Group Processes &Intergroup Relations

2004 Vol 7(1) 35–54

How do members of low status groupsrespond, socially and psychologically, to impu-tations of inferiority? This is an enduring andfundamental question for the psychology ofintergroup relations (e.g. Allport, 1954; Bobo &

GPIR

Copyright © 2004 SAGE Publications(London, Thousand Oaks, CA and New Delhi)

7:1; 35–54; DOI: 10.1177/1368430204039972

Author’s noteJohn T. Jost, Department of Psychology, NewYork University, New York, NY 10003, USA[email: john.jost@nyu.edu]

Resistant versus AcquiescentResponses to IngroupInferiority as a Function ofSocial Dominance Orientationin the USA and Italy

Jennifer R. OverbeckStanford University

John T. JostStanford University and Radcliffe Institute for Advanced Study, HarvardUniversity

Cristina O. MossoUniversity of Turin

Agnieszka FlizikStanford University

Social identity theory typically emphasizes how low status group members resist and challengeimputations of inferiority (Tajfel & Turner, 1979), whereas system justification theoryemphasizes the tendency to accept and justify status hierarchies ( Jost & Banaji, 1994). On thetheoretical assumption that responses to ingroup inferiority would vary according to individualdifferences in social dominance orientation (SDO; Pratto, Sidanius, Stallworth, & Malle, 1994),we predicted and found in two studies that low SDO members of low status groups engage inresistance and social competition, as social identity theory would predict, whereas high SDOmembers follow system justifying patterns of acquiescence and even active bolstering of thestatus quo. The fact that the studies were conducted in two cultures (USA and Italy) that differwith regard to hierarchical traditions and beliefs about social mobility increases thegeneralizeability of the results and strengthens the conclusion that SDO predicts acquiescent vs.resistant responses to ingroup inferiority.

keywords inferiority, intergroup relations, outgroup favoritism, power, socialdominance, social identity, status, system justification

03 GPI7-1 Overbeck (JB/D) 23/1/04 9:54 am Page 35

at UNIV OF SOUTHERN CALIFORNIA on May 23, 2011gpi.sagepub.comDownloaded from

Hutchings, 1996; Brewer, 1979; Crocker, Major,& Steele, 1998; Jackman, 1994; Major et al.,2002; Ridgeway, Diekema, & Johnson, 1995;Sidanius & Pratto, 1999; Tajfel, 1981). It is fairlyeasy to see why low status group memberswould seek to resist and challenge the statushierarchy, even in subtle and indirect ways, inorder to preserve self-esteem and positivegroup identification (e.g. Hogg & Abrams,1988; Hornsey, Spears, Cremers, & Hogg, 2003;Spears, Jetten, & Doosje, 2001; Tajfel & Turner,1979). Indeed, the research literature on inter-group relations has examined this issue in con-siderable detail (e.g. Brewer & Miller, 1996).

Less intuitive perhaps is the idea thatmembers of low status groups might accept andeven actively and willingly support the veryhierarchical system that places them at adisadvantage (e.g. Jost & Banaji, 1994; Jost,Pelham, & Carvallo, 2002). Although it mayseem surprising from the standpoint of theoriesthat stress self-interest and the drive for positivegroup distinctiveness, the acquiescent responsehas been widely observed in human society(e.g. Biko, 1978; Della Fave, 1980; Elster, 1982;Fanon, 1963; Glazer, 2002; Jost, 1995; MacKin-non, 1989; Marx & Engels, 1846/1970). Manywomen, for example, accept and perpetuategender inequality, internalize a diminishedsense of self-worth, and subscribe to stereotypesthat justify sexism against women (Glick &Fiske, 2001; Jackman, 1994; Major, 1994). Manylow-income workers defend and justify con-servative economic policies that maintain thestatus quo while rejecting proposed changesthat would directly benefit them (Kluegel &Smith, 1986; Lane, 1959; Stacey & Green,1971). In an even more extreme example,many Jews in Germany, Hungary, and else-where apparently urged one another to bequiet and go along with anti-Semitic policiesbefore and during World War II (Klemperer,1995). For some, manifestations of acquies-cence also included Jewish ‘self-hatred’(Allport, 1954; Lewin, 1941).

In general, resistance and acquiescence neednot be conceived of as mutually exclusiveresponses. Most members of low status groupsprobably engage in some resistance to the

status quo and some acceptance of it, possiblyeven at the same time (e.g. Glick & Fiske, 2001;Jackman, 1994). One intrapersonal conse-quence of being faced with a conflict betweenwhether to resist or acquiesce is attitudinalambivalence about one’s own group member-ship ( Jost & Burgess, 2000). It also seems likelythat internal conflicts will be resolved differ-ently by different people, so that some indi-viduals will resist imputations of inferiority,whereas others will acquiesce and even bolsterthe status quo. Recent research suggests thatsocial dominance orientation is a personalityvariable that may help to predict divergentresponses to being in a low status position(Sidanius & Pratto, 1999). In this article, wefocus on individual differences in predictingdifferent strategies for expressing resistant vs.acquiescent responses to group disadvantage.

Social identity theory and responsesto ingroup inferiorityAccording to Tajfel and Turner’s (1979) socialidentity theory, members of low status groupsmay accept their alleged inferiority, but onlyunder certain circumstances. Specifically, theyare expected to engage in outgroup favoritismwhen the system is perceived to be legitimateand stable (Turner & Brown, 1978), when inter-group comparisons are made on status relevantdimensions (van Knippenberg, 1978), and wheningroup identification is very low (Ellemers,Spears, & Doosje, 1997). By and large, narrativeand meta-analytic reviews have supported thenotion that these variables moderate the effectof low status on ingroup vs. outgroup favoritism(Bettencourt, Dorr, Charlton, & Hume, 2001;Hinkle & Brown, 1990; Hogg & Abrams, 1988;Mullen, Brown, & Smith, 1992).

Identity management strategiesThe main question addressed by social identitytheory in relation to low status groups is howthey respond to ‘negative or threatened identi-ties’ arising from their position in the hier-archy. Tajfel and Turner (1979) describedthree main possibilities. The first strategy forimproving one’s social identity is to exit one’s

Group Processes & Intergroup Relations 7(1)

36

03 GPI7-1 Overbeck (JB/D) 23/1/04 9:54 am Page 36

at UNIV OF SOUTHERN CALIFORNIA on May 23, 2011gpi.sagepub.comDownloaded from

group and attempt to join a higher statusgroup. Obviously, this option—which reflectsresistance to the imputation of individual (butnot necessarily group) inferiority throughexit—is only available when group boundariesare permeable and social mobility is possible(Ellemers, Wilke, & van Knippenberg, 1993;Lalonde & Silverman, 1994). The secondstrategy is called ‘social competition’, and itrefers to direct challenges made by members ofthe low status group against the alleged superi-ority of a higher status outgroup; thus, it mostclearly reflects resistance to allegations of infe-riority (e.g. Mummendey, Klink, Mielke,Wenzel, & Blanz, 1999). This strategy, it hasbeen suggested, is most likely when statusrelations are ‘insecure’, that is, when the statusquo is perceived as both illegitimate andunstable (Caddick, 1982; Tajfel & Turner, 1979;Turner & Brown, 1978).

The third strategy mentioned by Tajfel andTurner is referred to as ‘social creativity’, and itdescribes the tendency to seek out oppor-tunities for social comparisons that are favor-able toward the ingroup. One example occurswhen members of low status groups acknow-ledge their inferiority on status relevant dimen-sions like competence and intelligence butcompensate for it by expressing ingroupfavoritism on other, status irrelevant dimen-sions like warmth and friendliness (Ellemers &Van Rijswijk, 1997; Spears & Manstead, 1989;van Knippenberg, 1978). This strategy could beseen as involving either resistance or acquies-cence to the status quo, insofar as it allowsmembers of low status groups to affirm positivegroup distinctiveness without challenging inany way the original basis for inequality (Fiske,Cuddy, Glick, & Xu, 2002; Glick & Fiske, 2001;Jackman, 1994; Jost, 2001). Given that membersof low status groups typically exhibit strongingroup favoritism on status ‘irrelevant’ orcommunal traits, evidence of outgroupfavoritism on such traits would suggest the kindof generalized inferiority associated with systemjustifying responses.1 Thus, in the currentresearch we focused especially on status irrele-vant traits in seeking to understand individualdifferences in acquiescence and resistance.

Social mobility belief structuresThe choice of which identity-enhancingstrategy (individual mobility, social com-petition, or social creativity) members ofdisadvantaged groups follow is said to dependupon several things, including whether or notthere are social mobility beliefs that are preva-lent in society (Hogg & Abrams, 1988). Accord-ing to social identity theory, low status groupmembers are more likely to engage in thestrategy of individual mobility and less likely toengage in social competition in societies thatstress the possibility of social mobility than insocieties that do not (Tajfel & Turner, 1979).The USA is widely seen as possessing a culturalbelief system that strongly emphasizes prospectsfor upward mobility (e.g. Ho, Sanbonmatsu, &Akimoto, 2002), especially in comparison withEuropean countries (e.g. Glazer, 2002). In thecurrent research program, we conductedsimilar studies in the USA and Italy in part tocompare responses to ingroup inferiority incultures that differed significantly with respectto beliefs about social mobility.

System justification theory andresponses to ingroup inferioritySystem justification theory acknowledges thestrength and operation of ego justificationmotives to protect and enhance individual self-esteem and group justification motives to protectand enhance positive group distinctiveness( Jost & Banaji, 1994). However, its uniquefocus is on the relatively neglected phenom-enon of system justification, defined as thetendency to protect and enhance the perceivedlegitimacy of the status quo (e.g. Haines & Jost,2000; Jost, 2001; Jost et al., 2002; Jost, Pelham,Sheldon, & Sullivan, 2003). From this perspec-tive, members of low status groups often exhibitoutgroup favoritism because they, likemembers of high status groups, are motivatedto justify and rationalize the status quo, at leastto some degree.

System justification and social identitytheorists are in agreement that sociostructuralvariables such as legitimacy and stability play animportant role in determining the responses of

Overbeck et al. ingroup inferiority and social dominance orientation

37

03 GPI7-1 Overbeck (JB/D) 23/1/04 9:54 am Page 37

at UNIV OF SOUTHERN CALIFORNIA on May 23, 2011gpi.sagepub.comDownloaded from

low status groups. Specifically, both theoriesgenerally imply that as the perceived legitimacyof the system increases, so, too, does outgroupfavoritism among low status groups (seeEllemers et al., 1993; Hornsey et al., 2003; Jost,2001; Jost & Burgess, 2000; Mummendey et al.,1999).2 However, the two theoretical perspec-tives differ in a number of other ways. Whereassocial identity theory emphasizes resistance bysuggesting that members of disadvantagedgroups will seek to challenge the status quo andassert their own positive distinctivenesswhenever possible (Spears et al., 2001), systemjustification theory emphasizes acquiescence byasserting that (to some degree) members of lowstatus groups will seek to preserve the legiti-macy and stability of the system by endorsingstereotypes and other social judgments thatjustify status differences ( Jost & Banaji, 1994).

Perceived magnitude of the status gapIf it is the case that people use stereotypes andother evaluations to explain and justify per-ceived status differences between groups (e.g.Jost, 2001), then it follows that larger statusdifferences should be accompanied by greateroutgroup favoritism on the part of low statusgroup members, in comparison with smallerstatus differences. This suggests the relevanceof an additional sociostructural variable notpreviously investigated by social identitytheorists, namely the perceived magnitude ofstatus differences between groups. Whereas anacquiescent response pattern would entail thatlow status group members would exhibitstronger outgroup favoritism as the magnitudeof the alleged status gap increases, a pattern ofresistance would mean the opposite. To theextent that they challenge inequality, as socialidentity theory would predict, then they shouldexhibit stronger ingroup favoritism as themagnitude of the status gap increases.

Threats to the status quoAccording to social identity theory, members oflow status groups should be expected to takeadvantage of threats to the perceived stability(or security) of the status quo and use the oppor-tunity to advance their own cause and challenge

the inequality by engaging in social competitionand ingroup favoritism (e.g. Turner & Brown,1978). System justification theory, by contrast,suggests that a threat to the stability of the statusquo should (at least under some conditions)induce a defensive response whereby membersof low status groups should express increasedoutgroup favoritism in order to bolster thewaning legitimacy and/or stability of the system(e.g. Jost & Hunyady, 2002). In other words, atleast some members of disadvantaged groupsare expected to reject opportunities for socialchange in favor of maintaining and enhancingthe status quo (see also Haines & Jost, 2000; Jost& Burgess, 2000; Jost et al., 2002, 2003). Theindividual difference variable of social domi-nance orientation may help to predict andexplain which members of a low status group willacquiesce and which will tend to resist theiralleged inferiority (Pratto et al., 1994; Sidanius& Pratto, 1999), thereby reconciling divergingaccounts derived from theories of social identifi-cation and system justification.

Social dominance theory andresponses to ingroup inferiorityAccording to social dominance theory, thereare important individual differences in the waysin which people respond to group-basedsystems of inequality (e.g. Sidanius & Pratto,1999). Specifically, Pratto et al. (1994) devel-oped and validated a measure of social domi-nance orientation (SDO) that included itemssuch as the following: ‘Inferior groups shouldstay in their place’, ‘To get ahead in life, it issometimes necessary to step on other groups’,and ‘We would have fewer problems if wetreated people more equally’ (reverse-scored).Jost and Thompson (2000) argued that theSDO scale contained items that captured groupjustification and system justification, making itdifficult to determine whether high SDOmembers of low status groups would engage inresistance or acquiescence to the status quo.

Alternative interpretations of SDOThe interpretation of SDO favored by Alte-meyer (1998) is much closer to group

Group Processes & Intergroup Relations 7(1)

38

03 GPI7-1 Overbeck (JB/D) 23/1/04 9:54 am Page 38

at UNIV OF SOUTHERN CALIFORNIA on May 23, 2011gpi.sagepub.comDownloaded from

justification (and, in the context of low statusgroups, resistance) than it is to system justifi-cation (and acquiescence). He argued, forinstance, that ‘High SDO’s . . . see life as “dogeat dog” and—compared with most people—are determined to do the eating’ (p. 75).According to Altemeyer, SDO predicts aggres-sive desires to punish and humiliate outgroupmembers and to ‘become the alpha animal’ (p.87). This account differs considerably fromrecent interpretations offered by social domi-nance theorists, who now stress system justifi-cation (and acquiescence to the status quo)rather than group justification. For example,Sidanius, Levin, Federico, and Pratto (2001)clarified that:

While SDO has been conceptualized previously asa desire for ingroup domination over outgroups(e.g. Sidanius & Pratto, 1993), more recent workconceptually and operationally defines SDO interms of a more general desire for unequal relationsamong social groups, regardless of whether this meansingroup domination or ingroup subordination. (p. 312,emphasis added)

This interpretation is also consistent with thework of Jost and Burgess (2000), who arguedthat SDO is an individual difference variablethat measures variation in system justificationtendencies.

Theoretically, then, adopting an individualdifference perspective should help to predictwhich people will choose resistance and whichwill choose acquiescence. More specifically, theSDO scale should identify which groupmembers will respond to a position of inferior-ity by accepting and perpetuating status differ-ences through the use of stereotypes andevaluations, and which group members willchallenge and defy those differences. On theassumption that SDO (as a single, generalfactor) is closer to system justification thangroup justification, we hypothesized that,compared with those who are low in SDO,members of low status groups who score highon SDO should be more responsive to themagnitude of status differences between theirgroup and others, and therefore show strongeroutgroup favoritism as the status gap widens. Inaddition, they may express outgroup favoritism

as a way of enhancing the legitimacy and/orstability of the system when either of these arethreatened or in doubt. In other words, lowSDO respondents were expected to resist thestatus quo, whereas high SDO respondentswere expected to evince patterns of acquies-cence and even bolstering.

Past research on SDO and ingroup inferioritySome previous studies support the hypothesisthat members of low status groups who are highin SDO would be more sensitive to the per-ceived legitimacy and stability of the statusdifferences in comparison with group memberswho are low in SDO (e.g. Federico, 1999).Rabinowitz (1999), for instance, found thatmembers of racial and ethnic minority groupsin the US were most likely to oppose hierarchy-attenuating social policies (i.e. to acceptinequality) when they were both high in SDOand low in perceived injustice against their owngroup (i.e. high in perceived system legiti-macy). Similarly, Levin, Federico, Sidanius, andRabinowitz (2002) found that for members oftwo very different disadvantaged groups (Arabsliving in Israel and Latinos living in the US)SDO interacted with perceived legitimacy topredict levels of outgroup favoritism. Whereaslow status group members who were high inSDO tended to show more outgroup favoritismwhen perceived legitimacy was high than whenit was low, group members who were low inSDO did not. None of these studies investigatedthe perceived magnitude of status differences,and none demonstrated conclusively that lowSDO group members actively resist imputationsof inferiority or that high SDO group membersembrace them, but the results of past researchare at least consistent with these possibilities.

Overview of researchIn two studies reported here, high and low SDOscorers were required to draw comparisonsbetween their own group and a higher statusoutgroup. In the first study, students at the Uni-versity of California at Santa Barbara were pre-selected on the basis of their SDO scores andrecruited to participate in an experiment in

Overbeck et al. ingroup inferiority and social dominance orientation

39

03 GPI7-1 Overbeck (JB/D) 23/1/04 9:54 am Page 39

at UNIV OF SOUTHERN CALIFORNIA on May 23, 2011gpi.sagepub.comDownloaded from

which they evaluated their own group in com-parison with students at Stanford University,which is a higher status private university. In thesecond study, respondents from Sicily, in thesouth of Italy, completed a translated version ofthe SDO scale in addition to questionnairesabout the characteristics of Southern Italiansand Northern Italians, the latter of whom enjoyhigher social and economic status (e.g.Capozza, Bonaldo, & DiMaggio, 1982). Byinvestigating these two different social andcultural contexts, we were able to compareresults from a country that strongly emphasizessocial mobility (USA) and one that does not(Italy).

Because past research indicates thatmembers of low status groups frequentlyexhibit outgroup favoritism on status relevantcharacteristics and ingroup favoritism on statusirrelevant characteristics (Bettencourt et al.,2001; Mullen et al., 1992), research participantswere given the opportunity to draw intergroupcomparisons on both relevant (competence-related) and irrelevant (communal) dimen-sions. However, we expected that there wouldbe a wider range of responses for the lattercomparisons and that such responses would bemore theoretically interesting. Specifically, wefocused on communal (status irrelevant) traitson the assumption that outgroup favoritism onsuch traits would indicate generalized inferior-ity associated with system justifying responses,whereas outgroup favoritism on status relevantor competence traits would not necessarilydo so.

Respondents were asked about their percep-tions of the magnitude of the status differences,the legitimacy and stability of the status differ-ences, and the degree to which they identifiedwith their own group, because these variableshave been found to predict levels of ingroupand outgroup favoritism on the part oflow status groups (e.g. Ellemers et al., 1993;Turner & Brown, 1978), and because there aretheoretical reasons to think that high SDOpeople would be more sensitive than lowSDO people to the magnitude, legitimacy,and stability of status differences betweengroups.

Study 1MethodPre-selection procedure In group prescreen-ing sessions lasting approximately one hour,156 undergraduate students at the University ofCalifornia at Santa Barbara completed the 16-item ‘social dominance orientation’ scale vali-dated by Pratto et al. (1994), along withbatteries of measures that were used by otherresearchers.

Sampling procedure Several weeks later,experimental research participants weresampled from the quartile that scored highestand the quartile that scored lowest on the SDOscale. Potential research participants were con-tacted by telephone and asked to participate inan experiment in exchange for partial fulfill-ment of a course requirement. No mention wasmade of the SDO scale administered during theprescreening sessions. Approximately half ofthose people contacted actually participated infollow-up sessions, resulting in a sample of 19high SDO scorers and 19 low SDO scorers.

Experimental procedure During the actualexperimental sessions, which occurred at least6 weeks after the initial prescreening sessions,UCSB students were first asked to evaluate aseries of ingroup and outgroup thought-lists;this task reinforced the cover story (that thestudy was about abstract verbal reasoning) andhas been used in prior research (e.g. Jost, 2001;Jost & Burgess, 2000). Following the thought-list evaluation task, research participants wereasked to evaluate their ingroup and anoutgroup (Stanford University students) on aseries of status relevant and status irrelevantdimensions of comparison. Finally, they wereasked to estimate the magnitude of social statusdifferences between UCSB and Stanfordstudents, to assess the legitimacy of those differ-ences, and to indicate their level of identifi-cation with the ingroup of UCSB. students.

MaterialsIngroup and outgroup ratings In order toassess ingroup and outgroup evaluations, the

Group Processes & Intergroup Relations 7(1)

40

03 GPI7-1 Overbeck (JB/D) 23/1/04 9:54 am Page 40

at UNIV OF SOUTHERN CALIFORNIA on May 23, 2011gpi.sagepub.comDownloaded from

students were asked to complete a series ofLikert-type scales that have been used in priorresearch ( Jost, 2001; Jost & Burgess, 2000). Theitems of interest for this paper were six attrib-utes considered status irrelevant (honest, dis-honest, interesting, uninteresting, friendly,unfriendly). For balance, and in order todemonstrate a lack of contingent response onthe items, six attributes considered statusrelevant (intelligent, unintelligent, hard-working, lazy, and skilled and unskilled atverbal reasoning) were also included. Eachrating was made on a scale ranging from 0(labeled ‘Not at all’) to 9 (‘Extremely’), andrespondents evaluated the ingroup and theoutgroup on each trait independently.Ingroup/outgroup favoritism scores were calcu-lated separately for status-irrelevant and status-relevant traits by subtracting outgroup ratingsfrom ingroup ratings for each cluster.

Perceived magnitude of status differencesImmediately following the stereotyping task,respondents were asked about the extent towhich they believed that there is a status differ-ence between UCSB and Stanford Universitystudents. Specifically, the item read: ‘Somepeople think that Stanford and UCSB differ interms of social status. We are interested in youropinion. How much greater or less do you thinkStanford students’ social status is compared tothat of UCSB students?’ Responses were madeby circling a number on a 15-point scale withlabels ranging from ‘Much less’ (1–3) to‘Somewhat less’ (4–6) to ‘Neither greater norless’ (7–9) to ‘Somewhat greater’ (10–12) to‘Much greater’ (13–15).

Perceived legitimacy of status differences Tokeep the questionnaire brief and to avoidarousing suspicion and reactance on the part ofresearch participants, we used a single-itemindicator of perceived legitimacy. Specifically,participants were asked: ‘In general, how legiti-mate or illegitimate do you think the statusdifferences are between UCSB and Stanford?’Responses were made on a 15-point scalelabeled from ‘Extremely illegitimate’ (1–3) to‘Extremely legitimate’ (13–15).

Because methodological questions can beraised about the validity of a single-itemmeasure, we conducted a pretest using an inde-pendent sample of 127 students who respondedto multiple items assessing the legitimacy andstability of socioeconomic differences betweenNortherners and Southerners in the USA.Legitimacy items included: ‘Overall, how fair orunfair do you think it is that there are differ-ences between Northerners and Southerners interms of socioeconomic success?’; ‘Howstrongly do you agree or disagree that North-erners are entitled to be better off than South-erners?’; and ‘How justifiable or unjustifiabledo you think the differences are betweenNortherners and Southerners in terms ofsocioeconomic success?’ Items were drawnfrom several sources, including Mummendey etal. (1999), Jost and Burgess (2000), andSchmader, Major, Eccleston, and McCoy(2001). The seven legitimacy items formed avery reliable scale, ! = .80. The item we used inthe main study (‘In general, how legitimate orillegitimate do you think the status differencesare . . .’) was highly correlated with the overallscale, item-whole r(126) = .62, p < .01. Theseresults from the pretest suggest that our choiceof a single legitimacy item was an appropriateone.

Perceived identification with the ingroupFinally, respondents were asked ‘How stronglydo you think you identify with the group ofUCSB students?’ Responses again were madeon a 15-point scale labeled from ‘Not at all’(1–3) to ‘Very strongly’ (13–15).

ResultsCheck on perceived status difference Asexpected, UCSB students perceived their uni-versity to be lower in social status than StanfordUniversity. The mean rating of Stanford’s statusrelative to UCSB’s for the sample was 9.00,which differed significantly from 8, the neutralmidpoint of the scale (t(37) = 1.94, p < .03, one-tailed). Thus, it appears that the experimentdid involve a situation in which the ingroupsubjectively occupied a position of relativeinferiority.

Overbeck et al. ingroup inferiority and social dominance orientation

41

03 GPI7-1 Overbeck (JB/D) 23/1/04 9:54 am Page 41

at UNIV OF SOUTHERN CALIFORNIA on May 23, 2011gpi.sagepub.comDownloaded from

Individual and structural effects on ingroupand outgroup favoritism A general linearmodel was constructed in which ingroup/outgroup favoritism (calculated as ratings ofthe ingroup minus ratings of the outgroup) onstatus irrelevant traits3 was predicted by onecategorical variable, SDO level (high vs. low);three continuous variables, perceived magni-tude of status differences, perceived systemlegitimacy, and ingroup identification; and allresulting 2-way and 3-way interactions. Variableswere recentered at 0 prior to conductingregression analyses.

In order to interpret interaction effects, wegraphed the regression functions by identifyingsimple effects of one predictor variable oningroup favoritism at various specific levels ofthe interacting predictor variables. These levelswere defined either by group (for SDO) or as±1 standard deviation from the grand mean ofthe predictor (for continuous variables). Thesegraphs should be seen as illustrative of howinteractive relationships differ at various levelsof the predictor variables, but not as compre-hensively depicting all simple relationships.

The main prediction for this study was thatSDO, legitimacy, and magnitude of the status gapwould interact in predicting ingroup favoritism.Specifically, high SDO group members wereexpected to acquiesce to allegations of inferior-ity by showing strong outgroup favoritism (evenon status irrelevant traits), especially when statusdifferences were perceived to be large in magni-tude and legitimate. This is because a wide statusgap implies that there is more inequality tojustify, and legitimacy implies that the system isjust and should be supported. Low SDO partici-pants, on the other hand, were expected toresist their state of disadvantage by showingstrong ingroup favoritism, particularly whenstatus differences are large and illegitimate.From the perspective of resistance (as opposedto acquiescence), large status differences andillegitimate status differences are especially inneed of being contested.

These predictions were generally supported.Consistent with past research, a main effect oflegitimacy indicated that participants showedmore outgroup favoritism (on status-irrelevant

traits) as perceived legitimacy increased (F(1,24) = 9.00, p < .01). In addition, a significantmain effect of identification demonstrated thatmore highly identified UCSB students showedmore ingroup favoritism (F(1, 23) = 10.05, p <.01). The analysis also yielded a 3-way inter-action among SDO, legitimacy, and magnitudein predicting ingroup/outgroup favoritism(F(1, 24) = 7.84, p < .02). To interpret the inter-action, we examined separately the 2-wayinteractions between legitimacy and magnitudewithin each of the two levels of SDO. Theoverall pattern is depicted in Figure 1.

For those high SDO participants who per-ceived a larger status gap, outgroup favoritismwas inversely related to legitimacy (simple F(1,24) = 7.40, p = .01). That is, high SDO groupmembers seem to have bolstered the sagginglegitimacy of the status quo by expressingoutgroup favoritism when status differenceswere perceived as large. For low SDO partici-pants, the simple 2-way interaction of legitimacyand magnitude failed to attain significance(F(1, 24) = 2.02, p = .17). Nevertheless, largerperceived status differences were associatedwith stronger ingroup favoritism for this groupwhen legitimacy was low (F(1, 24) = 3.10,p = .09) (tested at "1 SD on legitimacy). Theseresults suggest that people who are low in SDOare more likely to exhibit the pattern of resist-ance expected on the basis of social identitytheory.

We also conducted an analysis that includedthe 4-way interaction of SDO, legitimacy, magni-tude, and group identification to determinewhether the observed effects depended onparticipants’ levels of identification with thegroup of UCSB students. The 4-way interactionwas nonsignificant (F(1, 23) = 0.03, p = .88).Identification did not interact with any of the vari-ables in predicting ingroup/outgroup favoritism.

DiscussionIn the first study, we obtained some supportfor predictions derived from social identity,system justification, and social dominancetheories. Ingroup identification was associatedwith ingroup favoritism on irrelevant traits.

Group Processes & Intergroup Relations 7(1)

42

03 GPI7-1 Overbeck (JB/D) 23/1/04 9:54 am Page 42

at UNIV OF SOUTHERN CALIFORNIA on May 23, 2011gpi.sagepub.comDownloaded from

Perceived legitimacy was associated withoutgroup favoritism on both relevant andirrelevant traits. These findings are consistentwith established research on social identitytheory (e.g. Bettencourt et al., 2001).

At the same time, additional findings suggestthat the ‘story’ of low status group members’responses to inferiority is somewhat morecomplex. When participants’ individual differ-ences in SDO were taken into account, it wasclear that low status group members did not allrespond the same. The most provocativefinding from the first study was that the effectsof structural variables (perceived magnitudeand perceived legitimacy) on ingroupfavoritism depended on SDO. Specifically,people who scored high on SDO manifested anincrease in outgroup favoritism as the per-ceived magnitude of the status differencesincreased and perceived legitimacy decreased.That is, they appear to have bolstered thewaning legitimacy of large (but not small)status differences by asserting the superiority of

the outgroup even on status irrelevant dimen-sions.

A second study was conducted in Italy tofurther examine the possibility that SDO isassociated with system justifying patterns ofresponse, most especially the tendency to useingroup and outgroup ratings in a manner thataccepts and perpetuates status differencesbetween groups. The opportunity to comparedata from two different cultural contexts wasparticularly appealing given that the USA iswidely assumed to possess a strong socialmobility belief structure, especially in compari-son with Europe (e.g. Ho et al., 2002; Hofstede,1997). The playwright David Mamet (2001)observed that American ‘national culture isfounded very much on the idea of strive andsucceed. . . . Instead of rising with the masses,one should rise from the masses’. Glazer (2002)reviewed findings from the World ValuesSurvey, including the fact that 71% of Americanssurveyed believed that the poor have a goodchance to work their way out of poverty,

Overbeck et al. ingroup inferiority and social dominance orientation

43

–6

–4

–2

0

2

4

6

8

10

0 2 4 6 8 10 12 14

Magnitude of status differences

Ing

rou

p/o

utg

rou

p f

avo

riti

sm

Low SDO/Low legit

Low SDO/High Legit

High SDO/Low legit

High SDO/High Legit

Figure 1. Study 1: Ingroup/outgroup favoritism as a function of SDO, perceived legitimacy, and perceivedmagnitude.

03 GPI7-1 Overbeck (JB/D) 23/1/04 9:54 am Page 43

at UNIV OF SOUTHERN CALIFORNIA on May 23, 2011gpi.sagepub.comDownloaded from

whereas only 40% of European respondentsshared this belief.

Tajfel and Turner’s (1979) influentialanalysis suggests that these two differentsocietal contexts might elicit different types ofresponses, given the different assumptionsabout upward mobility. According to socialidentity theory, members of disadvantagedgroups in societies that emphasize socialmobility are more likely to resist the conse-quences of inequality through individualmobility, whereas people in societies that de-emphasize social mobility are more likely toengage in direct social competition. On theother hand, obtaining consistent results acrossquite different cultures would augment thegeneralizeability of any conclusions drawnabout the role of SDO in predicting differentialresponses to ingroup inferiority.

Three additional changes to the basic pro-cedure were made in designing Study 2. First,we dropped the identification variable, becauseit did not qualify any of the effects of interest.Second, in order to facilitate comparisons withprevious studies assessing social identity theory(e.g. Turner & Brown, 1978), we added ameasure of perceived stability of the system.Finally, we also increased the sample size tomaximize our chances of detecting reliabledifferences.

Study 2MethodResearch participants Eighty-eight studentsenrolled at the University of Catania on theSouthern Italian island of Sicily participated inthe study.

Materials and procedure All research partici-pants completed (1) evaluations of NorthernItalians and Southern Italians on a series ofstatus irrelevant and status relevant traits, (2)items measuring the perceived magnitude ofsocial and economic status differences betweenNorthern Italians and Southern Italians, thelegitimacy of those differences, and the likeli-hood that these differences would change inthe future, and (3) an Italian translation of the

16-item SDO scale. The order in which (1) and(2) were administered was counterbalanced.The SDO scale in (3) was translated and back-translated by bilingual Italian-English speakers,and the resulting scale exhibited sufficientlyhigh reliability (! = .79). Because the SDO scalewas administered along with the othermeasures, we were able to retain the continu-ous scale values for analysis.

Ingroup and outgroup ratings Researchparticipants again completed a series of stereo-typing scales. In this study, four status-irrelevant(honest, interesting, friendly, happy) traits wereexamined. We also included four status-relevant traits (intelligent, hard-working, pro-ductive, efficient). Each rating was made on ascale ranging from 0 (labeled ‘Not at all’) to 9(‘Extremely’), and respondents evaluated theingroup and the outgroup on each trait inde-pendently.

Perceived magnitude and legitimacy of statusdifferences These items were identical tothose used in Study 1, except that they wererephrased to apply to perceptions of statusdifferences between Northerners and South-erners in Italy.

Perceived stability of the status differences Asin Study 1, we used single-item indicators tokeep the questionnaire brief and to avoid sus-picion. On a 15-point scale labeled from‘Extremely unlikely to change’ (1–3) to‘Extremely likely to change’ (13–15) respon-dents indicated how stable they perceived thestatus differences between the two groups to be.

To assess the validity of using a single-itemindicator, we administered a five-item scale ofperceived stability to the same pre-test sampleof 127 students described in the first study.Some items were drawn from Mummendey etal. (1999), and the scale included items such as:‘How strongly do you agree or disagree that therelationship between Northerners andSoutherners will remain stable for the nextseveral years?’ and ‘How strongly do you agreeor disagree that the current relationshipbetween Northerners and Southerners is just

Group Processes & Intergroup Relations 7(1)

44

03 GPI7-1 Overbeck (JB/D) 23/1/04 9:54 am Page 44

at UNIV OF SOUTHERN CALIFORNIA on May 23, 2011gpi.sagepub.comDownloaded from

temporary?’ The five stability items formed anadequately reliable scale (! = .63). The itemselected for use in the main study was highlycorrelated with the overall scale (item-wholer(126) = .45, p < .01).

ResultsCheck on perceived status difference Con-sistent with expectations, Southern Italians per-ceived their group to be lower in social status incomparison with Northern Italians. The meanrating of Southern Italy’s status relative to theNorth for the sample was 5.89, which differedsignificantly from 8, the neutral midpoint of thescale (t (87) = "8.55, p < .001, two-tailed). Thus,it appears that the experiment did involve asituation in which the ingroup subjectivelyoccupied a position of relative inferiority.

Individual and structural effects on ingroupand outgroup favoritism A general linearmodel was constructed in which ingroup/outgroup favoritism was predicted by continu-ous variables of SDO, perceived magnitude,perceived legitimacy, and perceived stability.4The model also included all 2-way, 3-way, and 4-way interactions. Variables were again recen-tered prior to analyses. The analysis yielded anumber of interesting effects.

As in the previous study, Southern Italiansshowed less ingroup favoritism (i.e. greateroutgroup favoritism) as perceived legitimacyincreased (F(1, 68) = 5.15, p = .03). In thepresent study, we also found that ingroupfavoritism increased on average as participants’SDO scores increased (F(1, 68) = 4.24, p = .04).This finding seems to reopen the questionraised by Jost and Thompson (2000) of whetherSDO is better conceptualized as group justifi-cation (Altemeyer, 1998) or system justification( Jost & Burgess, 2000; Sidanius et al., 2001).

A 2-way interaction between SDO and per-ceived legitimacy (F(1,68) = 10.56, p < .01)indicated that the participants higher in SDOtended to show greater outgroup favoritism assystem legitimacy increased, whereas thoselower in SDO tended to show greater ingroupfavoritism. These findings replicate patternsobserved by Levin et al. (2002).

A 2-way interaction between SDO and per-ceived stability was also obtained (F (1,68) =9.00, p < .01). For people who scored lower onSDO, greater instability was associated withincreased ingroup favoritism. This is consistentwith the notion offered by social identitytheorists that low status groups might take theopportunity to engage in ‘social competition’against higher status groups when statusrelations are ‘insecure’ (e.g. Caddick, 1982;Tajfel & Turner, 1979; Turner & Brown, 1978).For people who scored higher on SDO,however, instability was associated withdecreased ingroup favoritism (or increasedoutgroup favoritism). This finding is in contra-diction to a result reported by Federico (1999),but it is consistent with the system justificationargument that at least some members of lowstatus groups would rather maintain than chal-lenge the existing hierarchy in the face of athreat to the stability of the system.

The analysis yielded a 2-way interactioninvolving perceived legitimacy and the magni-tude of status differences (F (1,68) = 6.86, p =.01). Increased legitimacy was associated withdecreased ingroup favoritism (or increasedoutgroup favoritism) on irrelevant traits whenthe status gap was perceived as small in magni-tude but not when it was perceived as large.Finally, we also observed a significant 2-wayinteraction between SDO and the magnitude ofstatus differences (F(1, 68) = 5.34, p = .02), suchthat increasing status differences led to greateringroup favoritism on the part of lower-SDOrespondents and greater outgroup favoritismon the part of higher-SDO respondents. Thisfinding also suggests that low SDO scores areassociated with resistance, whereas high SDOscores are associated with acquiescent responsepatterns.

As in Study 1, a 3-way interaction involvingSDO, legitimacy, and magnitude was found topredict ingroup/outgroup favoritism. Thedegree of ingroup favoritism (on status-irrele-vant traits) was strongly affected by respon-dents’ SDO levels and their perceptions of thelegitimacy and magnitude of status differences(F(1, 68) = 9.30, p < .01). The overall pattern ofresults is illustrated in Figure 2.

Overbeck et al. ingroup inferiority and social dominance orientation

45

03 GPI7-1 Overbeck (JB/D) 23/1/04 9:54 am Page 45

at UNIV OF SOUTHERN CALIFORNIA on May 23, 2011gpi.sagepub.comDownloaded from

Across SDO levels, increases in the perceivedstatus gap did not change ingroup favoritismfor those who saw the system as highly legiti-mate. Among lower-SDO respondents who sawthe system as less legitimate, such increases ledto greater ingroup favoritism. Among highSDO scorers, however, this pattern wasreversed: increases in the perceived magnitudeof the status differences was associated withdecreased ingroup favoritism for those who sawthe system as low in legitimacy. Thus, as inStudy 1, people who were low on SDO acceptedtheir low status position when the system wasperceived as legitimate, but they challenged it(on status irrelevant traits) when the system wasperceived as illegitimate, as predicted by socialidentity theory (Caddick, 1982; Tajfel &Turner, 1979; Turner & Brown, 1978). Peoplewho were high on SDO once again exhibitedstrong outgroup favoritism (again on irrelevant

traits) when status differences were perceivedas large in magnitude and yet lacking in clearlegitimacy. Thus, it appears that they drewintergroup comparisons that would actuallybolster the waning legitimacy of large (but notsmall) status differences; this bolstering strategyis a counterintuitive possibility suggested bysystem justification theory (e.g. Haines & Jost,2000; Jost et al., 2003).

Finally, the complex 4-way interaction amongSDO, legitimacy, stability, and magnitude wasalso significant (F(1, 68) = 5.20, p = .03). Thepattern of results is illustrated in Figure 3 bypresenting the above 3-way interaction of SDO,legitimacy, and magnitude at ±1 SD of stability.The 4-way interaction appears to be driven bythe responses of higher-SDO participants whenthe social system is both unstable and highlylegitimate. As the magnitude of status differ-ences increases, these participants do express

Group Processes & Intergroup Relations 7(1)

46

1 3 5 7 9 11 13 15

Low SDO/Low legitLow SDO/High legitHigh SDO/Low legitHigh SDO/High legit

Ing

rou

p/o

utg

rou

p f

avo

riti

sm

4

3

2

1

0

Magnitude of status differences

Figure 2. Study 2: Ingroup/outgroup favoritism as a function of SDO, perceived legitimacy, and perceivedmagnitude.

03 GPI7-1 Overbeck (JB/D) 23/1/04 9:54 am Page 46

at UNIV OF SOUTHERN CALIFORNIA on May 23, 2011gpi.sagepub.comDownloaded from

Overbeck et al. ingroup inferiority and social dominance orientation

47

Figure 3. Ingroup/outgroup favoritism as a function of stability, SDO, perceived legitimacy, and perceivedmagnitude.

03 GPI7-1 Overbeck (JB/D) 23/1/04 9:54 am Page 47

at UNIV OF SOUTHERN CALIFORNIA on May 23, 2011gpi.sagepub.comDownloaded from

more ingroup favoritism, which is consistentwith a social competition response. However,collapsing across levels of stability, high SDOparticipants who see the system as illegitimateare consistent in bolstering the system byfavoring the outgroup more as the magnitudeof the inequality grows.

It is worth noting in passing that none of theabove effects included the legitimacy # stabilityinteraction hypothesized by social identitytheorists to account for responses to ingroupinferiority (e.g. Tajfel & Turner, 1979; Turner& Brown, 1978). Rather, SDO interacted withlegitimacy, and it also interacted with stabilityand with magnitude.

DiscussionA number of differences between high and lowSDO members of a disadvantaged group(Southern Italians) emerged in Study 2. Per-ceived legitimacy was associated with outgroupfavoritism to a significant degree as SDOincreased, but not at lower levels of SDO. Theperceived stability of the system had oppositeeffects, depending upon SDO levels. Instabilitywas associated with increased ingroupfavoritism among those with lower SDO scores,as social identity theory would predict, and withincreased outgroup favoritism among thosewith higher SDO scores, as system justificationtheory would predict.

The 3-way interaction obtained in Study 1 wasreplicated in Study 2. Once again, perceivedmagnitude had opposite effects on ingroupfavoritism as a function of SDO when perceivedlegitimacy was low. Consistent with a strategy ofbolstering the status quo, higher-SDO scorersshowed increased outgroup favoritism as theperceived magnitude of the status differencesincreased when perceived legitimacy was low(but not high). Lower-SDO scorers, by contrast,showed increased ingroup favoritism as the per-ceived magnitude of the status differencesincreased under low legitimacy (but not high),consistent with a pattern of resistance. Thus,SDO was again found to moderate theresponses of low status group members to theirposition of assumed inferiority.

Conclusions

In two studies carried out in the US and Italy,we found that scores on Pratto et al.’s (1994)SDO scale predicted which members of a lowstatus group would acquiesce to and whichwould rebel against the status quo. Comparedto low SDO group members, high SDO groupmembers were more sensitive in general to theperceived magnitude and legitimacy of statusdifferences between groups, and they weremore likely to engage in outgroup favoringcomparisons, even on status-‘irrelevant’ dimen-sions of comparison that are assumed to beemployed in identity-enhancing social creativitystrategies (e.g. Brewer & Miller, 1996; Tajfel &Turner, 1979). People who were high in SDOappeared to accept and even bolster statusdifferences by defending the status quo againstwaning legitimacy and/or stability, whereaspeople who were low in SDO appeared to chal-lenge the status differences whenever theopportunity arose. These findings have directrelevance to the question raised by Jost andThompson (2000) of whether SDO is betterconceptualized as a desire for ingroup domi-nance—as suggested by Altemeyer (1998) andothers—or system justification, as suggested byJost and Burgess (2000) and Sidanius et al.(2001). Our results generally support the latterinterpretation.

It is theoretically relevant that almost alldifferences emerged on status-irrelevantdimensions of comparison, on which low statusgroup members are expected to ‘compensate’for identity threats on status-relevant dimen-sions (e.g. Mullen et al., 1992; Spears &Manstead, 1989; van Knippenberg, 1978).Outgroup favoritism among members of lowstatus groups on irrelevant dimensions there-fore satisfies a relatively strict criterion forbeing considered a bias against the ingroup(Brewer & Miller, 1996, p. 95). The fact thatmost of our effects of interest occurred onirrelevant traits is also consistent with theresults of a meta-analysis by Bettencourt et al.(2001), who found that sociostructural vari-ables such as legitimacy were more likely toaffect intergroup comparisons on irrelevant

Group Processes & Intergroup Relations 7(1)

48

03 GPI7-1 Overbeck (JB/D) 23/1/04 9:54 am Page 48

at UNIV OF SOUTHERN CALIFORNIA on May 23, 2011gpi.sagepub.comDownloaded from

traits. Specifically, they concluded that: ‘whenstatus differences were perceived as illegitimate,low-status groups asserted the positivity of theirin-group by means of comparisons on the irrel-evant dimensions’ (p. 533). Our studies suggestthat this is an apt description of low but nothigh SDO group members. Because status-irrel-evant traits are less consensually defined andtherefore more open to flexible and strategicuse, it appears that they can be used in con-junction with a variety of motives, whetherthose are group-serving, as social identitytheory suggests, or system-serving, as theoriesof system justification and social dominancewould suggest. As expected, our evidenceconfirms that responses to a position of inferi-ority are complexly determined.

It seems, therefore, that members of lowstatus groups are more likely to engage in socialcompetition against higher status groups(Tajfel & Turner, 1979) when they are lowrather than high in SDO. Conversely, they aremore likely to adopt acquiescent responsepatterns, as suggested by system justificationtheory ( Jost & Banaji, 1994), when they arehigh rather than low in SDO. Thus, the indi-vidual difference variable of social dominanceorientation helps to integrate theoretical pre-dictions that emphasize resistance to inequalitywith those that emphasize acquiescence. Thissuggests that although the three theories ofsocial identification, system justification, andsocial dominance differ in important ways, theyare to a considerable degree compatible, com-plementary, and mutually informative, as hasbeen argued recently by proponents of each(e.g. Jost & Burgess, 2000; Levin et al., 2002;Sidanius et al., 2001; Spears et al., 2001).

In distinguishing our findings from previousresearch on social identity theory, it is import-ant to note how contingent the effects of manysociostructural variables are on intergroupbehavior. For example, social identity theoryhas long predicted that group members whoare more identified with their groups will showstronger resistance to inferiority (e.g. Ellemerset al., 1997). Here, although we find thatidentification does predict ingroup favoritismon average, it fails to affect favoritism once

SDO, legitimacy, and the magnitude ofdisadvantage are taken into account. Socialidentity theory emphasizes that more legitimatesystems will lead to greater acquiescence,particularly when these systems are stable (e.g.Turner & Brown, 1978). Here, we find thatindividuals who vary in SDO respond differ-ently to inferiority, even when they share per-ceptions of the system as legitimate and stable.That is, these two structural variables, whileimportant and informative, do not provide anaccount of low status perceivers’ responses thatis as clean or complete as is often assumed (e.g.Brewer & Miller, 1996). Indeed, we found nohint of an interaction between perceived legiti-macy and stability in predicting ingroup andoutgroup favoritism, although this prediction isseen as almost axiomatic in social identitytheory (e.g. Bettencourt et al., 2001; Ellemerset al., 1993; Hogg & Abrams, 1988; Tajfel &Turner, 1979). In sum, the current findingssuggest that a much more complex picture ofresistance and acquiescence—including onethat takes into account individual differences inthe motivation to defend and justify the socialsystem—is necessary to account fully forresponses to low status. Such complexity is alsoconsistent with recent findings obtained byMummendey et al. (1999).

That an individual difference variable such asSDO could be an important determinant ofintergroup behavior is in some ways antitheticalto traditional assumptions of social identitytheory (and its successor, self-categorizationtheory) that patterns of ingroup and outgroupbias follow not from personality characteristicsbut from the structure of intergroup relationsand social conformity to prototypical ingroupnorms (e.g. Billig, 1976; Tajfel, 1981; Tajfel &Turner, 1979; Turner, Hogg, Oakes, Reicher, &Wetherell, 1987). Nevertheless, it does seemthat there are personality and individual differ-ence dimensions such as SDO and right-wingauthoritarianism that play a major role in deter-mining intergroup attitudes and behavior aswell as reactions to social systems and theirauthorities (e.g. Altemeyer, 1998; Sidanius &Pratto, 1999). The studies presented heresuggest that individual differences may help to

Overbeck et al. ingroup inferiority and social dominance orientation

49

03 GPI7-1 Overbeck (JB/D) 23/1/04 9:54 am Page 49

at UNIV OF SOUTHERN CALIFORNIA on May 23, 2011gpi.sagepub.comDownloaded from

explain why all group members do not respondin the same way to a situation of relative inferi-ority. This prospect should certainly receivemuch greater attention in future work.

A fairly obvious methodological limitation ofthe two studies reported here is that all of thefindings involving SDO, legitimacy, stability,and magnitude are correlational in nature.Consequently, the direction of causality amongvariables is unclear. Future research would dowell to manipulate the variables of interest inexperimental settings. Our results suggest thata particularly promising candidate for avariable worthy of experimental manipulationis that of the perceived magnitude of statusdifferences, insofar as it was found here to haveopposite effects on ingroup favoritism for highand low SDO respondents under certaincircumstances.

An additional theoretical perspective thatmay help to shed light on how low status groupmembers respond to inferiority is Hogg’s(2000) uncertainty reduction model, whichholds that individuals are motivated to reduceuncertainty because it creates psychological dis-comfort. Strategies for reducing uncertainty,according to Hogg, include seeking out socialsupport and group identification and perceiv-ing others in rigid terms that reinforce groupboundaries. This approach suggests that whenconfronted with information confirming thelow status of their group, people may responddifferently according to the strategies theychoose for reducing uncertainty. One optionmay be to recategorize the self as separate fromthe low status group and to seek individualmobility and the opportunity to join a higherstatus group. Another option would be for thegroup member to reduce uncertainty by per-ceiving the social environment in stronglygroup-based terms, aligning the self with thelow status group, and then resisting the statusquo by working to change the group’s standing(the social competition strategy). Thus, Hogg’s(2000) model is generally consistent with theidentity management strategies suggested byTajfel and Turner (1979), but the reduction ofuncertainty is seen as the driving motivationalforce.

There are two related points that seemrelevant to a discussion of these issues. One isthat important individual differences in theneed to reduce uncertainty certainly do exist,and the second is that these individual differ-ences in uncertainty reduction are predictive ofpolitically conservative and authoritarian atti-tudes (e.g. Jost, Glaser, Kruglanski, & Sulloway,2003; Sorrentino & Roney, 2000; Wilson, 1973).With regard to the results of studies reportedhere, it is possible that people who are high onSDO are also more likely to find uncertainty tobe aversive and therefore more eager to imposestructure on the social world by clinging tohierarchy, order, and rigid classification. Thus,the desire to manage uncertainty may be oneunderlying factor that leads to acquiescent,system justifying patterns of behavior. Futureresearch is needed to examine this possibilitymore directly.

In conclusion, we have argued that member-ship in a low status group leads different peopleto adopt different types of intergroup behavior.Specifically, some group members challengethe status quo, while others accept and perpetu-ate it. Social identity theory is especially well-suited to account for social competition andresistance to status inequality (e.g. Tajfel &Turner, 1979), whereas system justificationtheory has focused more on addressingpatterns of acquiescence and bolstering ( Jost &Banaji, 1994). Consistent with a perspectivistphilosophy of social science (McGuire, 1997),the present research suggests that both theoriesare right, despite their different emphases.Sidanius and Pratto’s (1999) concept of socialdominance orientation, it seems, helps us tounderstand which low status group memberswill follow the identity-enhancing strategiesdescribed so well by social identity theory andwhich individuals will instead opt for maintain-ing the legitimacy and stability of the statusquo, as emphasized by system justificationtheory.

Notes1. Brewer and Miller (1996) have argued that

outgroup favoritism among members of low status

Group Processes & Intergroup Relations 7(1)

50

03 GPI7-1 Overbeck (JB/D) 23/1/04 9:54 am Page 50

at UNIV OF SOUTHERN CALIFORNIA on May 23, 2011gpi.sagepub.comDownloaded from

groups ‘should probably not be labeled a “bias” atall’, because ‘this effect is found to occur almostexclusively on status-relevant dimensions ofevaluation’ (p. 95). By demonstrating that somemembers of low status groups exhibit outgroupfavoritism on irrelevant dimensions, then, we areshowing that their behavior satisfies Brewer andMiller’s criterion for being biased against theirown group.

2. Social identity and system justification theories dodiffer, however, with respect to the effects ofperceived (il)legitimacy on high status or highpower groups. According to social identity theory,members of advantaged groups should expressmore ingroup favoritism when intergrouprelations are seen as illegitimate and thereforeinsecure than when they are seen as legitimateand secure (Caddick, 1982; Hornsey et al., 2003;Turner & Brown, 1978). By contrast, systemjustification theory suggests that members ofadvantaged groups should express more ingroupfavoritism when their position is seen aslegitimate (and therefore justified andideologically defensible) than when it is seen asillegitimate and unjustified (Chen & Tyler, 2001;Jost, 2001; Jost & Hunyady, 2002).

3. We also included trait type as a factor in a GLManalysis in order to confirm that status relevantand irrelevant traits did indeed produce differentpatterns of favoritism. Because virtually all effectsoccur only among irrelevant traits, we report onlythese results. The four-way interaction of SDO #magnitude # legitimacy # trait type wassignificant (F(1, 24) = 4.28, p < .05), confirmingthat results were significantly different for the twotypes of traits. The sole reliable finding forrelevant traits was a simple effect of legitimacy: onaverage, participants showed greater outgroupfavoritism on relevant traits when they perceivedstatus differences to be more legitimate (F(1, 24)= 4.93, p = .04).

4. As in Study 1, we also conducted analyses thatincluded trait type as a factor. No significanteffects of the predictors were found on relevanttraits. The 4-way interaction of SDO # magnitude# legitimacy # trait type was significant (F(1, 68)= 6.25, p = .01), confirming that results weredifferent for the two types of traits.

ReferencesAllport, G. W. (1954). The nature of prejudice.

Cambridge, MA: Addison-Wesley.

Altemeyer, R. A. (1998). The other ‘authoritarianpersonality’. Advances in Experimental SocialPsychology, 30, 47–91.

Bettencourt, B. A., Dorr, N., Charlton, K., & Hume,D. L. (2001). Status differences and ingroup bias:A meta-analytic examination of the effects ofstatus stability, status legitimacy, and grouppermeability. Psychological Bulletin, 127, 520–542.

Biko, S. (1978). I write what I like. San Francisco:Harper & Row.

Billig, M. (1976). Social psychology and intergrouprelations. London: Academic Press.

Bobo, L., & Hutchings, V. L. (1996). Perceptions ofracial group competition: Extending Blumer’stheory of group position to a multiracial socialcontext. American Sociological Review, 61, 951–972.

Brewer, M. B. (1979). Ingroup bias in the minimalintergroup situation: A cognitive–motivationalanalysis. Psychological Bulletin, 86, 307–324.

Brewer, M. B., & Miller, N. (1996). Intergrouprelations. Buckingham, UK: Open University Press.

Caddick, B. (1982). Perceived illegitimacy andintergroup relations. In H. Tajfel (Ed.), Socialidentity and intergroup relations (pp. 137–154).Cambridge: Cambridge University Press.

Capozza, D., Bonaldo, E., & DiMaggio A. (1982).Problems of identity and social conflict: researchon ethnic groups in Italy. In H. Tajfel (Ed.), Socialidentity and intergroup relations (pp. 299–334).Cambridge: Cambridge University Press.

Chen, E. S., & Tyler, T. R. (2001). Cloaking power:Legitimizing myths and the psychology of theadvantaged. In A. Y. Lee-Chai & J. Bargh (Eds.),The use and abuse of power: Multiple perspectives on thecauses of corruption. Philadelphia: Psychology Press.

Crocker, J., Major, B., & Steele, C. M. (1998). Socialstigma. In D. T. Gilbert, S. T. Fiske, & G. Lindzey(Eds.), The handbook of social psychology (pp.504–553). Boston: McGraw-Hill.

Della Fave, L. R. (1980). The meek shall not inheritthe earth: Self-evaluation and the legitimacy ofstratification. American Sociological Review, 45,955–971.

Ellemers, N., Spears, R., & Doosje, B. (1997).Sticking together or falling apart: In-groupidentification as a psychological determinant ofgroup commitment versus individual mobility.Journal of Personality and Social Psychology, 72,617–626.

Ellemers, N., & Van Rijswijk, W. (1997). Identityneeds versus social opportunities: The use ofgroup-level and individual-level identitymanagement strategies. Social Psychology Quarterly,60, 52–65.

Overbeck et al. ingroup inferiority and social dominance orientation

51

03 GPI7-1 Overbeck (JB/D) 23/1/04 9:54 am Page 51

at UNIV OF SOUTHERN CALIFORNIA on May 23, 2011gpi.sagepub.comDownloaded from

Ellemers, N., Wilke, H., & van Knippenberg, A.(1993). Effects of the legitimacy of low group orindividual status on individual and collectivestatus-enhancement strategies. Journal of Personalityand Social Psychology, 64, 766–778.

Elster, J. (1982). Belief, bias, and ideology. In M.Hollis & S. Lukes (Eds.), Rationality and relativism(pp. 123–148). Oxford: Basil Blackwell.

Fanon, F. (1963). The wretched of the earth. New York:Grove Press.

Federico, C. M. (1999). The interactive effects ofsocial dominance orientation, group status, andperceived stability on favoritism for high-statusgroups. Group Processes and Intergroup Relations, 2,119–144.

Fiske, S. T., Cuddy, A. J. C., Glick, P., & Xu, J.(2002). A model of (often mixed) stereotypecontent: Competence and warmth respectivelyfollow from perceived status and competition.Journal of Personality and Social Psychology, 82,878–902.

Glazer, N. (2002, February 11). Why Americans don’tcare about income inequality. Paper presented at theJohn F. Kennedy School of Government, HarvardUniversity, Cambridge, MA.

Glick, P., & Fiske, S. T. (2001). An ambivalentalliance: Hostile and benevolent sexism ascomplementary justifications for genderinequality. American Psychologist, 56, 109–118.

Haines, E. L., & Jost, J. T. (2000). Placating thepowerless: Effects of legitimate and illegitimateexplanation on affect, memory, and stereotyping.Social Justice Research, 13, 219–236.

Hinkle, S., & Brown, R. (1990). Intergroupcomparisons and social identity: Some links andlacunae. In D. Abrams & M. A. Hogg (Eds.), Socialidentity theory: Constructive and critical advances(pp. 48–70). New York: Springer-Verlag.

Ho, E. A., Sanbonmatsu, D. M., & Akimoto, S. A.(2002). The effects of comparative status on socialstereotypes: How the perceived success of somepersons affects the stereotypes of others. SocialCognition, 20, 36–57.

Hofstede, G. H. (1997). Cultures and organizations:Software of the mind. New York: McGraw-Hill.

Hogg, M. A. (2000). Subjective uncertaintyreduction through self-categorization: Amotivational theory of social identity processes.European Review of Social Psychology, 11, 223–253.

Hogg, M. A., & Abrams, D. (1988). Socialidentifications: A social psychology of intergrouprelations and group processes. London: Routledge.

Hornsey, M. J., Spears, R., Cremers, I., & Hogg, M.A. (2003). Relations between high and low power

groups: The importance of legitimacy. Personalityand Social Psychology Bulletin, 29, 216–227.

Jackman, M. R. (1994). The velvet glove: Paternalismand conflict in gender, class, and race relations.Berkeley, CA: University of California Press.

Jost, J. T. (1995). Negative illusions: Conceptualclarification and psychological evidenceconcerning false consciousness. Political Psychology,16, 397–424.

Jost, J. T. (2001). Outgroup favoritism and thetheory of system justification: An experimentalparadigm for investigating the effects of socio-economic success on stereotype content. In G.Moskowitz (Ed.), Cognitive social psychology: ThePrinceton Symposium on the legacy and future of socialcognition (pp. 89–102). Hillsdale, NJ: Erlbaum.

Jost, J. T., & Banaji, M. R. (1994). The role ofstereotyping in system-justification and theproduction of false consciousness. British Journal ofSocial Psychology, 33, 1–27.

Jost, J. T., & Burgess, D. (2000). Attitudinalambivalence and the conflict between group andsystem justification motives in low status groups.Personality and Social Psychology Bulletin, 26,293–305.

Jost, J. T., Glaser, J., Kruglanski, A. W., & Sulloway, F.(2003). Political conservatism as motivated socialcognition. Psychological Bulletin, 129, 339–375.

Jost, J. T., & Hunyady, O. (2002). The psychology ofsystem justification and the palliative function ofideology. European Review of Social Psychology, 13,111–153.

Jost, J. T., Pelham, B. W., & Carvallo, M. (2002).Non-conscious forms of system justification:Cognitive, affective, and behavioral preferencesfor higher status groups. Journal of ExperimentalSocial Psychology, 38, 586–602.

Jost, J. T., Pelham, B. W., Sheldon, O., & Sullivan, B.N. (2003). Social inequality and the reduction ofideological dissonance on behalf of the system:Evidence of enhanced system justification amongthe disadvantaged. European Journal of SocialPsychology, 33, 13–36.

Jost, J. T., & Thompson, E. P. (2000). Group-baseddominance and opposition to equality asindependent predictors of self-esteem,ethnocentrism, and social policy attitudes amongAfrican Americans and European Americans.Journal of Experimental Social Psychology, 36, 209–232.

Klemperer, V. (1995). I will bear witness. New York:Random House.

Kluegel, J. R., & Smith, E. R. (1986). Beliefs aboutinequality: Americans’ view of what is and what oughtto be. Hawthorne, NJ: Aldine de Gruyter.

Group Processes & Intergroup Relations 7(1)

52

03 GPI7-1 Overbeck (JB/D) 23/1/04 9:54 am Page 52

at UNIV OF SOUTHERN CALIFORNIA on May 23, 2011gpi.sagepub.comDownloaded from

Lalonde, R. N., & Silverman, R. A. (1994).Behavioral preferences in response to socialinjustice: The effects of group permeability andsocial identity salience. Journal of Personality andSocial Psychology, 66, 78–85.

Lane, R. E. (1959). The fear of equality. AmericanPolitical Science Review, 53, 35–51.

Levin, S., Federico, C. M., Sidanius, J., & Rabinowitz,J. L. (2002). Social dominance orientation andintergroup bias: The legitimation of favoritism forhigh-status groups. Personality and Social PsychologyBulletin, 28, 144–157.

Lewin, K. (1941). Self-hatred among Jews. InResolving social conflicts: Selected papers on groupdynamics (pp. 186–200). New York: Harper &Brothers.

MacKinnon, C. A. (1989). Toward a feminist theory ofthe state. Cambridge, MA: Harvard University Press.

Major, B. (1994). From social inequality to personalentitlement: The role of social comparisons,legitimacy appraisals, and group memberships.Advances in Experimental Social Psychology, 26,293–355.

Major, B., Gramzow, R. H., McCoy, S. K., Levin, S.,Schmader, T., & Sidanius, J. (2002). Perceivingpersonal discrimination: The role of group statusand legitimizing ideology. Journal of Personality andSocial Psychology, 82, 269–282.

Mamet, D. (2001, January). Stagebill for AmericanConservatory Theatre production of GlengarryGlen Ross, San Francisco.

Marx, K., & Engels, F. (1846/1970). The Germanideology (C. J. Arthur, Ed.) New York: InternationalPublishers.

McGuire, W. J. (1997). Creative hypothesisgenerating in psychology: Some useful heuristics.Annual Review of Psychology, 48, 1–30.

Mullen, B., Brown, R., & Smith, C. (1992). Ingroupbias as a function of salience, relevance, andstatus: An integration. European Journal of SocialPsychology, 22, 103–122.

Mummendey, A., Klink, A., Mielke, R., Wenzel, M.,& Blanz, M. (1999). Sociostructural characteristicsof intergroup relations and identity managementstrategies: Results from a field study in EastGermany. European Journal of Social Psychology, 29,259–285.

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.

Rabinowitz, J. L. (1999). Go with the flow or fightthe power? The interactive effects of social

dominance orientation and perceived injustice onsupport for the status quo. Political Psychology, 20,1–24.

Ridgeway, C., Diekema, D., and Johnson, C. (1995).Legitimacy, compliance, and gender in peergroups. Social Psychology Quarterly, 58, 298–311.

Schmader, T., Major, B., Eccleston, C. P., & McCoy,S. K. (2001). Devaluing domains in response tothreatening intergroup comparisons perceivedlegitimacy and the status value asymmetry. Journalof Personality and Social Psychology, 80, 782–796.

Sidanius, J., Levin, S., Federico, C., & Pratto, F.(2001). Legitimizing ideologies: The socialdominance approach. In J. T. Jost & B. Major(Eds.), The psychology of legitimacy: Emergingperspectives on ideology, justice, and intergroup relations(pp. 307–331). New York: Cambridge UniversityPress.

Sidanius, J., & Pratto, F. (1993). The inevitability ofoppression and the dynamics of social dominance.In P. Sniderman, P. E. Tetlock, & E. G. Carmines(Eds.), Prejudice, politics, and the American dilemma(pp. 173–211). Stanford, CA: Stanford UniversityPress.

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

Sorrentino, R. M., & Roney, C. J. R. (2000). Theuncertain mind: Individual differences in facing theunknown. Philadelphia: Psychology Press/Taylor &Francis.

Spears, R., Jetten, J., & Doosje, B. (2001). The(il)legitimacy of ingroup bias: From social realityto social resistance. In J. T. Jost & B. Major (Eds.),The psychology of legitimacy: Emerging perspectives onideology, justice, and intergroup relations (pp.332–362). New York: Cambridge University Press.

Spears, R., & Manstead, A. S. R. (1989). The socialcontext of stereotyping and differentiation.European Journal of Social Psychology, 19, 101–121.

Stacey, B. G., & Green, R. T. (1971). Working-classconservatism: A review and an empirical study.British Journal of Social and Clinical Psychology, 10,10–26.

Tajfel, H. (1981). Human groups and social categories.Cambridge: Cambridge University Press.

Tajfel, H., & Turner, J. C. (1979). An integrativetheory of intergroup conflict. In W. G. Austin & S.Worchel (Eds.), The social psychology of intergrouprelations (pp. 33–47). Monterey, CA: Brooks-Cole.

Turner, J. C., & Brown, R. (1978). Social status,cognitive alternatives, and intergroup relations. InH. Tajfel (Ed.), Differentiation between social groups(pp. 201–234). London: Academic Press.

Overbeck et al. ingroup inferiority and social dominance orientation

53

03 GPI7-1 Overbeck (JB/D) 23/1/04 9:54 am Page 53

at UNIV OF SOUTHERN CALIFORNIA on May 23, 2011gpi.sagepub.comDownloaded from

Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher,S. D., & Wetherell, M. S. (1987). Rediscovering thesocial group: A self-categorization theory. Oxford:Blackwell.

van Knippenberg, A. (1978). Status differences,comparative relevance and intergroupdifferentiation. In H. Tajfel (Ed.), Differentiationbetween social groups (pp. 171–199). London:Academic Press.

Wilson, G. D. (Ed.) (1973). The psychology ofconservatism. London: Academic Press.

Paper received 28 March 2003; accepted 20 April 2003.

Biographical notesjennifer r. overbeck completed a postdoctoral

fellowship at Stanford University and is nowassistant professor of organizational behavior atthe University of Southern California’s MarshallSchool of Business. Her research focuses on effectsof (and beliefs about) social power and status.

john t. jost is currently associate professor ofpsychology at New York University. Until recently,he was associate professor of organizationalbehavior at Stanford University. His researchinterests include stereotyping, prejudice,intergroup relations, social justice, ideology,political psychology, and the theory of systemjustification.

cristina o. mosso is a researcher in socialpsychology at the University of Turin, Italy. Herwork addresses issues of stereotype transmission,the dynamics between personal and social identity,and, more broadly, psychological and structuralfactors in intergroup relations.

agnieszka flizik is pursuing a doctorate inorganizational behavior at the StanfordGraduate School of Business. Her interests lie instatus and power influences on intergrouprelations and the perception of time inorganizational functioning.

Group Processes & Intergroup Relations 7(1)

54

03 GPI7-1 Overbeck (JB/D) 23/1/04 9:54 am Page 54

at UNIV OF SOUTHERN CALIFORNIA on May 23, 2011gpi.sagepub.comDownloaded from