“Erst kommt das fressen, dann kommt die Moral” (“First grub, then ethics”) wrote Bertolt Brecht in his Threepenny Opera, expressing the preferences of disadvantaged groups. These preferences are seldom accurately understood by even the best intentioned members of advantaged groups.

Recently, social psychologists started addressing this problem by switching the focus from the perpetrators' (high status/advantaged/majority group) feelings, needs, and attitudes into the deeper study of the victims' (low status/disadvantaged/minority group members) perspective. This paradigmatic shift occurred simultaneously in studies of intergroup helping (Nadler Reference Nadler2002), reconciliation (Shnabel & Nadler Reference Shnabel and Nadler2008), intergroup contact (Saguy et al. Reference Saguy, Tausch, Dovidio and Pratto2009), intergroup emotions (Imhoff et al. Reference Imhoff, Bilewicz and Erb2012), and collective action (Wright & Lubensky 2008). All these lines of research suggest a divergence of goals between advantaged and disadvantaged groups. Dixon et al. declare that the prevalent psychological approach to prejudice – treating negative evaluations as hallmarks of discrimination – was based on researchers' narrow focus on majority groups. The modern psychology of intergroup relations, by addressing minority groups' needs, should rather focus on subtle biases and paradoxical discriminatory consequences of seemingly unprejudiced attitudes. Acknowledging the importance of such an approach in understanding the dynamics of discrimination in the West, this comment will suggest some cultural limitations of it.

The classic theories of prejudice and prejudice reduction (Allport Reference Allport1954; Zawadzki Reference Zawadzki1948) were developed in countries where prejudiced attitudes were culturally accepted and sometimes even normatively supported by legal, political, and religious authorities of majority groups. Dixon et al. argue that current structures of oppression are more nuanced: modern discrimination sometimes has a “benevolent” expression, so traditional prejudice reduction strategies (e.g., contact, recategorization) have limited effects on improving intergroup relations. This situation seems specific to developed Western societies, where expression of ethnic prejudice among high-status groups is suppressed by strong societal norms of political correctness (Crandall et al. Reference Crandall, Eshleman and O'Brien2002). Most recent psychological studies performed on North American or West European student samples hardly detect any overt prejudice. The progress of integration and the civil rights movement influenced Americans' responses in racial attitudes surveys, while not changing structural power relations. As early as 1992, more than 95% of American students declared acceptance of a black neighbor in a study of social distance (compared to 41% in 1949; Dovidio et al. Reference Dovidio, Brigham, Johnson, Gaertner, Macrae, Stangor and Hewstone1996). At the same time, most Americans still chose to live in racially segregated rather than integrated cities (Goldsmith & Blakely Reference Goldsmith and Blakely2010). If people do not express their attitudes openly, then it should not be surprising that discriminatory behaviors cannot be well predicted by explicit measures of attitudes. We should expect “benevolent” and implicit prejudice to be the dominant expressions of intergroup hatred in such societies (McGrane & White Reference McGrane and White2007). It is also obvious that in such societies, changes in expressed prejudice (caused by intergroup contact or tolerance education) do not automatically affect discriminatory intentions and behaviors.

In societies where ethnic conflicts are more intense and political correctness norms are weaker than in the West, blatant forms of prejudice are still operating – and systematically result in discrimination. The most recent European Value Survey (EVS 2011) shows that prejudice expression significantly differs across countries. In typical countries of Western Europe (such as Belgium, France, or Germany), less than 10% of participants declared that they would not accept people of different race as their neighbors, whereas in several countries of post-Communist Eastern Europe such prejudice was openly expressed by more than 20 percent of respondents. Our recent survey of prejudicial beliefs about Jews showed that in Poland such negative beliefs are good predictors of discriminatory intentions and behavioral non-helping (Bilewicz & Krzemiński Reference Bilewicz and Krzemiński2010; Bilewicz et al., in press). More subtle measures (e.g., secondary anti-Semitism scale) were less predictive than traditional, blatant ones. In places of actual or recent blatant ethnic conflicts, such as Armenia, Turkey, or Northern Cyprus, up to 56% of citizens openly expressed unwillingness to accept neighbors of different race. In other European countries, anti-immigrant prejudice is still a strong predictor of discriminatory intentions – particularly among people of low income (Küpper et al. Reference Küpper, Wolf and Zick2010). In regions where prejudice currently causes much more harm – in the global South and East, in impoverished areas – the link between prejudicial beliefs and discriminatory behavior remains strong and stable. These regions belong to most underresearched by social psychologists who often base their theories on American or West European students samples (i.e., “WEIRD” people: Western, educated, industrialized, rich, democratic; Henrich et al. Reference Henrich, Heine and Norenzayan2010).

Dixon et al. suggest that old, blatant prejudice is currently being replaced with more nuanced ambivalent stereotypes or implicit infra-humanization. However, the same groups (i.e., Gypsies) who are subtly infra-humanized in Britain are still harshly and openly dehumanized in Romania, where they are subject to everyday discrimination (Marcu & Chryssochoou Reference Marcu and Chryssochoou2005; Tileagă Reference Tileagă2007). And in most acts of genocide or ethnic cleansing, it was not subtle infra-humanization, but rather overt dehumanization that shaped propaganda and intergroup perceptions (Bandura Reference Bandura1999; Bilewicz & Vollhardt, Reference Bilewicz, Vollhardt, Golec De Zavala and Cichocka2012). Ambivalent stereotyping is the result of a complex, two-dimensional structure of intergroup perception in Western societies. Recent research in Poland found systematic correlations between these two dimensions: Groups perceived as incompetent were also perceived as unfriendly (Winiewski Reference Winiewski2010). It seems that a more general prejudiced attitude was responsible for all forms of out-group negativity.

Dixon et al. shed some light on the subtle mechanisms of discrimination that inhibit collective action in modern democratic societies. However, most ethnic violence occurs outside of this WEIRD world. In many underresearched societies of the global South and East, prejudiced remarks are still normatively accepted and hate speech is neither a felony nor a taboo. In such societies, social justice programs, as well as prejudice reduction programs, are effectively changing attitudes of both advantaged and disadvantaged group members. Gacaca tribunals in Rwanda – an effort leading to justice restoration after genocide of Tutsis – reduced prejudice among perpetrators and victims of the genocide (Kanyangara et al. Reference Kanyangara, Rimé, Philippot and Yzerbyt2007). Changing power asymmetries and changing attitudes went hand in hand. And this is where prejudice reduction campaigns and interventions could still be applicable – as they change the situation of both advantaged and disadvantaged social groups.