial evidence that particular laytheories either facilitate or impede tolerance,we suggest that some lay theories may be opento more than one intergroup implication. Onereason a lay theory could have two, evenopposite, intergroup implications is that a laytheory can have a surface meaning and anassociated meaning, which is the outgrowth ofcultural or personal experience. In otherwords, a lay theory may be initially learned interms of its surface meaning; then, with experi-ence, people may learn that the lay theoryholds additional meanings as part of the laytheory’s associative network. We propose thatPWE, on the surface, suggests the sociallytolerant stance that people from all socialcategories are basically equal and can allsucceed. Promising successful outcomes fordiligent effort, PWE is indeed used to motivateindividuals, as in the classic US children’s book,The Little Engine that Could. This is the story of alittle engine who, through diligent effort, wasable to reach a valued outcome that appearedinsurmountable (hauling cars filled with giftsover a tall mountain). If children accept the

    Group Processes & Intergroup Relations 9(1)

    96

  • PWE message at face value, as taught in suchbooks and by teachers encouraging all studentsto work hard, they will likely believe that anyonewho works hard can succeed. In this way, effortcan be something that equalizes people ofdifferent social categories. Everyone can putforth effort and succeed, so everyone is basicallyequal.

    Other associations, however, may becomelinked to a lay theory through experiences.With experience in US culture, people may seethat PWE can be used to support the con-clusion that lack of success reflects dispositionalfactors, such as laziness, which can in turn beused to justify inequalities and differential treat-ment of social groups (e.g. Crandall, 1994,2000; Katz & Hass, 1988; Somerman, 1993).Thus, with experience in the US, peopleeventually should also associate PWE with argu-ments justifying inequality such as theargument that members of disadvantagedgroups are to blame for their predicament,which could be alleviated through their owneffort.

    Our reasoning about the development ofpotential associated meanings of PWE suggeststhat, through experience, people accumulateand refine their understandings of PWE; thus,adults are likely to be more familiar with theintolerance meaning of PWE, but children(people less familiar with the culture orenvironment) likely see PWE through itssurface (‘social equalizer’) meaning only.Accordingly, in our studies, we compare theresponses of children to adults to help revealthe social equalizer meaning of PWE and ourhypothesized associated meanings mechanism(experience with PWE as a justifier of inequal-ity).

    In summary, we suggest that a lay theory mayaccrue new, even opposite, associatedmeanings, thereby having both a tolerance andan intolerance meaning. PWE seemed to be aprime candidate for dual intergroup meaningsbecause it is somewhat vaguely and broadlydefined, allowing for heterogeneity in its impli-cations. Additionally, people in the US arehighly invested in this culturally pervasive laytheory. Because it is difficult to give up, people

    may instead attempt to accommodate PWE toserve both tolerant and intolerant needs acrosssituations and over time.

    By contrast, social dominance orientation(SDO), the view that some groups are superiorto others (e.g. Pratto et al., 1994), although per-vasive in some settings, is not easily framed tofit social tolerance. SDO, in its direct suggestionthat some groups are inferior to others and inits preference for social hierarchies, specificallyendorses intolerance of lower status groups(e.g. Pratto et al., 1994; Sidanius & Pratto,1999). Thus, SDO is unlikely to also have anegalitarian implication, and we will examine theproposed unitary intergroup implication ofSDO in one developmental study.

    Moreover, lay theories such as PWE that canserve as justifiers of intolerance (e.g. Crandall,2000) in a seemingly egalitarian society aregood candidates for having a surface meaningpromoting social tolerance and an associatedmeaning promoting intolerance. That is, tojustify socially unacceptable and often person-ally unacceptable levels of prejudice in a societythat espouses egalitarian values, a lay theorymust appear to be somewhat egalitarian so thatany social intolerance appears to be a ‘fair’response.

    Before describing our investigations into thedual intergroup implications of PWE, weprovide some further background on PWE.

    Prior intergroup research on the PWE

    As noted from the outset, in the US, muchresearch and theorizing supports the notionthat PWE is a justifier of social inequalityamong adults. Crandall (2000) aptly noted that‘people, in the process of stigmatizing others,believe that the rejection, avoidance, andinferior treatment they dole out to stigmatizedothers is fair, appropriate, judicious—in otherwords, justified’; thus, as a result of justifierssuch as PWE, ‘one can continue to treat peopleas second-class citizens, apply a lower moralstandard, and practice exclusion with a clearconscience’ (p. 126).

    Although a large body of work with USadults is consistent with the notion that PWE is

    Levy et al. dual implications

    97

  • a justifier of social inequality, there is someevidence to the contrary. That is, there havebeen inconsistencies in PWE’s link to socialintolerance across samples, studies, or relevantmeasurement tools (and there may be moreinconsistencies that have not been published).For instance, in a large-scale study of nineEuropean American college samples, Katz andHass (1988; Study 1) found that, in one of thenine samples, PWE was not significantly posi-tively correlated with ‘Anti-Black attitudes’(seeing African Americans as ‘deviant’; e.g. ‘Onthe whole, Black people do not stress educationand training’) and in four of the nine samples,PWE was not significantly negatively related to‘Pro-Black attitudes’ (seeing African Americansas ‘disadvantaged’; e.g. ‘Too many Blacks stilllose out on jobs and promotions because oftheir skin color). Subsequently, Monteith andWalters (1998) found that PWE (using Katz &Hass’s measure) was unrelated to racism towardAfrican Americans as assessed by the ModernRacism Scale (e.g. McConahay & Hough,1976).

    Furthermore, Katz and Hass (1988) foundthat PWE was unrelated to, rather than nega-tively related to, a general measure of socialtolerance (humanitarianism/egalitarianism;e.g. ‘There should be equality for everyone—because we are all human beings’) in eight ofthe nine samples, a finding that has been sub-sequently replicated (e.g. Biernat et al., 1996;Monteith & Walters, 1998).

    In reflecting on some of the unexpectedmixed findings in their data set, such as thoseconcerning the relation between PWE and atti-tudes toward African Americans as notedabove, Katz and Hass (1988, p. 902) wrote: ‘wefind these exceptions puzzling.’ We suggest thatthose exceptions and other ones may not be‘noise’ in the data but rather suggest that PWEis not solely used to support intolerance.

    Overview of the present investigation

    Our main aim is to show that PWE has dualintergroup implications, namely that PWEsuggests egalitarianism and social tolerance tosome people (or in some contexts) and, in

    contrast, prejudice and social intolerance toothers (or in other contexts). Our secondaryaim is to provide preliminary evidence for anassociated meanings mechanism—that theintolerance implication is linked to PWEthrough social experience.

    These aims were addressed in four studiesconducted in the United States. In Studies 1and 2, we examined how age (experience) mayalter associations between PWE and social toler-ance among three age groups (people approx-imately 10, 15, and 20 years old). We expectedto show that PWE would be related to greaterlevels of social tolerance among children, whopresumably think of PWE in a socially tolerantway, relative to adults who, according to pastwork, tend to think of PWE in a socially intoler-ant way (e.g. Biernat et al., 1996; Katz & Hass,1988; Kinder & Sears, 1981; Levy et al., 2002;McConahay & Hough, 1976; Monteith &Walters, 1998; Somerman, 1993). We alsoexamined SDO (e.g. Pratto et al., 1994) as acomparison to PWE in Study 1, expecting SDOto hold a single (intolerant) meaning across thedifferent age groups.

    Thus, in Studies 1 and 2, we intended toprovide the first evidence to our knowledgethat PWE is related to social tolerance, in thiscase, among children. At the same time, weaimed to replicate prior findings among adultsshowing that, by contrast, PWE tends to supportintolerance among this age group. Importantly,however, our theorizing about adults directlysuggests that focusing on a particular aspect ofPWE should give rise to increased intolerance.In Studies 3 and 4, therefore, we attempted todirectly manipulate adults’ interpretations ofPWE and to demonstrate a situational triggerfor their use of PWE in an intolerant manner.

    Study 1

    In Study 1, we examined whether shifts in age(experience) among US students (approxi-mately 10, 15, and 20 years old) relate to differ-ent patterns of relations between PWE andsocial tolerance. Consistent with past work, weassessed the relation between a standardmeasure of PWE (Katz & Hass, 1988) and

    Group Processes & Intergroup Relations 9(1)

    98

  • beliefs about general social tolerance (Katz andHass’s (1988) egalitarianism/humanitarianismmeasure) as well as a measure of intendedbehavior toward a disadvantaged group in USsociety—African Americans (e.g. Katz & Hass,1988; Kinder & Sears, 1981; McConahay &Hough, 1976). As a first step, our sample waslimited to European Americans, given the largehistory of theorizing and research on therelation between PWE and social intolerance(particularly toward African Americans) amongEuropean Americans (e.g. Biernat et al., 1996;Katz & Hass, 1988; Kinder & Sears, 1981;McConahay & Hough, 1976; Monteith &Walters, 1998).

    As noted, we suggest that PWE’s surfacemeaning is the assertion that anyone who workshard can succeed, which should suggest theequivalence of people of different socialcategories. Thus, in the youngest group(approximately 10 years old), PWE shouldrelate positively to egalitarianism and negativelytoward desired social distance from AfricanAmericans. By approximately 15 years of age,the relations between PWE and markers ofsocial tolerance should weaken. Continuingthis hypothesized developmental trend, incollege, the relation between PWE and greatersocial tolerance should shift further away froma strong positive relation. Given that we expectadults to be familiar with both the egalitarianand justifier of inequality meanings of PWE,taken together with previous findings, we wouldexpect to find that PWE is either (1) unrelatedto measures of tolerance (according to our the-orizing, this would suggest that across partici-pants, both meanings were being used aboutequally often) or (2) significantly related togreater intolerance (suggesting that, across par-ticipants, the intolerance meaning is moreprevalent). As reviewed earlier, there is themost precedence in past research withEuropean American college students for (2)and some precedence for (1). Among collegestudents, then, we expected to see that therelation between PWE and tolerance would benonsignificant or negative. In the general dis-cussion, we address the issue of whether adults,once familiar with the intolerance meaning of

    PWE, use PWE in a purely egalitarian way there-after.

    As a contrast to our theorizing about PWE,we also assessed a lay view that did not appearto be open to additional meanings, but ratherappeared to have an unequivocal, stablerelation to prejudice. As reviewed earlier, SDO(Sidanius & Pratto, 1999) intrinsically pre-scribes the non-egalitarian view that somegroups are inherently superior to others;further, a consistent demonstration of researchfindings across cultures supports this relation(e.g. Pratto et al., 1994; Sidanius & Pratto,1999). We posit no associated meaning of SDOcontradicting this direct intolerance meaning.Accordingly, across the different age groups,SDO should be negatively correlated with egali-tarianism and positively correlated with desiredsocial distance from African Americans.

    Also, to help isolate the relation betweenPWE and social tolerance, we controlled statis-tically for other variables that could explain ourhypothesized findings such as participants’ self-esteem and their concerns with reportingsocially pleasing responses.

    MethodParticipants Participants were recruited fromnearby schools in New York. The collegestudent sample included 23 males and 82females, aged 17 to 25 years (M = 20.64), drawnfrom psychology courses requiring researchparticipation. The middle group, aged 14 to 16years (M = 15.01), included 37 males and 97females from 9th and 10th grade classrooms,and the youngest group, aged 9 to 12 years (M = 10.48), included 47 males and 60 femalesfrom 5th and 6th grade classrooms. Participantsincluded those students who agreed to partici-pate and, in the case of the non-college partici-pants, whose parents or legal guardiansprovided written consent. Because this studywas intended to focus on European Americans,the samples were limited to European Ameri-cans. All samples came from, on average,middle income environments.

    Measures Table 1 presents means, standarddeviations, and internal reliabilities of the

    Levy et al. dual implications

    99

  • measures. As can been seen, for each of the agegroups, there was good internal reliability.

    Lay theories and egalitarianism We slightlymodified existing scales of PWE (Katz & Hass,1988), SDO (Pratto et al., 1994), and egali-tarianism (Katz & Hass, 1988) from the adultliterature to create single measures of each thatwould be appropriate for use with all three agegroups. In a pilot study with non-study partici-pants (35 girls, 39 boys) of the same age as theyoungest participants in our study (9- to 11-year-old children, M = 10.05), we examined thereliability of these modified scales. Each of themeasures demonstrated good internalreliability (PWE’s � = .80; SDO’s � = .61; Egali-tarianism’s � = .70) and good test-retestreliability over a four month period (PWE’s r =.42, p < .001, SDO’s r = .53, p < .001, and Egali-tarianism’s r = .43, p < .001).

    Examples of items include: ‘If people workhard, they can get a very good job’ (PWE);‘Some groups of people are not as good asother groups of people’ (SDO); ‘Everyoneshould be treated equally because we are allhuman’ (egalitarianism). All statements wererated on a 6-point scale (1 = very stronglydisagree, 6 = very strongly agree). For eachscale, participants’ responses to all items(reverse-scoring when needed) were averagedto create three separate indices such thathigher scores indicated greater agreement withthe construct.

    Desired social distance from African AmericansOur measure of desired social distance wasdrawn from measures commonly used in thesocial psychological (e.g. Esses & Dovidio,2002) and developmental (e.g. Karafantis &Levy, 2004) literatures. Specifically, participantswere asked ‘How much would you like to livenear [be friends with] Black persons?’ (1 = notat all; 5 = very much). For ease of interpretationand discussion, participants’ responses to eachof the two questions were reverse-scored andthen averaged such that higher numbersindicate greater social distance from AfricanAmericans.

    Self-esteem and social desirability To assess theirtendency to provide socially desirable answers,participants were asked to rate five statements(e.g. ‘Have you ever said things to get people tolike you?’) on a 6-point scale (never to many,many times). As a brief assessment of self-esteem,respondents were asked to rate how good theyfelt about themselves (not at all good to very, verygood) and selected the face drawing that best fithow they felt about themselves (1 = large frown;7 = large smile; Andrews & Withey, 1976). Foreach scale, participants’ responses wereaveraged to create two separate indices, suchthat higher scores indicate greater agreementwith the construct.

    Procedure Two experimenters conducted thestudy at each site. All participants were tested in

    Group Processes & Intergroup Relations 9(1)

    100

    Table 1. Means, standard deviations, and internal reliability of measures according to age group, Study 1

    Mean (standard deviation) Internal reliability

    No. of Age group Age group Age group Age group Age group Age groupMeasure items 1 2 3 1 2 3

    PWE 5 4.58 (0.87) 4.25 (0.65) 3.78 (0.77) .79 .76 .81SDO 7 2.64 (0.67) 2.62 (0.66) 2.78 (0.76) .64 .79 .81Egalitarianism 6 4.91 (0.71) 4.32 (0.67) 4.09 (0.73) .76 .82 .80Interracial distance 2 0.79 (0.87) 1.14 (0.87) 1.71 (0.84) .91 .83 .70Social concerns 5 2.90 (0.66) 2.96 (0.67) 3.05 (0.61) .57 .71 .73Self-esteem 2 5.47 (1.07) 4.95 (1.09) 4.00 (1.10) .84 .82 .73

    Notes: PWE = Protestant work ethic; SDO = social dominance orientation; interracial distance = desired socialdistance from African Americans: Age group 1 = 10- to 12-year-old children: age group 2 = 14- to 16-year-oldadolescents: age group 3 = approximately 20-year-old college students.

  • classrooms of 25 to 35 students. For privacy,students were asked to separate their desks orsit one seat apart while completing the survey.

    ResultsWe began by testing our main hypothesis, thatPWE would have different intergroup-relevantmeanings with age. In separate simultaneousregression analyses, we regressed participants’egalitarianism and interracial social distancescores on variables representing PWE, partici-pants’ age groups, and the age group � PWEinteraction. As predicted, demonstrating sig-nificant differences in slopes representingPWE’s relations to egalitarianism and inter-racial social distance across the age groups, theinteraction term (Age Group � PWE) was sig-nificant for both egalitarianism (F (1, 340) =8.49 p < .01), and interracial social distance (F(1, 340) = 4.95, p < .05). As depicted in Figure1, with increasing age, the relation betweenPWE and egalitarianism went from significantlypositive to nonsignificant; whereas the relationbetween PWE and interracial social distancewent from significantly negative to significantlypositive. For egalitarianism and PWE, the corre-lation for Age Group 1 was significantly largerthan the correlation for Age Group 3 (Z = 2.69,p < .01), whereas the differences between thecorrelations for Age Groups 1 and 2 (Z = 1.58)and Age Groups 2 and 3 (Z = 1.26) were not sig-nificant. For PWE and interracial socialdistance, the correlation for Age Group 3 wassignificantly different from the correspondingcorrelations among Age Group 1 (Z = 2.82, p <.01), and Age Group 2 (Z = 3.32, p < .01),whereas there was no significant differencebetween Age Groups 1 and 2 (Z = –0.23).

    We also tested whether the predicted signifi-cant findings would remain significant whencontrolling for other relevant variables, namelyparticipants’ levels of social concerns and self-esteem. Thus, in separate simultaneous regres-sion analyses, we regressed participants’egalitarianism and interracial social distancescores on variables representing PWE, socialconcerns, self-esteem, participants’ age groups,and the age group � PWE interaction. Consist-ent with our theorizing, the interaction term

    (Age Group � PWE) remained significant forboth egalitarianism (F (1, 330) = 10.70, p < .01),and interracial social distance (F (1, 330) =6.33, p < .05), despite the inclusion of theseother relevant predictor variables in theanalysis.1

    Next, hoping to show conditions in whichassociated meanings do not accrue with age, weanalyzed the relation between SDO and inter-group tolerance across the three age groups. Inseparate simultaneous regression analyses, weregressed participants’ egalitarianism and inter-racial social distance scores on variables repre-senting SDO, participants’ age groups, and theage group � SDO interaction. SDO was signifi-cantly related, in line with our predictions (andpast work), to egalitarianism (F (1, 340) = 26.27,p < .001), and to interracial social distance (F(1, 340) = 16.16, p < .001), showing that SDOwas significantly related to greater intolerance.Consistent with our theorizing, the interactionterm (Age group � SDO) was nonsignificantfor both egalitarianism (F (1, 340) = .67, p >.40), and interracial social distance (F (1, 340)= 3.27, p > .07). The correlations between SDOand egalitarianism were virtually identicalacross the age groups (rs = –.57, –.55, –.56, allps < .001) whereas the correlations betweenSDO and desired social distance from AfricanAmericans, varied from the youngest to oldestgroup (rs = –.40, –.36, –.22, all ps < .05),although there were no significant differences(comparing age group 1 to 3, Z = 1.44, compar-ing age group 2 to 3, Z = 1.16).

    DiscussionIn this study, we provided the first evidence ofa shift in PWE’s intergroup meaning. Oneweakness of Study 1 was that the oldest partici-pants were drawn from a more select group ofthe population—college students—than theyounger groups who were drawn from the sur-rounding public school system. Thus, the shiftin meaning could be due to something specificto a college education or the college experi-ence. To address this, we asked adult com-munity members to complete a brief surveycontaining the PWE and egalitarianismmeasures from Study 1. These participants were

    Levy et al. dual implications

    101

  • recruited from malls, libraries, and restaurantsin close vicinity to the public elementary schoolattended by the younger participants in Study1. They included 47 European American adultsaged 23 to 66 (mean age = 41.83) who did nothave a four year college education. The corre-lation between PWE and egalitarianism wasnonsignificant among the community sample(r = .10) and virtually identical to the correla-tion found among the college students in Study1 (r = .11). Thus, the shift in PWE’s meaning

    cannot be easily explained by our use of collegestudents as a comparison.

    Taken together, then, our findings thus farshow that PWE has different intergroup impli-cations with age. For the younger samples, PWEwas related positively to egalitarianism andnegatively to desired social distance fromAfrican Americans, suggesting that, at theseages, PWE has a meaning that promotes socialtolerance. For the oldest sample, however, therelations between PWE and these same social

    Group Processes & Intergroup Relations 9(1)

    102

    -0.3

    -0.2

    -0.1

    0

    0.1

    0.2

    0.3

    0.4

    0.5

    1 2 3

    Age Group

    )r( tneiciffe

    oC

    noitalerr

    oC t

    nem

    oM-tc

    ud

    orP

    nosrae

    P

    PWE-Egalitarianism

    PWE-InterracialDistance

    Figure 1. Correlation of egalitarianism and interracial distance with the Protestant work ethic, amongelementary school, high school, and college students. Note: PWE = Protestant work ethic; Age group 1 = 10- to12-year old children; Age group 2 = 14- to 16-year old adolescents; Age group 3 = approximately 20-year oldcollege students.

  • tolerance measures were mixed (unrelated toegalitarianism, significantly positively related todesired social distance from African Ameri-cans), consistent with past findings and with theexpectation that adults do not solely use PWEin an intolerant way. In Studies 3 and 4, weaimed to provide a more direct test of the intol-erance meaning of PWE among adults.

    Evidence that markers of intergroup intoler-ance related to SDO consistently across agegroups, whereas the intergroup intolerancerelation to PWE was moderated by age,provides initial support for the proposition thatPWE’s implications shift with age due to anassociated meanings mechanism. The findingthat the age-related pattern of the relationbetween PWE and intergroup measuresremained significant when controlling for par-ticipants’ levels of social concerns and self-esteem suggests that our results do not likelyreflect one age group giving more socially desir-able answers or having a greater need to boosttheir self-esteem by justifying their higher socialstatus or derogating others.

    Study 2

    In Study 2, using an experimental induction ofPWE, we aimed to replicate the findings fromStudy 1 demonstrating that PWE relates tosocial tolerance among young European Amer-icans. Members of the same three age groupsfrom Study 1 were asked to read one of twobrief articles (equated for length at the readinglevel of the youngest age groups), each of whichreported allegedly credible and extensivepsychological research; however, the articlesdiffered in that they concluded that thefindings either supported or opposed PWE.The impact of each induction was then assessedwith the measure of egalitarianism used inStudy 1. Consistent with our theorizing andwith results from Study 1, we expected that apro-PWE (vs. anti-PWE) message would triggergreater egalitarianism among younger partici-pants, who it is assumed construe PWE in termsof its surface meaning relevant to egalitarian-ism, relative to college students, who arepresumably also familiar with PWE’s inequality-

    justifying associations. Checks on participants’temporary acceptance, understanding, andenjoyment of the PWE-relevant articles inaddition to a self-esteem measure wereincluded.

    MethodParticipants Participants were recruited fromthe same schools described in Study 1. Theoldest group (39 males, 130 females) was aged18 to 25 years (M = 21.31). The middle group(49 males, 106 females) was aged 14 to 16 years(M = 14.99). The youngest group (41 males and40 females) was aged 10 to12 years (M = 10.80).Consistent with Study 1, the samples werelimited to European Americans. Data from 14participants who did not provide an accuratesummary of the PWE-relevant articles wereexcluded.

    Procedure Participants were randomlyassigned to read either a pro-PWE (781 words)or anti-PWE (788 words) induction article. Pur-porting to convey psychological research, thearticles described identical methods (such as alarge, longitudinal study at Harvard University),but concluded either that ‘people who workhard do well and have a successful life’ (pro-PWE) or that ‘people who work hard are notalways successful’ (anti-PWE). Participants wereasked to briefly summarize the report afterreading it.

    Following the induction, participants com-pleted the measures of egalitarianism, PWE,and self-esteem used in Study 1 and rated theirunderstanding and enjoyment of the report(‘How much did you understand [enjoy] thereport that you read?’; 1 = not at all; 7 = very,very much). Finally, participants weredebriefed. As in Study 1, each measure exhib-ited good internal reliability for the youngest,middle, and oldest age groups, as demonstratedrespectively, �s for PWE = .92, .82, and .84; �sfor egalitarianism = .62, .76, and .82; rs for self-esteem = .88, .87, and .90.

    ResultsPreliminary analyses We began by examiningwhether all three age groups temporarily

    Levy et al. dual implications

    103

  • accepted the message encouraged by the PWE-relevant articles. As expected, across the threeage groups, participants who read the pro-PWEarticle subsequently endorsed PWE to a greaterextent (M = 5.21) than did participants whoread the anti-PWE article (M = 4.20, t(388) =8.20, p < .001; Cohen’s d = .83). This effect wasobtained among the youngest (t(67) = 5.57, p <.001; Cohen’s d = 1.37), middle (t(150) = 4.45,p < .001; Cohen’s d = .73), and oldest groups(t(167) = 4.86, p < .001; Cohen’s d = .75). Sug-gesting that the two articles were received insimilar ways by participants, participantsreported similar levels of enjoyment of the anti-PWE (M = 3.49) and pro-PWE articles (M =3.49; t(388) = 0.06, ns), and similar levels ofunderstanding the anti-PWE (M = 6.22) andpro-PWE articles (M = 6.36; t(388) = 1.39, p >.16), and participants’ membership in thedifferent age groups did not significantlymoderate the impact of the lay theory manipu-lation on their enjoyment (F < 1) or under-standing (F < 1) of the articles.2

    Primary analyses Our main hypothesis was thatage group would moderate the impact of thePWE-relevant messages on reported egalitarian-ism. To test this, we analyzed participants’ egali-tarianism scores in a 2 (PWE Message) � 3 (AgeGroup) analysis of variance. This analysisrevealed a significant Age Group effect (F (2,384) = 14.84, p < .001), which, as in Study 1 (seeTable 1), reflected younger participants’stronger endorsement of egalitarianism. Therewas no main effect of the PWE message (F (2,384) = 1.56, p > .21). Of greatest relevance is thefinding that the predicted Age Group � PWEMessage interaction was significant (F (2, 384) =8.49, p < .001). In an analysis of covariance, thiseffect remained significant when controlling forparticipants’ self-esteem and enjoyment andunderstanding of the articles (F (2, 380) = 6.89,p < .001).

    As illustrated in Figure 2, follow-up compari-sons revealed the nature of this interaction.Among the youngest group, those assigned toread the pro-PWE article subsequently reportedsignificantly higher levels of egalitarianism (M= 5.74) than did those assigned to read the anti-

    PWE article (M = 5.35; t(67) = 2.14, p < .05;Cohen’s d = .52). Similarly, among the middleage group, those assigned to read the pro-PWEarticle subsequently reported significantlyhigher levels of egalitarianism (M = 5.10) thandid those assigned to read the anti-PWE article(M = 4.70; t(150) = 2.61, p < .01; Cohen’s d =.43). In contrast, among the oldest group, thoseassigned to read the pro-PWE article subse-quently reported significantly lower levels ofegalitarianism (M = 4.58) than did thoseassigned to read the anti-PWE article (M = 4.96;t(167) = 2.17, p < .05; Cohen’s d = –.34).

    DiscussionConsistent with our theorizing and findingsfrom Study 1, results from Study 2 revealed thata pro-PWE message had markedly differenteffects on the reported egalitarianism of peopleof different ages. Although participants withinall three age groups endorsed the pro- or anti-PWE views presented to them, among theyoungest two groups only, those encouraged totemporarily adopt a pro-PWE view endorsedegalitarianism to a greater extent than didthose encouraged to temporarily adopt an anti-PWE view.

    Although these results are consistent withour theorizing about the dual intergroup impli-cations of PWE, particularly that young peoplewould be familiar with mainly the egalitarianmeaning of PWE, it is noteworthy that, amongthe oldest group, those encouraged to adopt apro-PWE view endorsed egalitarianism to a sig-nificantly lesser extent than did those encour-aged to adopt an anti-PWE view. The finding isconsistent with past theorizing and work sug-gesting that PWE has an intolerance meaning(e.g. Biernat et al., 1996; Crandall, 1994, 2000;Katz & Hass, 1988; Kinder & Sears, 1981;McConahay & Hough, 1976), but is incon-sistent with a smaller body of previous work,including our Study 1, which found a nonsignif-icant rather than negative relation betweenPWE and egalitarianism (e.g. Biernat et al.,1996; Katz & Hass, 1988; Montieth & Walters,1998). The finding from Study 2 indicates that,on average, this sample of college studentsviewed PWE as supporting intolerance.

    Group Processes & Intergroup Relations 9(1)

    104

  • Study 3

    Findings from Studies 1 and 2 as well as theprior work on PWE indicate that collegestudents tend to think of PWE in an intolerantway (e.g. Crandall, 1994; Katz & Hass, 1988).Importantly, however, our theorizing aboutadults directly suggests that specifically focusingon a particular aspect of PWE should give riseto its intolerance meaning. In Study 3, there-fore, we aimed to directly manipulate adults’interpretations of PWE to demonstrate a situa-tional trigger of the intolerance meaning ofPWE.

    To briefly review our theorizing, we hypothe-

    size that PWE’s implication for intolerancedevelops in part from social and cultural experi-ence. One way in which the justifier of inequal-ity meaning may arise is through exposure toothers using PWE to justify the status quo(inequality), as in the argument that disadvan-taged groups and group members are to blamefor their disadvantage and that they could pullthemselves out of their dire situation by simplyputting forth some effort. Repeatedly experi-encing PWE used in this way should increasethe likelihood that thinking about how peopleuse PWE to justify their arguments cues theintolerant meaning of PWE. Thus, havingparticipants consider others’ use of PWE in

    Levy et al. dual implications

    105

    4

    4.2

    4.4

    4.6

    4.8

    5

    5.2

    5.4

    5.6

    5.8

    6

    1 2 3

    Age Group

    msi

    nair

    atila

    gE

    htiw t

    ne

    me

    erg

    A

    Pro-PWE

    Anti-PWE

    Figure 2. Difference in endorsement of egalitarianism for participants assigned to read articles supporting(Pro-PWE) versus undermining (Anti-PWE) the Protestant work ethic, among elementary school, highschool, and college students. Note: Age group 1 = 10- to 12-year old children; Age group 2 = 14- to 16-year oldadolescents; Age group 3 = approximately 20-year old college students.

  • arguments is expected to trigger the associatedmeaning of PWE.

    In this study, all college student participantswere instructed to engage in a thoughtexercise. Half of the participants were asked tothink and write about instances of others using‘people who work hard succeed’ in support oftheir arguments (justification condition)whereas the other half of participants wereasked to think and write about what ‘peoplewho work hard succeed’ means (definition con-dition). Adults who thought about others’ useof PWE in arguments were expected to endorseegalitarianism to a less extent than adults whoconsidered the definition of PWE.

    Addressing the possibility that merely con-sidering any instances of others using a state-ment in an argument (justification condition)would impact one’s level of egalitarianism, twoadditional, control conditions included thesame justification and definition conditioninstructions regarding the lay theory, ‘Absencemakes the heart grow fonder; sometimesspending too much time together is bad for acouple’, which was expected not to hold differ-ent intergroup implications across conditions.

    MethodParticipants A total of 135 undergraduates(63 women, 72 men), aged 18–29 (M = 21.45),all native speakers of English, received US$6.00for participating. Race/ethnicity informationwas not assessed in this sample due to an errorin the computer program. The racial/ethniccomposition of all undergraduates enrolled atthis university during the year of the experi-ment was: 9% African American, 23% Asian,7% Latino, 36% European American, 18%unknown, and 8% other.

    Procedure Participants were seated in individ-ual cubicles with computers and were randomlyassigned to one of four conditions. The PWE-justification instructions stated: ‘Please thinkabout times you may have heard others make aparticular statement in discussions or argu-ments, in order to support their views. That is,please think about instances in which otherpeople have used this statement to help support

    a particular point they were trying to make.Think about how others have argued that:People who work hard succeed; people who donot work hard fail’. The PWE-definition induc-tion instructions stated: ‘Please think specifi-cally about the meaning of a particularstatement. That is, think about what this state-ment means. Think about the statement:People who work hard succeed; people who donot work hard fail’. Participants randomlyassigned to one of the control conditions weregiven identical justification and definitioninstructions for: ‘Absence makes the heart growfonder; sometimes spending too much timetogether is bad for a couple.’

    To facilitate participants’ involvement inthese thought exercises, they were instructed tospend five to ten minutes typing some of theirthoughts via computer. It is important to notethat neither task instructions (justification vs.definition), task content (PWE vs. Control), northeir two-way interaction significantly impactedtime spent on this exercise, recorded in milli-seconds via computer (all Fs < 1).

    Participants lastly completed the measure ofegalitarianism used in Studies 1 and 2 and indi-cated their level of agreement (using the same6-point scale) with the two statements used inthe study to define PWE: ‘People who workhard succeed’, and ‘People who do not workhard fail’.

    Results and discussionAs predicted, participants assigned to the PWE-justification condition subsequently reportedlower levels of egalitarianism (M = 3.88) thandid those assigned to the PWE-definition con-dition (M = 4.86; t(64) = 3.06, p < .01; Cohen’sd = .76). Among the control conditions, incontrast, there was no significant difference inreported egalitarianism between those whoreceived the justification instructions (M =4.85) and definition instructions (M = 4.90;t(67) = 0.17, ns). As shown in Figure 3, thesefindings constituted a significant interactionbetween task instructions (justification vs. defi-nition) and task content (PWE vs. control) (F(1, 131) = 3.98, p < .05, �p

    2 = 0.03). Only amongparticipants considering the justifying uses of

    Group Processes & Intergroup Relations 9(1)

    106

  • PWE did we find lower levels of egalitarianism.Thus, merely thinking about people using anylay theory in an argument does not promotelower levels of egalitarianism.

    To better understand the differences betweenthe PWE-justification and PWE-definition con-ditions, we also coded participants’ freeresponses in those conditions. As expected,justification-condition participants (66.7% vs.5.6% of definition-condition participants) men-tioned significantly more instances of blamingpeople for their misfortune (e.g. ‘it is a shamethat we support such things as welfare, whichenables people to be lazy and unambitious, to sitaround all day long’) whereas definition-condition participants (94.4% vs. 33.3% ofjustification-condition participants) were signifi-cantly more likely to simply restate the PWE (e.g.‘people who work hard by putting in time andeffort succeed and those who don’t will notsucceed. Those who work hard achieve in lifeand those who do not work hard do not achievein life’), χ2(1, N = 66) = 14. 57, p < .001, � = 0.64).

    Moreover, participants assigned to the PWEconditions did not significantly differ in agree-

    ment with the PWE statements (Mjustification =4.36; Mdefinition = 4.30; t(64) = 0.16, ns). Thus, itappears that the meaning of PWE, not endorse-ment of PWE, was influenced by the experimen-tal induction. Thinking about PWE used insupport of arguments (e.g. as a justifier ofinequality), then, seems to contribute to thePWE-intolerance relation.

    Study 4

    In Study 4, we aimed to provide furtherevidence that experience with PWE as a justifiercontributes to PWE’s intolerance meaning byusing the same justification and definitionthought exercises from Study 3, but this time,assessing their impact on actual intergroupbehavior—monetary donations to a homelessshelter. College student participants receivedeither the justification or definition inductionsand then were introduced to an ostensiblyunrelated task for which they were paid twodollars. Borrowing from successful prior induc-tions promoting greater helping (e.g. Batsonet al., 1997; Levy et al., 2002), participants read

    Levy et al. dual implications

    107

    3

    3.5

    4

    4.5

    5

    PWE Control

    Statement Content

    hti

    w tne

    meerg

    Amsi

    nairatilag

    E

    Definition Condition

    Justification Condition

    Figure 3. Endorsement of egalitarianism for college students assigned to think about statements of differingcontent, reflecting either PWE (Protestant work ethic) or a control topic (‘Absence makes the heart growfonder’), under instructions to focus on either how others have used the statements to justify their views(justification conditions) or what the statement means (definition conditions).

  • about a local homeless shelter and were giventhe opportunity to donate money. Adultsfocused on others’ use of PWE in support oftheir arguments were expected to donate lessmoney than adults focused on the definition ofPWE. A control condition in which participantsdid not engage in a thought exercise wasincluded for comparison purposes.

    MethodParticipants A total of 132 undergraduates(18 men, 67 women, 47 unidentified), aged 18to 22 years (M = 19.96), participated inexchange for extra credit in their psychologycourse and unexpectedly received US$2. Par-ticipants were 12.9% African American, 7.6%Asian, 8.3% Latino, 25.8% European American,9.8% other, 35.6% unavailable.

    Procedure Participants received a largeenvelope containing the study materials.Except for participants randomly assigned tonot participate in the thought exercises(control condition), the first pages of thesurvey consisted of ‘Study 1’, in which partici-pants were provided either the justification ordefinition thought exercises. The subsequentpages consisted of ‘Study 2’, which included anews column about a local homeless shelterand instructions for donating money to thatshelter. Study 2 was introduced by a letterostensibly from the professor of the coursestating that she was ‘serving on a committee onstudent affairs and thus was asked to distributethis survey . . . The committee has some fundsso you are being paid $2’. Stapled to the lastpage of the survey packet was a sealed envelopecontaining, in US currency, four quarters andten dimes. Instructions contained in the surveydirected participants to take their payment andto consider whether or not to make a cashdonation to the homeless shelter. In thedebriefing, no participants reported being sus-picious of any aspect of the study.

    MaterialsThought exercise Participants in the experimen-tal conditions were asked to complete the samethought exercise regarding PWE (justification

    and definition instructions) as described inStudy 3.

    Description of a homeless shelter The cover storythat was used to introduce the homeless shelterwas adopted from Levy et al.’s (2002) adapta-tion of Batson and colleagues’ (1997) investi-gations of helping among college students.Participants were told that they would bereading and evaluating a new column for oneof the university newspapers and moreover thatthe article contains transcripts from actualinterviews with homeless persons.

    After reading the sample article, participantsreceived a note from the supposed studentcommittee who needed the feedback on thearticle: ‘It occurred to us that some peoplereading the news article about homelesspersons . . . might wish to help them . . .donations would be most helpful . . . envelopeattached to this survey packet contains your $2payment . . . If you would like to donate somemoney . . . place it in the large yellow envelopethat your survey came in . . . your participationin this study in no way obligates you to donate’.

    Results and discussionAs predicted, justification-condition partici-pants subsequently donated significantly lessmoney (M = $1.28) than definition-conditionparticipants (M = $1.77) (t(88) = 2.90, p < .01),and control-condition participants (M = $1.66)(t(91) = 2.28, p < .05). The definition andcontrol conditions did not significantly differfrom one another (t(85) = 0.65, ns) This set offindings constituted a significant effect acrossconditions (F (2, 131) = 4.72, p < .05).3

    We also analyzed participants’ free responsesfrom the thought exercises. Consistent withfindings from Study 3, justification-conditionparticipants (37.8%) mentioned significantlymore instances of blaming people for their mis-fortune than did definition-condition partici-pants (9.5%) whereas definition-conditionparticipants (90.5%) were significantly morelikely to simply restate the PWE than justifi-cation-condition participants (62.2%) (p < .001).Differences in the study environment likelycontributed to the fewer blame explanations

    Group Processes & Intergroup Relations 9(1)

    108

  • given in this study than in Study 3. Study 3 par-ticipants typed their essays into a computer in aprivate cubicle, which likely made participantsfeel more comfortable writing socially undesir-able responses, compared to Study 4 partici-pants who wrote their essays while in aclassroom with other participants.

    In summary, the PWE-justification conditionseemed to temporarily decrease people’sdonations relative to the other conditions. Inline with the results from Study 3, these resultssuggest that experience with PWE used in argu-ments as a justifier of inequality contributes tothe PWE-intolerance relation.

    General discussion

    In this paper, we aimed to show that a lay theorycould have dual intergroup implications.Specifically, we aimed to show that PWE is notsimply related to intergroup intolerance in theUnited States, as past work has indicated, butrather PWE has implications for both intoler-ance and tolerance. We also aimed to provideevidence for an associated meaning mechan-ism, namely that the intolerance implication islinked to PWE through social experience.Results from four studies support these aims.

    Because previous research (conducted withyoung adults) has tended to demonstrate thatPWE relates to endorsing intolerance towardless advantaged or more stigmatized groups,such work seems to have implied either (a) thatthe intrinsic logic of PWE directly prescribesintolerance, or (b) that these relations solelyreflect people’s use of PWE to justify one’s pre-existing, prejudiced beliefs. In Study 1,however, PWE positively related to egalitarian-ism and desired social closeness to AfricanAmericans among those roughly 10 and 15years old. Accordingly, an assumption that thelogic of PWE intrinsically prescribes intolerancedoes not appear to tell the whole story. Study 2showed that manipulating participants’adoption of PWE influenced their intergroupbeliefs, promoting higher levels of egalitarian-ism among younger participants and lowerlevels of egalitarianism among older partici-pants. Moreover, then, the idea that PWE

    relates to intergroup attitudes only as a justify-ing afterthought also does not appear to tell thewhole story. Instead, we suggest that relationsbetween PWE and intolerance reflect in partthe acquired meanings that become associatedwith PWE through social experience in the US.Further supporting this explanation, Studies 3and 4 showed that adults focused on instancesof others using PWE to justify their viewsendorsed egalitarianism to a lesser extent anddonated less money to a homeless shelter thandid adults focused on the definition of PWE.

    In environments in which two or moremeanings of a lay theory are present, it ispossible that the initial meaning of a lay theoryfor example as understood by children is notthe surface meaning of a lay theory but insteadan associated meaning. The socioculturalenvironment could emphasize an associatemeaning to children thereby overriding thesurface meaning. Research is needed to addressthis possibility.

    Some remaining issuesOur findings show that younger persons tend touse the social equalizer meaning of PWE and,consistent with prior work, that adults tend touse the justifier of inequality meaning of PWE.An important remaining issue is whether thejustifier of inequality meaning of PWE replacesthe social equalizer meaning of PWE orwhether the social equalizer meaning remainsaccessible such that some adults (or adults insome contexts), use the social equalizermeaning of PWE. We suspect that once aware ofthe justifier of inequality meaning of PWE,people are unlikely to think of PWE exactly asthey did before; however, we suggest that adultscan still view PWE in an egalitarian way. As men-tioned previously, for PWE to function effec-tively for adults as a justifier of inequality in aseemingly egalitarian society, it would need toat least appear egalitarian among adults. Someadults may sincerely continue to believe thathard work is a social equalizer, which is sup-ported anecdotally by ‘rags to riches’ stories(figures such as Andrew Carnegie, OprahWinfrey) that have long captured headlines andhave been the basis for popular books in the US

    Levy et al. dual implications

    109

  • (e.g. Heykoe & Hock, 2003; Liberman &Lavine, 2000).

    In our recent work, we found that adultsreport using the social equalizer meaning ofPWE. Study participants were asked to explicitlyreport the extent to which they use PWE indifferent ways (e.g. ‘When you say things like“People who work hard succeed”’, tell us howmuch you mean this: ‘Anyone can work hardand succeed because people in different groupshave similar abilities and the potential to dowell’ [social equalizer]; ‘Hard work is all that’snecessary for success, so it is not fair to givepreferences to race-minority groups like Blacksand Latinos’ [deny racial inequality]).European American college students, onaverage, report strong use of the egalitarianmeaning and relatively weaker use of the denialof racial inequality meaning. Although such ameasure has its disadvantages, the findingssuggest that both meanings are likely availableand used by adults. It is important to note thata weakness of the current investigation is thatwe did not include direct measures of thedifferent intergroup meanings of PWE.

    Future work may aim to induce an egalitarianmeaning of PWE among adults—that is, aninduction that boosts people’s social tolerancelevels significantly above and beyond a controlcondition. A successful induction of this sortwould suggest that the developmental trendtoward intolerance shown in Studies 1 and 2 ofthe current investigation could be reversed.

    Findings from this investigation point to theneed for greater focus on different age groupsto better understand how, why, and whenpeople endorse a particular lay theory or par-ticular implication of that lay theory. Thisinvestigation, however, was limited in its abilityto address the important issue of lay theoryendorsement and use by groups differing onfactors other than age. Does PWE develop the‘justifier of inequality’ meaning for all groupsand in all cultures? Following from the assump-tion that people accumulate and refine theirunderstandings of certain lay theories such asPWE through particular social and culturalexperiences, PWE should not accumulate thesame meanings in all environments and

    cultures. Conceiving of culture in broad terms,PWE should not develop the justifier ofinequality meaning in cultures where peopletend to blame others less for their disadvantage.Prior work suggests that Latin American adultstend to blame others less for their disadvantageor stigma (being overweight, failing at a task)than US adults (e.g. Betancourt & Weiner,1982; Crandall & Martinez, 1996). In a recentstudy of Colombians (Mestizos) of ages similarto those in Studies 1 and 2, we found that thecorrelation between PWE and egalitarianismwas significantly positive and similar across thethree age groups (Levy, West, & Ramirez, inpress). Thus, unlike our current findings withEuropean Americans, PWE appears to consist-ently relate to egalitarianism among Mestizos,the majority group in Colombia. This resultseems to support the idea that the egalitarianmeaning of PWE is the surface rather thanassociated meaning of PWE and more broadlyrepresents additional evidence that socio-cultural experiences shape the meaning ofPWE.

    Even within cultures such as the UnitedStates in which the justifier of inequalitymeaning of PWE is available, not everyone inthe culture may be equally exposed to it.European American children’s and early ado-lescents’s stronger report of the egalitarianmeaning of PWE suggests the justifier ofinequality meaning of PWE is not as prevalentin their immediate environment compared toEuropean adults. The justifier of inequalitymeaning of PWE also may be less likely to bedirectly highlighted to members of relativelydisadvantaged groups in the US. After all, thatmeaning of PWE justifies advantaged groupmembers’ place in society. Members of disad-vantaged groups seem more likely to be repeat-edly exposed to the social equalizer meaning ofPWE by, for example, family and friendsbecause that meaning conveys a positivepathway in society despite their disadvantage.Indeed, our recent findings indicate that PWEhas a predominately egalitarian meaningamong African American and US Latino adultscompared to European American adults (Levyet al., in press). It is also possible that members

    Group Processes & Intergroup Relations 9(1)

    110

  • of disadvantaged groups reject PWE because oftheir familiarity with the intolerant meaningrather than focusing on the egalitarianmeaning, which is an important issue requiringfurther study.

    Thus, by studying a wider diversity of laytheorists, we may be able to tease apart the com-plexities behind sociocultural, political, andmotivational factors that influence the degreeto which members of different groups endorseand use particular interpretations of the laytheories.

    Future work is also needed to examine inmore detail what accounts for the develop-mental shift in the meaning of PWE as revealedin our present studies. The shift is likelybrought on by a combination of cognitive,sociocultural, and motivational factors.Although even the youngest participants in ourstudies have already passed through the majordevelopmental milestones, cognitive matura-tion is ongoing during the adolescent period.Still, our recent findings with AfricanAmerican, US Latino, and Mestizo Colombianadults which indicated their tendency to usePWE in an egalitarian way suggest that theEuropean American children do not simply usethe egalitarian meaning of PWE because of cog-nitive immaturity.

    As emphasized in the present work, sociocul-tural factors likely play a key role in the shift inPWE’s meaning; however, they likely work inconcert with motivational forces. The justifierof inequality meaning of PWE seems lessrelevant to children relative to the social equal-izer meaning of PWE. Parents, teachers, andother adults in the immediate environment ofchildren likely want to motivate children towork hard and be successful.

    Studying life transitions, which have motiva-tional effects (e.g. Higgins & Parsons, 1983;Ruble, 1994), appears to be a fruitful avenuefor developmental investigations in this area.For example, while preparing to compete forcollege or for job placement at any age whenone’s place in the system is mostly directly inquestion, members of relatively advantagedgroups may be particularly likely to receive andbe receptive to the justifier of social inequality

    meaning of PWE. We are addressing this incurrent work.

    Another key remaining issue is whether ourtheorizing about the dual intergroup impli-cations of lay theories is specific to PWE or gen-eralizes to other lay theories. As noted earlier,lay theories that are justifiers of inequality in aseemingly egalitarian society seem to be goodcandidates for having more than one inter-group meaning, namely for having a surfacemeaning promoting social tolerance and anassociated meaning that serves as a justifier ofintolerance.

    In our current work, we are studying anotherjustifier of inequality in the US, the colorblindtheory, and have found promising preliminaryresults for two contradictory intergroupmeanings. The colorblind theory, whichsuggests that social category information suchas race is irrelevant, can be used in certaincontexts (or by certain individuals) to supportthe socially tolerant stance that people shouldbe judged as individuals rather than asmembers of a group (e.g. Allport, 1954;Wolsko, Park, Judd, & Wittenbrink, 2000).However, in other contexts (and for other indi-viduals) the colorblind theory can be used inother ways (see Richeson & Nussbaum, 2004),for example, to deny the existence of racismand to blame disadvantaged groups for theirdisadvantage (e.g. Neville, Lilly, Duran, Lee, &Browne, 2000; Schofield, 1986).

    Serving as a justifier of inequality within aseemingly egalitarian environment is just oneconsideration in determining whether laytheories have more than one meaning and par-ticularly contradictory ones. Consistent with amotivational interpretation, lay theories likelyhave more than one intergroup implication tobetter serve the needs of lay theorists. If a laytheory is salient and socially acceptable in asetting in which people have a range of socialbeliefs such as tolerant and intolerant beliefs,people may attempt to use that lay theory inmultiple ways. For example, PWE and color-blind theory are pervasive theories in whichAmericans are highly invested in; so, peoplemay be motivated to accommodate such laytheories to their varying needs across situations

    Levy et al. dual implications

    111

  • and over time rather than simply discarding thelay theories. Still, a lay theory would need to besomewhat vaguely and broadly defined to allowfor flexibility in its implications and socialappeal to people with a wide range of beliefsand goals. Social dominance orientation,although potentially salient in settings wherepeople have wide range of beliefs, is not easilyframed to fit social tolerance. In contrast, PWEis more broadly defined and thus potentiallyhas wider appeal to people both high and lowin social tolerance. Further, the social andpolitical climate would need to be receptive todifferent uses of that lay theory. For instance, assuggested earlier, PWE is unlikely to be used asa justifier of inequality in environments inwhich people tend not to be blamed for theirnegative outcomes, but rather external forcesare blamed. Political movements or campaigns,nonetheless, may be able to bring about a tem-porary or long-lasting interpretative shift in alay theory.

    Some of our findings suggest that US adultsmay move somewhat seamlessly between differ-ent intergroup meanings of PWE. Findingsfrom the current Studies 3 and 4 suggest thatadults can be spontaneously led to think aboutPWE as justifier of inequality. In addition, ourrecent work shows that European Americanadults spontaneously invoke the justifier ofinequality meaning of PWE in response to a situ-ational trigger—reading about an AffirmativeAction program that would negatively impactmembers of their ingroup. Thus, self and groupinterests as well the broader sociocultural andpolitical context likely provoke interpretiveshifts in PWE and other lay theories.

    This investigation also speaks to the stableand flexible aspects of lay theories. Onceadopted, a lay theory could become a pos-session (see Abelson, 1986), which ‘predisposesthe follower to observe and accept facts compat-ible with the theory, and it becomes the corefeature in group allegiances and social dis-course’ (Furnham, 1988, p. 226). At the sametime, when a lay theory is environmentally acti-vated, people can rather readily (but notnecessarily permanently) switch to viewing theirworld, and rendering relevant intergroup

    judgments, through a different lay theory (onethat is likely familiar to them through previoussocial experiences; e.g. Dweck, Chiu, & Hong,1995; Hong et al., 2001). Results from ourstudies of PWE and theorizing about laytheories with dual intergroup implicationssuggest the potential for additional flexibilityand stability not afforded to lay theories withunitary intergroup implications. Becausepeople may be able to use lay theories such asPWE to support divergent positions, attitudes,or behaviors, they can then use the same laytheory in a wide variety of contexts—ones pro-moting either tolerance or intolerance. Thisflexibility additionally provides a mechanismfor the potential stability of lay theories withmore than one meaning. Because people canuse the same lay theory across time and situa-tions, they do not need to give up the lay theorywhen the social implications are not relevant orappropriate in a particular context. Futurework, however, is needed to test this conjecture.

    In conclusion, the current investigationshowed that PWE has two opposite intergroupimplications; that is, PWE not only has relationsto intolerance (as past work has shown) but alsohas relations to tolerance. We also providedpreliminary evidence for an associated meaningmechanism underlying the development of theintolerance meaning among adults. Further,there is preliminary evidence that adults maynot always use the intolerance meaning of PWE.The intolerance implication of PWE likelydepends on adults focusing on and being moti-vated by the justification aspect of PWE. Morework is needed to understand the extent towhich adults use the tolerance meaning ofPWE.

    More broadly, results from the presentinvestigation suggest that intolerance coulddevelop at least partly through adopting a laytheory with opposite intergroup implications.Also, people who endorse lay theories such asPWE in the US may only appear to be speakingthe same language; these individuals mayactually intend to communicate vastly differentmessages. Further study of lay theory develop-ment may help illuminate how intergroup atti-tudes develop and change.

    Group Processes & Intergroup Relations 9(1)

    112

  • Notes1. The interaction term (Age group � PWE) also

    remained significant for both egalitarianism(F (1, 339) = 8.35, p < .01), and interracial socialdistance (F (1, 339) = 4.84, p < .05), whenconsidering the gender of the participant. Therewas a significant main effect for gender, ashowing consistent with past work (e.g. Prattoet al., 1994), that females report greater socialtolerance (F (1, 339) = 6.33, p < .05).

    2. In our preliminary analyses, we also examinedpotential gender differences in a 2 (PWEmessage) � 3 (Age group) � 2 (Gender) analysisof variance on participants’ egalitarianism scores.The gender of the participants contributed toone statistically significant effect, a gender maineffect, which, consistent with past work includingStudy 1, indicated that that females report higherlevels of social tolerance than do males (F (2, 384)= 4.44, p < .05).

    3. This same pattern of findings was revealed whenconsidering donated money as a discrete variable(whether or not donated). That is, there was asignificant effect across conditions (F (2, 131) =5.80, p < .01). Follow-up analyses revealed thatjustification-condition participants weresignificantly less likely to donate money (68%)than definition-condition participants (92.7%)(t(88) = 3.13, p < .01) and control-conditionparticipants (88.6%) (t(91) = 2.67, p < .01). Thedefinition and control conditions did notsignificantly differ from one another (t(85) =0.51, ns).

    Also noteworthy is that in preliminary analyses,we examined potential gender of and race ofparticipant differences and found no significantmain effects or interactions with condition inpredicting donations to the shelter. However, oursample was limited in terms of the race andgender of the participants; plus, about one thirdof participants did not provide their race orgender.

    AcknowledgmentsWe are grateful to Carol Dweck and Antonio Freitasfor valuable comments on this line of work. Thismaterial is in part based upon work supported by theNational Science Foundation under Grant No. 0213660 to the first author.

    ReferencesAbelson, R. P. (1986). Beliefs are like possessions.

    Journal for the Theory of Social Behavior, 16, 223–250.Allport, G. W. (1954). The nature of prejudice.

    Cambridge, MA: Addison-Wesley.Altemeyer, B. (1998). The other ‘authoritarian

    personality’. Advances in Experimental SocialPsychology, 30, 47–92.

    Andrews, F. W., & Withey, S. B. (1976). Socialindicators of well-being: American’s perceptions of lifequality. New York: Plenum.

    Batson, C. D., Sager, K., Garst, E., Kang, M.,Rubchinsky, K., & Dawson, K. (1997). Is empathy-induced helping due to self–othermerging? Journal of Personality and Social Psychology,73, 495–509.

    Betancourt, H., & Weiner, B. (1982). Attributions forachievement-related events, expectancy, andsentiments: A study of success and failure in Chileand the United States. Journal of Cross-CulturalPsychology, 13, 362–374.

    Biernat, M., Vescio, T. K., & Theno, S. A. (1996).Violating American values: A ‘value congruence’approach to understanding outgroup attitudes.Journal of Experimental Social Psychology, 32,387–410.

    Bruner, J. S. (1957). Beyond the information given.In J. S. Bruner et al. (Eds.), Contemporary approachesto cognition. Cambridge, MA: Harvard UniversityPress.

    Crandall, C. S. (1994). Prejudice against fat people:Ideology and self-interest. Journal of Personality andSocial Psychology, 66, 882–894.

    Crandall, C. S. (2000). Ideology and ideologies ofstigma: The justification of stigmatization. In T. F.Heatherton, R. E. Kleck, M. R. Hebl, & J. G. Hull(Eds.), The social psychology of stigma (pp. 126–150).New York: Guilford.

    Crandall, C. S., & Martinez, R. (1996). Culture,ideology, and antifat attitudes. Personality and SocialPsychology Bulletin, 22, 1165–1176.

    Dweck, C. S., Chiu, C., & Hong, Y. (1995). Implicittheories and their role in judgments andreactions: A world from two perspectives.Psychological Inquiry, 6, 267–285.

    Esses, V. M., & Dovidio, J. F. (2002). The role ofemotions in determining willingness to engage inintergroup contact. Personality and Social PsychologyBulletin, 28, 1202–1214.

    Furnham, A. (1988). Lay theories: Everydayunderstandings of problems in the social sciences. NewYork: Pergamon Press.

    Levy et al. dual implications

    113

  • Haslam, N., Rothschild, L., & Ernst, D. (2002). Areessentialist beliefs associated with prejudice? BritishJournal of Social Psychology, 41, 87–100.

    Heykoe, K., & Hock, S. (2003). Harvesting the dream:The rags-to-riches tale of the Sutter Home Winery. NewYork: Wiley.

    Higgins, E. T., & Parsons, J. E. (1983). Socialcognition and the social life of the child: Stages assubcultures. In E. T. Higgins, D. N. Ruble, & W. W. Hartup (Eds.), Social cognition and socialdevelopment (pp. 15–62). New York: CambridgeUniversity Press.

    Hong, Y., Chiu, C., Young, G., & Tong, Y. (1999).Social comparison during the political transition:Interaction of entity versus incremental beliefsand social identities. Journal of InterculturalRelations, 23, 257–279.

    Hong, Y. Y., Levy, S. R., & Chiu, C. Y. (2001). Thecontribution of the lay theories approach to thestudy of groups. Personality and Social PsychologyReview, 5, 98–106.

    Karafantis, D. M., & Levy, S. R. (2004). The role ofchildren’s lay theories about the malleability ofhuman attributes in beliefs about andvolunteering for disadvantaged groups. ChildDevelopment, 75, 236–250.

    Katz, I., & Hass, R. G. (1988). Racial ambivalence andAmerican value conflict: Correlational and primingstudies of dual cognitive structures. Journal ofPersonality and Social Psychology, 55, 893–905.

    Kinder, D. R., & Sears, D. O. (1981). Prejudice andpolitics: Symbolic racism versus racial threats tothe good life. Journal of Personality and SocialPsychology, 40, 414–431.

    Levy, S. R., Freitas, A. L., & Salovey, P. (2002).Construing action abstractly and blurring socialdistinctions: Implications for perceivinghomogeneity among, but also empathizing withand helping, others. Journal of Personality and SocialPsychology, 83, 1224–1238.

    Levy, S. R., West, T., & Ramirez, L. (in press). Laytheories and intergroup relations: A socialdevelopmental perspective. In W. Stroebe & M.Hewstone (Eds.) The European Review of SocialPsychology.

    Levy, S. R., Stroessner, S. J., & Dweck, C. S. (1998).Stereotype formation and endorsement: The roleof implicit theories. Journal of Personality and SocialPsychology, 74, 1421–1436.

    Liberman, G., & Lavine, A. (2000). Rags to riches:Motivating stories of how ordinary people achievedextraordinary wealth. Dearborn trade.

    McConahay, J. B., & Hough, J. C. (1976). Symbolicracism. Journal of Social Issues, 32, 23–45.

    Monteith, M. J., & Walters, G. L. (1998).Egalitarianism, moral obligation, and prejudice-related personal standards. Personalityand Social Psychology Bulletin, 24, 186–199.

    Neville, H. A., Lilly, R. L., Lee, R. M., Duran, G., &Browne, L. (2000). Construction and initialvalidation of the Color-Blind Racial AttitudesScale. Journal of Counseling Psychology, 47, 59–70.

    Plaks, J. E., Stroessner, S. J., Dweck, C. S., &Sherman, J. (2000). Person theories and attentionallocation: Preferences for stereotypic versuscounter stereotypic information. Journal ofPersonality and Social Psychology, 80, 876–893.

    Pratto, F., Sidanius, J., Stallworth, L. M., & Malle,B. F. (1994). Social dominance orientation: Apersonality variable predicting social and politicalattitudes. Journal of Personality and Social Psychology,67, 741–763.

    Quinn, D. M., & Crocker, J. (1999). When ideologyhurts: Effects of belief in the Protestant ethic andfeeling overweight on the psychological well-beingof women. Journal of Personality and SocialPsychology, 77, 402–414.

    Richeson, J., & Nussbaum, R. J. (2004). The impactof multiculturalism versus color-blindness onracial bias. Journal of Experimental Social Psychology,40, 417–423

    Ross, L. & Nisbett, R. E. (1991). The personal and thesituation. New York: McGraw-Hill.

    Ruble, D. N. (1994). A phase model of transitions:Cognitive and motivational consequences. In M. Zanna (Ed.), Advances in experimental socialpsychology (pp. 163–214). New York: Academic Press.

    Schofield, J. W. (1986). Causes and consequences ofthe colorblind perspective. In J. F. Dovidio & S. L.Gaertner (Eds.), Prejudice, discrimination, and racism(pp. 231–253). Orlando, FL: Academic Press.

    Sidanius, J., & Pratto, F. (1999). Social dominance: Anintergroup theory of social hierarchy and oppression.New York: Cambridge University Press.

    Somerman, F. B. (1993). Value, attitude, and beliefdeterminants of willingness to accept a facility forthe homeless. Journal of Social Distress and theHomeless, 2, 177–192.

    Wegener, D. T., & Petty, R. E. (1998). The naivescientist revisited: Naive theories and socialjudgment. Social Cognition, 16, 1–7.

    Whitley, B. E., Jr. (1999). Right-wingauthoritarianism, social dominance orientation,and prejudice. Journal of Personality and SocialPsychology, 77, 126–134.

    Wolsko, C., Park, B., Judd, C. M., & Wittenbrink, B.(2000). Framing interethnic ideology: Effects ofmulticultural and color-blind perspectives on

    Group Processes & Intergroup Relations 9(1)

    114

  • judgments of groups and individuals. Journal ofPersonality and Social Psychology, 78, 635–654.

    Yzerbyt, V., Leyens, J-P., & Corneille, O. (1998).Social judgeability and the bogus pipeline: Therole of naïve theories of judgment in impressionformation. Social Cognition, 16, 56–77.

    Paper received 28 October 2004; revised version accepted24 June 2005.

    Biographical notesSHERI R. LEVY is an assistant professor of psychology

    at the State University of New York at Stony Brook.Her research interests include understandingprocesses that contribute to prejudice and the roleof people’s lay theories in intergroup relations.

    TARA L. WEST completed her doctoral degree insocial psychology at the State University of NewYork at Stony Brook. She is currently pursuing alaw degree at New York University. Her research

    interests include increasing pro-social attitudesand behaviors and understanding and reducingprejudice.

    LUISA RAMIREZ is a doctoral student in thedepartment of psychology at the State Universityof New York at Stony Brook. She earned her BA inPsychology and MA in Political Science fromUniversidad de Los Andes (University of LosAndes). Her research interests includeunderstanding prejudice processes and theformation of collective identities in ethnicminority groups.

    DINA M. KARAFANTIS completed her doctoraldegree in social psychology at the State Universityof New York at Stony Brook. She is currently anassistant professor of psychology at Borough ofManhattan Community College. Her researchinterests include understanding mechanisms thatcan promote positive intergroup attitudes andrelations among children and adults.

    Levy et al. dual implications

    115