Jussim (Reference Jussim2012) presents a compelling case against the notion that the self-fulfilling prophecy and expectancy effects of descriptive stereotypes (hereafter “stereotypes”) are potent shapers of social reality. We accept Jussim's claims that (a) the evidence for self-fulfilling prophecy and expectancy effects is weak, fragile, and fleeting, and (b) stereotype beliefs are not inherently inaccurate. Nevertheless, it is premature and reductionist to conclude that stereotypes do not shape social reality. Stereotypes have a substantial influence in shaping social reality through their influence on collective action.

Tajfel's (Reference Tajfel1974) and Moscovici's (Reference Moscovici and Forgas1981) critiques of social psychology as overly reductionist emphasised that explanations of social phenomena, particularly coherent collective behaviour, must incorporate the psychology of shared social perception. Theory must account not only for the interpersonal level of judgment and perception, but also for the consensual understanding of the macro-level social relations in which different groups are embedded (Abrams Reference Abrams, Mikulincer and Shaver2015; Abrams & Hogg Reference Abrams, Hogg, Kruglanski and Higgins2004; Tajfel & Turner Reference Tajfel, Turner, Austin and Worchel1979). Specifically, the common collective behaviour of geographically dispersed, socially diverse, groups of individuals is grounded in their understanding of consensually shared stereotypes (Tajfel Reference Tajfel, Turner and Giles1981). The case that more complete explanations in social psychology require attention to both the micro- and macro-levels of analysis has been reinforced by numerous scholars (e.g., Abrams & Grant Reference Abrams and Grant2012; Dixon et al. Reference Dixon, Levine, Reicher and Durrheim2012; Oishi et al. Reference Oishi, Kesebir and Snyder2009; Pettigrew Reference Pettigrew2006; Wright & BarayReference Wright, Baray, Dixon and Levine2012). The self-fulfilling prophecy and expectancy effects described by Jussim exist at the interpersonal level of analysis: they involve a perceiver and a (stereotyped) target interacting directly or indirectly. Though Jussim notes the macro-level influence of stereotypes, this is typically to refute the assumption that stereotypes are inherently inaccurate. However, the effects of stereotypes on behaviour extend beyond the issue of whether they are accurate or not: after all, the accuracy of a belief is not a prerequisite for that belief to affect behaviour. Thus, irrespective of veracity, the role of consensual stereotype beliefs in motivating or justifying the collective behaviour of groups of individuals is overlooked in Jussim's argument, thereby missing an important route by which stereotypes shape social reality.

“Collective action against collective disadvantage is one of the major pathways to social change” (van Zomeren et al. Reference van Zomeren, Postmes and Spears2012, p. 52). History is replete with examples of collective action (CA) stimulating pervasive and profound changes in social reality. Prominent examples include the end of apartheid in South Africa, the abolition of slavery in the New World, and the host of civil rights movements throughout the 20th century (e.g., see Dixon et al. Reference Dixon, Levine, Reicher and Durrheim2012; Hardin Reference Hardin1982; Tilly & Wood Reference Tilly, Wood, Diani and McAdam2003). We note that CA can range from violent revolutions and terrorism, to peaceful demonstrations, petition signing, campaigning, and voting (Abrams & Grant Reference Abrams and Grant2012; Tausch et al. Reference Tausch, Becker, Spears, Christ, Saab, Singh and Siddiqui2011). Furthermore, CA can be directed at improving the position of one's own group, or can be “sympathetic” on behalf of another group (Saab et al. Reference Saab, Tausch, Spears and Cheung2014; Stewart et al. Reference Stewart, Pratto, Zeineddine, Sweetman, Eicher, Licata, Morselli, Saab, Aiello, Chryssochoou, Cichocka, Cidam, Foels, Giguere, Lui, Prati and van Stekelenburg2016). We now present evidence to support our contention that stereotypes influence engagement in CA and thus shape social reality indirectly.

Complementary stereotyping may serve to pacify CA engagement by enhancing support for the status quo. Complementary stereotyping involves the assignment of benevolent traits that off-set the presence of negative trait assignments, or vice versa (e.g., see Cuddy et al. Reference Cuddy, Fiske and Glick2008; Glick & Fiske Reference Glick and Fiske2001). Studies show that people who engaged in more complementary stereotyping of Northerners and Southerners as agentic and communal, respectively (in Italy), or communal and agentic, respectively (in England), viewed the social system as fairer and more legitimate (Jost et al. Reference Jost, Kivetz, Rubini, Guermandi and Mosso2005). Even in countries where general support for the status quo is low, people who endorse complementary stereotypes express greater satisfaction with the current socioeconomic and political reality (Cichocka et al. Reference Cichocka, Winiewski, Bilewicz, Bukowski and Jost2015). Indeed, across 37 different countries, such complementary stereotype beliefs are strongest in societies where income inequality is higher (Durante et al. Reference Durante, Fiske, Kervyn, Cuddy, Akande, Adetoun, Adewuyi, Tserere, Ramiah, Mastor, Barlow, Bonn, Tafarodi, Bosak, Cairns, Doherty, Capozza, Chandran, Chryssochoou, Iatridis, Contreras, Costa-Lopes, Gonzalez, Lewis, Tushabe, Leyens, Mayorga, Rouhana, Castro, Perez, Rodriguez-Bailon, Moya, Marente, Galvez, Sibley, Asbrock and Storari2013). The proposition that the consensual complementary stereotyping of various social groups pacifies engagement in CA that might otherwise change prevailing socioeconomic inequality is confirmed by experimental evidence. Jost and colleagues revealed that complementary stereotypes of the “poor” as “happy/honest,” or the “rich” as “unhappy/dishonest” (compared to unhappy/dishonest or happy/honest, respectively) led college students to report increased satisfaction with the socioeconomic and political status quo in the US (Kay & Jost Reference Kay and Jost2003; also see Kay et al. Reference Kay, Czapliński and Jost2009). Furthermore, exposure to similar stereotypes of the poor were found to increase support for government policy, and diminish support for disruptive protest against government pension reform, among demonstrators at a 2008 May Day rally in Greece (Jost et al. Reference Jost, Chaikalis-Petritsis, Abrams, Sidanius, Van Der Toorn and Bratt2012).

Effects of complementary stereotyping on CA engagement are also evident in research on benevolent sexism (e.g., stereotyping women as more “caring” than men, see Glick & Fiske Reference Glick and Fiske1996). Across 19 countries, women endorsed complementary stereotypes of their own gender (e.g., women as more communal and less agentic) most strongly in countries where average levels of sexism were highest (Glick et al. Reference Glick, Fiske, Mladinic, Saiz, Abrams, Masser, Adetoun, Osagie, Akande, Alao, Annetje, Willemsen, Chipeta, Dardenne, Dijksterhuis, Wigboldus, Eckes, Six-Materna, Exposito, Moya, Foddy, Kim, Lameiras, Sotelo, Mucchi-Faina, Romani, Sakalli, Udegbe, Yamamoto, Ui, Ferreria and López2000; Glick & Fiske Reference Glick and Fiske2001). Thus, complementary stereotypes may pacify engagement in CA to change a prevailing social reality of substantial gender inequality. Experimental evidence supports this proposition. In four studies Becker and Wright (Reference Becker and Wright2011) found that women's engagement in CA to address gender inequality (e.g., petition signing, flyer distribution, self-reported intentions) decreased when they were exposed to complementary stereotypes of their gender (also see Becker Reference Becker2012). Moreover, when women are exposed to such complementary stereotypes, they show greater satisfaction with the status quo of gender relations (Jost & Kay Reference Jost and Kay2005), reduced CA engagement (Foster Reference Foster1999), and a greater resistance to changing the social order (Di Bella and Crisp Reference Di Bella and Crisp2015).

In contrast to complementary stereotyping, which pacifies CA of either kind (own-group-directed or sympathetic), hostile stereotype beliefs (overtly negative stereotypes, e.g., “women are less intelligent than men”) tend to have divergent effects; promoting greater CA engagement among stereotyped targets, whilst attenuating sympathetic CA engagement. For example, when targets of hostile stereotypes see such beliefs expressed publicly, they demonstrate greater engagement in CA (Becker & Wright Reference Becker and Wright2011; Ellemers & Barreto Reference Ellemers and Barreto2009). A number of studies have documented the association between perception of hostile stereotype beliefs and anger amongst the stereotyped (Bosson et al. Reference Bosson, Pinel and Vandello2010; Ellemers & Barreto Reference Ellemers and Barreto2009; Swim et al. Reference Swim, Hyers, Cohen and Ferguson2001). Given that group-based anger is a critical driver of engagement in CA (Leach et al. Reference Leach, Iyer and Pedersen2006; Tausch et al. Reference Tausch, Becker, Spears, Christ, Saab, Singh and Siddiqui2011; van Zomeren et al. Reference van Zomeren, Spears and Leach2008), this provides one mechanism through which hostile stereotype beliefs operate to shape social reality. Indeed, Ellemers and Barreto (Reference Ellemers and Barreto2009) found that women who were confronted with the stereotype that women are less intelligent than men reported significantly greater anger, support for CA, and intentions to protest.

Moreover, those who endorse hostile stereotypes are less likely to engage in, or may even oppose, CA on behalf of the stereotyped group (i.e., sympathetic CA). Considering the critical role of sympathetic CA in social change movements (e.g., Leach et al. Reference Leach, Snider, Iyer, Walker and Smith2002; Simon & Klandermans Reference Simon and Klandermans2001; Tilly & Wood Reference Tilly, Wood, Diani and McAdam2003), this constitutes an equally important avenue through which stereotypes influence social reality. Stewart et al. (Reference Stewart, Pratto, Zeineddine, Sweetman, Eicher, Licata, Morselli, Saab, Aiello, Chryssochoou, Cichocka, Cidam, Foels, Giguere, Lui, Prati and van Stekelenburg2016) combined data from twelve countries to examine sympathetic support for Arab CA in the Arab uprisings that began in 2010. Endorsement of the hostile stereotype “Arabs are not competent enough to govern themselves” predicted reduced intentions to engage in sympathetic CA for the Arab peoples. Similarly, across five studies, participants who more strongly stereotyped the agents of social change (e.g., feminists as “militant”) were less likely to engage in sympathetic CA on their behalf (Bashir et al. Reference Bashir, Lockwood, Chasteen, Nadolny and Noyes2013).

Experimental evidence shows that hostile stereotypes can also directly affect public policy support. Johnson et al. (Reference Johnson, Olivo, Gibson, Reed and Ashburn-Nardo2009) found that activating the “Black criminal” and “promiscuous Black female” stereotypes significantly diminished support for public policy intended to benefit Black males and Black females, respectively. Similarly, when Hurwitz and Peffley (Reference Hurwitz and Peffley2005) activated racial stereotypes (e.g., “Blacks are lazy”) using minor changes in the language presented to participants, this increased White participants’ support for punitive policies, such as the building of extra prisons (over less punitive policies such as antipoverty programs). Maurer et al. (Reference Maurer, Park and Judd1996) provided specific evidence that the effects of stereotype beliefs extend beyond the interpersonal level of analysis. Participants’ endorsement of stereotypes of two different social groups (gay people and welfare recipients) predicted their public policy positions, independent of any interpersonal judgments made in individual cases involving welfare recipients and gay people. Indeed, Maurer et al. argued that “the nature of public policy judgments requires thought at the superordinate level – what the group-as-a-whole is like” (p.412). Other scholars have supported this, contending that public policy stance is group-centric, that is, “shaped in powerful ways by the attitudes citizens possess towards the social groups they see as [affected by] the policy” (Nelson & Kinder Reference Nelson and Kinder1996, p.1055).

These experimental studies reinforce conclusions from extensive field evidence. Gilens’ (Reference Gilens2009) comprehensive analysis of survey data, opinion polls, and public policy actions identified hostile stereotypes (e.g., “Blacks are lazy”) as one of the primary factors in U.S. citizens’ opposition to CA intended to address systemic socioeconomic inequality (also see Kluegel & Smith Reference Kluegel and Smith1986). Other scholars have identified that the adoption of a “racial frame” in policy discourse (e.g., stereotyping Hispanics as undeserving and lazy) profoundly influenced support for, and engagement with, public policy action in the United States between 1990 and 1997 (Brown Reference Brown2013). Similar patterns have also been observed in the United Kingdom. Bamfield and Horton (Reference Bamfield and Horton2009) examined large scale opinion surveys conducted in 2008–2009, and found that tacitly stereotyping the poor as irresponsible and lazy (i.e., attributing their socioeconomic status to individual fault) predicted opposition to welfare policy initiatives (also see de Vries Reference de Vries2015). Though we are aware survey results cannot imply causation, we emphasize the diversity of evidence attesting to the influence of stereotypes upon public policy support (both experimental and correlational). We also note evidence that suggests people spontaneously generate the prototypical member (stereotypic exemplar) of a relevant social group (as opposed to other policy-relevant principles) when thinking about public policy actions that will affect that social group (Lord et al. Reference Lord, Desforges, Fein, Pugh and Lepper1994; also see Reyna et al. Reference Reyna, Henry, Korfmacher and Tucker2006).

In sum, taken together, the weight of evidence supports the contention that stereotypes do exert a substantial influence upon social reality – through their impact on CA engagement and people's priorities for public policy. We therefore note the irony of Jussim's assertion that, having been liberated from their false assumptions regarding stereotype inaccuracy, scholars are now free to focus on addressing the “actual” causes of social inequality and oppression (p.425, para. 2). There are multiple roots of inequality (besides stereotypes), such as socioeconomic disparity, deprived socialization, or inadequate healthcare. However, as our review demonstrates, stereotype beliefs are intimately related to motivating the necessary collective action that would address some of these alternative causes of inequality. Finally, we emphasize that our review makes no normative assumptions about the moral or political “rightness” of engaging in CA (whether the means, or ends, are justified or desirable). Rather, we have advanced an empirical case that stereotype beliefs influence CA engagement, and thus, do have a hand in shaping social reality. We concur strongly with Jussim that the economic, political and other roots of group-based inequality need to be addressed by economic, political and other means. There are real differences between groups that have to be understood. However, we also contend that it is people's shared, collective, understanding of these differences that is the vehicle for coordinated and meaningful social change.

As a consequence of distinguishing between the interpersonal- and-collective levels of analysis, we acknowledge the case for some key claims of Jussim's book, namely that (a) the evidence for self-fulfilling prophecy and expectancy effects is weak, fragile, and fleeting, and (b) stereotype beliefs are not inherently inaccurate. However, we reject the conclusion that stereotype beliefs do not influence social reality.