R2. Prejudice, prejudice reduction, and the limits of collective action models of social change

In our target article, we basically made four points:

1. 1. Long term relationships of inequality between groups (e.g., relations of class, race, and gender in many societies) are accompanied not only by negative intergroup attitudes, but also by ambivalent or even positive attitudes.

2. 2. The latter sometimes help to sustain broader forms of inequality and discrimination. That is, however strange the idea may initially appear, warm feelings and positive thoughts about others are sometimes part of the social and political cement that solidifies an unequal status quo.

3. 3. As such, interventions aimed at getting us to like one another are not necessarily an effective mechanism to promote social change. They can have the paradoxical consequence of inhibiting the social psychological impetus for change, particularly for members of historically disadvantaged groups.

4. 4. In consequence, we believe it is time to reevaluate the efficacy of the prejudice reduction model in relation to other models of social change. In this regard, we have discussed an alternative model based around collective action to achieve equality. We have also invited critical reflection on the relationship between these two models of social change, which might be conceived, for example, as incommensurable, reconcilable, or complementary (see also Wright & Lubensky Reference Wright, Lubensky, Demoulin, Leyens and Dovidio2009).

Several commentators have challenged these central strands of our argument. To begin with, they have worried that our critique of the orthodox conception of prejudice goes too far, underestimating the extent to which disadvantaged groups continue to be targeted for antipathy, hostility and violence. Although conceding that prejudice may take diverse forms, both Abrams, Vasiljevic, & Wardrop (Abrams et al.) and Brown argue that its “old-fashioned” expressions remain an urgent problem in many societies. Animosity towards immigrants and Muslims in Britain and other parts of Western Europe provides a stark illustration. Outside of Western democracies, in societies where comparatively little research on prejudice has been conducted, the situation is probably worse. Bilewicz's reflections on relations in Armenia, Romania, Poland, Turkey, and Northern Cyprus, for example, suggest that blatant forms of prejudice are alive and well and that they predict discriminatory actions more strongly than the “subtle” indices that have dominated recent psychological research in the West. In this vein, Brown warns that: “if we should be aware that the wolf of prejudice (towards women and some minority groups) can sometimes come disguised in benevolent sheep's clothing, it is also important to remember that the prejudice experienced by many groups is far from ‘benevolent’” (para. 2).

Developing a related argument, several commentators have vigorously defended the utility of traditional prejudice reduction interventions, particularly interventions based on Allport's (Reference Allport1954) contact hypothesis (Abrams et al.; Brown; N. Haslam; Hewstone et al.). After all, if old-fashioned antipathy remains an urgent problem in many societies, then surely we need interventions that can encourage us to like one another more or at least “hate one another less” (Brown)? Intergroup contact – especially when it occurs under favourable circumstances – has consistently been found to accomplish this goal. It works even on the attitudes of the highly prejudiced (Hodson Reference Hodson2011). Moreover, as Hewstone et al. observe, contact has closely related benefits: it promotes intergroup forgiveness, empathy, and perspective-taking and reduces intergroup anxiety and distrust. Unlike collective action interventions, moreover, it facilitates peaceful coexistence, reducing the likelihood of disturbances, unrest, or bloodshed.

In any event, as several commentators warn, the collective action model has problems of its own and is no panacea for discrimination and inequality. Some disadvantaged groups are simply too fragmented, numerically small, or lacking in political clout to mount a meaningful challenge to the status quo, leaving prejudice reduction as a more feasible option for improving intergroup relations (N. Haslam). Relatedly, in many cases of collective mobilization, protagonists ultimately fail to achieve meaningful change, and they may even create conditions that are appreciably worse (Abrams et al.). Struggles to transform the status quo may meet with violent resistance from members of advantaged groups; and this may set in motion a downward, destructive cycle of intergroup conflict that is ultimately costly to both sides (N. Haslam; Hewstone et al.; Schaller & Neuberg). Because it sharpens and magnifies group differences, collective action may also help to “essentialize” social categories and relations (N. Haslam). It may thus encourage processes of mutual dehumanization and hinder long-term prospects of reconciliation. In light of these problems, it might seem that the perils of collective action outweigh its prospects. Hence, out of pragmatism if nothing else, some may concur with Abrams et al. that the objective of getting the powerful to like the less powerful “can surely only be helpful” (para. 2).

In responding to these criticisms, we wish to make three initial points that were perhaps not articulated clearly enough in our target article.

1. 1. The focus of our argument was (and remains) on intergroup relations marked by relatively stable, institutionally embedded, relations of inequality. Certainly there can be prejudices in the absence of inequalities, between redheads and brunettes, between Manchester United and Manchester City fans, between people from Yorkshire and those from Lancashire, and so on. However, like most researchers who have studied prejudice, our interest derives from the concept's relevance to the entrenched inequalities that scar our societies such as racism, sexism, and anti-Semitism.

2. 2. In societies where equality has been achieved and institutionalised disadvantages are negligible or nonexistent, we wholeheartedly embrace the goal of achieving intergroup harmony as a valuable end in itself. We also acknowledge that in such contexts collective action to transform unequal and discriminatory relations is, by definition, unnecessary.

3. 3. Our concern is therefore with the idea that prejudice reduction is the way forward where there are existing inequalities and practices of discrimination between groups. More specifically, we identify and contest the widespread assumption that enhanced liking will be a “bottom up” driver that leads to transformations at a cultural, institutional, and legal level (contra Alicke).

Bearing in mind these points, we offer the following additional responses to our critics. Our first response addresses the prevalence of “hostile” versus “benevolent” forms of prejudice. Those who point out that hostile prejudice remains alive and well throughout the world are quite right. However, they are quite wrong to believe that we think otherwise. Indeed, building on the work of Leach (Reference Leach2005), we would conjecture that the supposedly universal shift from “old” to “new” forms of prejudice has been overstated in recent psychological literature, an idea reinforced by Bilewicz's thoughtful commentary.

In discussing research on “benevolent” prejudices, we were not making a point about prevalence at all. Rather, we were underscoring a conceptual argument. That is, if researchers think that the answer to social inequality is to promote harmony, then they should look closely at those instances where intergroup harmony seems to reign and majorities talk positively about minorities. They will find that things are not necessarily much better for the minority. It might even be that their prospects are worse.

To put this point more precisely, prejudice research has, throughout its long history, focused almost exclusively on the problem of measuring, explaining, and changing negative evaluations. Until recently, this focus has largely obscured the role of ostensibly positive and ambivalent attitudes in sustaining relational inequality between groups. It has masked the fact that “unalloyed antipathy” (Glick & Fiske Reference Glick and Fiske2001) is neither practical nor ideologically functional in such relationships and in fact impedes their smooth operation. “Paternalistic” power relations entail (and require) more complex constellations of social attitudes and beliefs – often blending positive attributions about those who follow ascribed roles and display valued attributes with hostility towards those who step out of line. These attitudes and beliefs in turn shape the unfolding actions and reactions of both dominant and subordinate groups and help to reproduce their unequal relationship in ways that cannot be readily captured by a simple model of prejudice as negative evaluation. As our target article elaborated, they find expression in seemingly benevolent forms of (conditional) inclusion, relations of helping, idealization of (certain) attributes of disadvantaged groups, and heartfelt support for general principles of social equality amongst the advantaged (offset by continuing resistance to its implementation).

In drawing together recent evidence on these kinder, gentler forms of prejudice, we have sought to highlight some limitations of mainstream work in the field, flagging the need for alternative theoretical, methodological, and applied perspectives. It is in the arena of social change that these limitations are most troubling to us and, as we shall see, to several other commentators on our target article. We do not deny that the prejudice reduction has a beneficial impact on negative intergroup emotions and beliefs. Nor do we deny that, in many intergroup contexts, getting people to like one another more is a vitally important objective. However, we do deny that prejudice reduction is an effective mechanism for improving intergroup relations (i.e., reducing discrimination and promoting equality) in contexts where intergroup relations are marked by long-standing patterns of institutional discrimination and where intergroup relations are relatively stable and “harmonious.” Moreover, even in contexts where intergroup relations are far from stable and harmonious but remain unequal, we argue that prejudice reduction is sometimes more clearly in the interest of advantaged than disadvantaged groups. As outlined in our target article and developed by the commentaries discussed in the next section, we have reason to suspect that interventions to reduce prejudice may, paradoxically, entrench existing status and power relations in unequal societies.

This takes us to the second set of concerns raised in the responses. Our central argument is that, in this kind of context, collective action becomes an essential motor of social change (see also Drury; S. Haslam & Reynolds; Maoz; Tausch & Becker; Wright & Bitacola). It is important to note, however, that this is a conditional and not an absolute statement in support of collective action. On the one hand, it does not suggest that collective action is essential in all contexts. We have already pointed to this, but it is sufficiently important to bear repetition. Collective action is not essential where intergroup hostility occurs in the absence of a history of intergroup inequality. Equally, it is no longer essential where intergroup hostility remains after a history of intergroup inequality has been overcome. As Wright & Bitacola suggest, we may need to introduce a temporal dimension to our analysis. It may well be that prejudice reduction is particularly relevant precisely when bitter and prolonged collective struggles have defeated a system of inequality. In South Africa, the anti-apartheid struggle gave way to a Truth and Reconciliation process. Intergroup theorists have much to learn from this kind of process, and we agree with Wright & Bitacola that the value of interventions may be a matter of timing.

On the other hand, to argue that collective action is essential to overcoming inequality is not to argue that any form of collective action will do or that collective action will always be successful. That would be far too stringent a test for any intervention to pass. Commentators are therefore right to note that the ability of protest to undermine institutional inequalities is contingent on a range of factors, including the material and communicative resources available to protesting groups and their capacity to form effective coalitions and third-party alliances. We also acknowledge that collective action may have unintended negative consequences (e.g., Abrams et al.; Brown; N. Haslam; Schaller & Neuberg), including an upsurge in the violence faced by the disadvantaged – and that therefore it is indeed irresponsible to advocate revolt when the conditions are not ripe. Yet the lesson here for us is not to abandon collective action, but rather to develop our analyses of when and how people can be mobilised to act together, how they can build alliances and win third parties to their side, how (in other words) we can better understand the conditions under which collective action succeeds. To this end, we draw on a wealth of evidence that shows the value of collective action in transforming unequal relations between groups, only a fraction of which was discussed in our target article. Indeed, the extensive psychological literature on social identity and collective action – a literature to which some of our most trenchant critics have contributed – is premised on the assumption that, where group members are in a subordinated social position, social change occurs precisely to the extent that they act together as group members and directly challenge the dominant group (Tajfel & Turner Reference Tajfel, Turner, Austin and Worchel1979).

This brings us to a related point, which concerns the relationship between collective action and social categorization. We disagree with N. Haslam's point that collective action tends to “essentialize” human categories, encouraging us to treat particular social divisions in society as pre-given, natural, or inevitable. To the contrary, by disrupting the institutional structures that entrench and reify categories (e.g., structures of segregation), it often has precisely the opposite effect! It reveals the contingent, historical specific, and fluid nature of our social identities and relationships. It also enables the realignment of established ways of constructing such identities and relationships. To use more everyday terms, constructions of “us” and “them” are often profoundly altered as collective action unfolds, a point emphasized by Drury.

In sum, our argument is not about the attempt to change hostile attitudes in itself. Rather, it is about the means by which fundamental changes to inequality come about. To repeat, the issue is about models of social change. Everything else is secondary. We are not arguing that contemporary prejudice is primarily benevolent rather than hostile. We are not arguing per se against contact between social groups or denying that prejudice reduction interventions can have beneficial consequences under some conditions. What we are suggesting is that the way to reduce inequality is not to get people to like each other but to get them to mobilise collectively against the structures of inequality. This, if nothing else, is the point to retain. On this our position stands or falls. But that is not the end of the matter. Once this point is acknowledged, we can then go back and ask how intergroup hostility or benevolence contribute to (or stifle) collective mobilization. We can reexamine prejudice reduction interventions (e.g., intergroup contact) from the perspective of how they affect our readiness to acknowledge and challenge social inequality and to participate in collective action (and not simply how they affect our liking of other groups). Indeed, that is precisely what we do over the course of the following sections.

R3: “Let them eat harmony?” The limits of “prejudice” and “prejudice reduction” revisited

Our critique of the Prejudice Reduction Model of social change was based in part on the claim that prejudice reduction can have ironic effects on the political attitudes of the historically disadvantaged. In promoting warmer feelings and thoughts about the historically advantaged, we argued, such interventions may also decrease the disadvantaged's support for policies designed to combat inequality, acknowledgement of institutional discrimination, and willingness to engage in collective action to transform social inequalities. As one of the original reviewers of our target article pointed out, however, empirical evidence to support this claim remains comparatively sparse. Moreover, the contextual boundary conditions for these “ironic” effects are poorly understood, and theoretical explanations of how and why they occur remain under-specified. Several commentaries on our target article have begun to address these gaps.

Eagly & Diekman at once support and qualify our core argument. They agree that it is insufficient to equate prejudice with negative attitudes precisely because discrimination is sometimes grounded in seemingly positive attributions about other groups (e.g., women are more socially sensitive than men). They agree, too, that standard prejudice reduction interventions may be ineffective in challenging institutionalised patterns of discrimination; indeed, in some settings, it may sometimes be more effective to foster less positive attributions about groups such as women and senior citizens (e.g., attributions of ambition, aggression). However, Eagly & Diekman also argue that many researchers, including ourselves, commonly fail to recognise that it is context specific rather than abstract evaluations that are most critical in sustaining discrimination. For example, whereas women's social sensitivity may be a quality that contributes to their generally positive evaluation by men, it may carry costs in contexts where normative role requirements specifically valorise other qualities (e.g., corporate executive), leading to discrimination and exclusion.

Related to this, Eagly & Diekman hold that the key to political change lies in widening the access of disadvantaged groups to desirable and powerful social roles (or, presumably, in redefining role requirements so that they are no longer incongruent with the attributed qualities of disadvantaged group members). Over time, this process alters group stereotypes (e.g., the belief that women are too nice to take the hard-nosed decisions required of a corporate executive) and gradually erodes the likelihood of future discrimination. Eagly & Diekman suggest that neither prejudice reduction nor collective action is likely to achieve this goal. The former tends to promote a general positive evaluation whilst leaving intact the context specific attitudes that underpin discrimination. The latter does not directly target access to the social roles associated with political power and economic advantage.

In our view, Eagly & Diekman's analysis raises several important points. We accept that discrimination is often grounded in context-specific evaluations – both positive and negative – rather than in generic intergroup attitudes. We also accept that this point was somewhat neglected in our target article. However, we disagree with Eagly & Diekman's dismissal of the value of collective action in transforming access to social roles associated with power and social advantage. In discussing how such transformation occurs, they rightly point to the role of broader historical and economic shifts in society (e.g., how World War II opened up opportunities for women to enter employment roles that were traditionally reserved for men). Yet we would argue that disadvantaged communities are not simply the passive beneficiaries (or victims) of such shifts nor, crucially, can they afford to patiently await reforms generated by those in powerful roles already. We fear this will be a very long wait indeed. Rather, under the right social conditions, the disadvantaged can and must also become active agents in the transformation of role discrimination; and collective action is often the primary mechanism through which this agency finds expression. To be sure, as Eagly & Diekman imply, individual instances of social mobility into valued social roles may form a vanguard that eases the path for others. Equally, though, in the absence of collective action to alter the institutional structures of opportunity, such instances will all too often result in tokenism. Members of disadvantaged groups will be cast as “exceptions to the rule,” and normative practices of discrimination will remain intact.

Langdridge applies our core thesis to a quite different sociopolitical process – one not anticipated by our target article – namely the process of “coming out” within the lesbian, gay, and bisexual communities. Orthodox models of “coming out,” he contends, tend to propose a developmental pathway that takes as its ideal endpoint a “quiet acceptance of the wider social world” (para. 3). Whereas responses to living in a homophobic society may initially involve negative feelings such as identity confusion and anger, there is an underlying assumption that individuals should accommodate to, and become assimilated within, the wider heterosexual majority, becoming “good homosexual citizen[s]” (para. 3). In this framework, the significance of “coming out” as a political, as well as personal, event is marginalised. Members of lesbian, gay, and bisexual communities are encouraged to accept inequalities as “well-adjusted” individuals rather than challenge them as a mobilized collective.

Jost, Stern, & Kalkstein (Jost et al.) similarly extend our argument by drawing on the insights of Systems Justification Theory. This theory was originally devised to explain how, when, and why individuals and communities adopt beliefs that legitimate inequitable political systems. Jost et al. suggest that it also challenges the mainstream psychology of prejudice. First, in line with our position, they hold that acceptance of systemic inequalities is encouraged not only by negative but also by positive and complementary stereotypes about the disadvantaged. For example, endorsement of stereotypes of the working class as “poor but happy” is related to acceptance of inequalities of class. Second, Jost et al. hold that prejudice reduction interventions may actually facilitate rather than undermine this ideological process, encouraging the disadvantaged to view the status quo as legitimate or to misattribute the causes of economic or political disadvantage to in-group failings (e.g., lack of ambition or intelligence). If the commentaries discussed in the previous section suggest that our critique of the mainstream psychology of prejudice reduction went too far, Jost et al. suggest that it did not go far enough! They contend that prejudice reduction not only has a “sedative” effect on collective action to combat social injustice, but also a “palliative” effect, assuaging a deep-seated motivation to believe that we exist in a just world where people generally get what they deserve.

We welcome Jost et al.'s commentary as an important extension of our argument, with two qualifications. First, we think it is important to be careful not to imply that the disadvantaged have a natural or inevitable inclination to support the status quo. Second and related, we agree with Howarth, Wagner, Kessi, & Sen's (Howarth et al.'s) warning about the dangers of imputing a uniform “false consciousness” to the disadvantaged, implying that they simply “misperceive” their own situation as fairer than it “really is” – a criticism that could, with some justice, be levelled at our target article as well as at Jost et al.'s commentary. To clarify, for us the key point is that not that the disadvantaged are somehow a singular group of ill-informed or misguided “dupes” or that they have a natural tendency to accept passively the legitimacy of (unequal) social systems. Rather, we argue that the project of establishing warmer relations with the advantaged may, under certain sociopolitical circumstances, decrease the likelihood that members of particular communities will actively acknowledge and challenge inequality.

Jost et al. touch upon another more specific aspect of our argument. They cite emerging evidence that harmonious contact between the advantaged and the disadvantaged acts as a mechanism through which the system-justifying effects of prejudice reduction occur (e.g., see Cheung et al. Reference Cheung, Noel and Hardin2011; and for related examples, see Saguy & Chernyak-Hai Reference Saguy and Chernyak-Hai2012; Sengupta et al., in press). This theme is developed in other commentaries.

Maoz reflects upon a compelling tradition of research on programmes of “planned encounters” between Arab and Jewish Israelis that have sought to apply Allport's (Reference Allport1954) contact hypothesis within a context of long-standing (and some believe intractable) intergroup conflict. She argues that the outcomes of such programmes may be less favourable than the optimistic tone of recent work on intergroup contact might suggest. Yes, they have improved intergroup attitudes, particularly amongst Israeli Jewish participants. In seeking to implement a pleasant atmosphere of social harmony that promotes peaceful “coexistence,” however, they have also sidelined or suppressed Arab concerns over continuing social injustices in Israeli society. At worst, Maoz argues, such programmes “can be viewed as intentionally perpetuating existing asymmetrical power relations by focusing on changing individual-level prejudice while ignoring the need to address collective and institutionalized bases of discrimination” (para. 6).

Both Tausch & Becker and Wright & Bitacola also extend our analysis of the limits of intergroup contact as mechanism for reducing prejudice and promoting social change. Tausch & Becker highlight the need for a more complete theoretical model that can explain how and why contact has “demobilizing effects” on members of disadvantaged groups. In our target article, we suggested that such effects are partly explained by the role of contact in: (1) reducing perceptions of personal discrimination that then generalise to an intergroup level and (2) increasing positive emotional responses towards the advantaged. Building on the work of Jackman (Reference Jackman1994) and Wright and Lubensky (Reference Wright, Lubensky, Demoulin, Leyens and Dovidio2009), Tausch & Becker posit a third theoretical mechanism, namely, the role of positive contact in perpetuating the belief that social systems are permeable and that it is therefore possible for individual members of disadvantaged groups to climb the ladder of social and economic advancement. They cite some preliminary evidence that confirms this idea, showing how a so-called “individual mobility” orientation partly mediates the relationship between intergroup contact and collective demobilization. Clearly, this line of analysis dovetails with the systems justification perspective proposed by Jost et al.

Wright & Bitacola posit a fourth mechanism through which positive interactions between members of advantaged and disadvantaged groups may legitimate social inequalities and diminish motivations to participate in collective action. Building on Ridgeway's (Reference Ridgeway, Jost and Major2001) Status Construction Theory, they argue that such interactions may serve as an arena for the display (and situated reproduction) of wider status relations, expressed via an array of subtle cues that convey the superiority of one group over another. For example, linguistic and nonverbal markers of gender identity may subtly reinforce gender hierarchies during cross-sex conversations, a process evidenced by the extensive literature on gender and language use (e.g., Uchida Reference Uchida1992). Wright & Bitacola's commentary, then, demonstrates the need to investigate the nature of unfolding interactions between the members of advantaged and disadvantaged groups in order to understand how intergroup status and power differences are reproduced. We see this as a promising avenue for future research.

Another promising avenue might attempt to establish the conditions under which intergroup contact – and similar interventions to reduce prejudice – do not decrease the likelihood that disadvantaged group members will recognise, reject, and resist social inequalities or mobilize collectively to challenge the status quo. Tausch & Becker argue that nature and content of intergroup communication is crucial in this regard. In interactions where advantaged group members treat social inequality as illegitimate, for example, contact may not undermine participation in collective action. Similarly, Wright & Bitacola argue that positive feelings towards particular members of an advantaged group will undermine the collective action orientation of the disadvantaged only if such members are perceived as representative of that group. If they are perceived as a distinct subgroup or as “exceptions to the rule,” then positive interactions are unlikely to exercise “ironic” effects on political attitudes.

In sum, in this section we have explored commentaries that have developed or refined the core arguments of our target article, identifying valuable directions for future research. These commentaries have further demonstrated the inadequacies of a model of social change based on getting the advantaged to like the disadvantaged more. At the same time, they have explored the boundary conditions under which intergroup contact may be commensurable with a collective action model of social change, a point that we develop in the next section. Although this line of research remains at an early stage of development, we note that commentators have cited several recent studies that did not feature in our target article (e.g., Becker & Wright Reference Becker and Wright2011; Becker et al., under review; Cheung et al. Reference Cheung, Noel and Hardin2011). In the next section, reflecting in more detail on how this emerging paradigm of research might develop, we discuss themes that recurred across a number of commentaries.

R4. Emerging themes in the psychology of prejudice, discrimination, and social change