Lewin (1951) recognized that it “is usually easier to change individuals formed into a group than to change any one of them separately” (p. 228). More than sixty-five years later, social identity theory (SIT; Tajfel & Turner, 1979) and self‐categorization theory (SCT; Turner et al., 1987; collectively referred to as the social identity perspective/approach) offer many insights into why and how this is the case. At the heart of the social identity perspective is a comprehensive and systematic theory of the “group,” which generated a new view of the self-process where humans are both individuals and group members with both personal (“I”) and social (“we”) identities. Importantly, social and personal identities can change and, in turn, so too can behavior. An important part of the behavior change “puzzle”, often overlooked by researchers, policy makers, and practitioners, is that it is necessary to engage not only the “I” or “me” but also the “we.” This chapter outlines the potential of social identity processes, including in-group norms and social influence, in advancing understanding of behavior change. Taken together, research and practice applying the social identity approach to behavior change demonstrate considerable promise in promoting change in group contexts and for multiple behaviors in multiple domains such as work, education, and community settings.

In any given era, there may be a widespread consensus concerning what groups are seen as deserving of negative treatment. Indeed, such social-group memberships can represent a negative identity for those who are categorized as “one of them” (e.g., murderers, child abusers, smokers; see Crandall, Eshleman, & O'Brien, 2002). Likewise, groups that are consensually agreed upon as being positive (e.g., volunteer groups) have the potential to confer a positive social identity on their members. Social psychologists have assumed that a basic human desire to achieve and maintain a positive identity motivates people to seek membership in groups that are seen as positive and to avoid membership in groups that are seen as negative (Cialdini et al., 1976; Tajfel & Turner, 1979). Because people derive their identity in part from the groups to which they belong, it might seem, at first glance, to be rather contradictory to this central tenet in the field to find that some people remain loyal to groups that are seen as negative by other social groups or by society as a whole. Yet, clearly, many people choose to belong to a group even though they know they may face discrimination as a result.

What is more, we can point to many examples of people joining groups fully aware that this may make them vulnerable to discrimination (e.g., joining a feminist group or gay rights movement). In some extreme cases, people might even join such groups precisely because of their devaluation by the powers that be.

An extensive body of research has examined the factors that shape attitudes toward employment-based redistributive policies such as affirmative action and comparable worth. Empirical work has focused particularly on individualistic predictors of opposition toward these policies, including prejudice or justice beliefs (Bobocel, Son Hing, Davies, Stanley, and Zanna, 1998; Sears, Henry, and Kosterman, 2000). Although some research has considered whether policy support varies as a function of the type of affirmative action program – preferential treatment versus equal opportunity (Bobocel et al., 1998) – little research has assessed how features of employment settings themselves might affect attitudes toward redistributive policies and ideological beliefs. Aspects of the work setting are likely to shape people's beliefs about inequality and group-based policies by cueing different self-categorizations and comparison standards that people draw on when they evaluate their employment outcomes (Tajfel and Turner, 1986; Turner, Hogg, Oakes, Reicher, and Wetherell, 1987). In this chapter, we argue that the presence or absence of gender-based redistributive policies in employment settings convey different identity and comparison information, which then affects people's responses to gender differences in employment outcomes and whether they support or oppose policies that alter these outcomes. We expect that women and men who have conscious experience with redistributive policies will respond differently to these policies and related ideological beliefs than will those who do not have this experience.

Social identity theory (Tajfel & Turner, 1986) claims that how people interpret group-relevant events depends on their own group membership and the structural relations that exist between the ingroup and a salient comparison outgroup. When people think of themselves as a member of a social group, they are motivated to protect the identity of their group, and this can lead them to appraise the ingroup's actions in ways that benefit their ingroup (Baumeister & Hastings, 1997; Iyer, Leach, & Pedersen, in press). We argue that such ingroup-serving motivations affect the appraisal of ingroup actions and the emotions that group members experience when they self-categorize as members of their group (Branscombe, Doosje, & McGarty, 2002; Montada & Schneider, 1989).

Even when the self has clearly played no causal role in an event, depending on the meaning that event has for a valued social identity, differential emotional reactions can occur (Branscombe et al., 1993; Cialdini et al., 1976; Doosje et al., 1998; Yzerbyt et al., 2002). As Smith (1999) notes, group-based emotional responses rest on appraisals that consider the event's implications for one's social identity, whereas personal emotional responses are based on appraisals that consider the event in terms of its implications for personal identity. In this chapter, we explore the processes that are instigated when people are reminded of their ingroup's past or present harmful actions toward another group. We focus on how appraisals that reduce the severity and injustice of the ingroup's actions can undermine the experience of collective guilt.

History is replete with instances of great harm being committed by one group against another. Regrettably, while engaged in such harm doing, group members tend to perceive their own group's harmful actions as morally justifiable. Such legitimization protects people from the distress and guilt that they might otherwise experience when confronted with the harm done by their ingroup. Indeed, choosing not to legitimize the ingroup's actions and questioning its morality can be seen as traitorous acts. Thus, there are ample reasons to suppose that people will be strongly inclined to protect their social identity by perceiving their group's actions from the vantage point of the “moral high ground.”

Collective guilt, then, might seem to be an unlikely emotion to experience, given that its occurrence requires the self to be seen as a member of a group that has acted immorally. Despite the many impediments to doing so, around the world there have been instances where people have both publicly and privately debated whether their group's treatment of another group was justified or not. Such a reassessment of the legitimacy of the ingroup's actions can focus on historical injustices from long ago, as well as more recent and ongoing forms of intergroup harm doing. Feelings of collective guilt that are widely shared have resulted in formal apologies and other forms of reparations being issued to the harmed group as a means of making amends for the past and its continuing effects in the present.

Emotions can be ephemeral. How events are appraised and the subjective experience they generate can rapidly shift with changes in the social context. In order to capture people's emotional responses to events, psychologists have employed a variety of methods including the assessment of facial expressions (Ekman, 1984), physiological reactivity (Pennebaker, 1982) and, most commonly, self-report methods (Shaver et al., 1987). Particularly for the self-conscious emotions of pride, shame, and guilt, for which differential physiological symptoms are not expected, researchers have primarily relied on various participant self-assessments (Tangney & Fischer, 1995). Individual differences in the tendency to experience these emotions are generally assessed via self-ratings of the frequency, duration, or intensity with which they are felt. Indices that capture agreement or disagreement with Likert-type statements concerning how much the emotion is experienced with respect to a particular referent are also often employed. In this chapter, we report on the construction and validation of a self-report scale to assess various aspects of the collective guilt experience.

Theoretical bases of guilt

There is considerable agreement concerning the antecedent conditions that are necessary for personal guilt to be experienced. Weiner (1995) argues that guilt is most likely to be experienced when the self is seen as responsible for a negative outcome. From an attribution perspective, guilt occurs when an internal attribution is made for a controllable negative outcome that results in harm to another.

Inequality between women and men is perhaps the most pervasive form of group-based inequality facing humankind; it has a long history stretching back thousands of years (Johns, 1947; Seagle, 1947/1971) and exists to some degree in essentially all modern societies (Peterson & Runyan, 1993). In societies where gender inequality exists, its form is exclusively patriarchal with men having more power, resources, and status than women (Abel & Nelson, 1990; Beck & Keddie, 1978; Busch, 1990; Collier & Yanagisako, 1987; Keegen, 1993; Wood & Eagly, 2002). Women continue to be disadvantaged in the labor force compared to men (Burn, 1996; Kanter, 1977; Vasquez, 2001; Williams, 1992), even when controlling for qualifications, job type, and required skills (Bartol, 1999; Jacobs, 1995; Jacobs & Steinberg, 1990; Reskin & Padavic, 1994; Stroh, Brett, & Reilly, 1992). Furthermore, women are more likely than men to be poor (McClanahan & Kelly, 1999), engage in unpaid work (Shelton, 1999), and are the primary victims of domestic abuse and sexual assault (Rozee & Koss, 2001).

Individual men vary in the extent to which they are responsible for maintaining gender inequality. Furthermore, the form and extent of male privilege varies depending on social class, ethnicity, age, and sexual orientation (Connell, 1987). However, at the group level, gender inequality clearly benefits men at the expense of women. In this chapter, we consider a number of factors that are likely to influence the extent to which gender inequality will lead men to experience a sense of collective guilt.

Although more than a half century has passed since the liberation of the Nazi concentration camps, the extermination of some six million Jews remains the prototype of the violence that one group can commit against another group (Barkan, 2000; Staub, 1989). The extent to which contemporary Germans should and do continue to feel collective guilt for the Nazi genocide is a topic of much ongoing political debate (see Churchill, 1997; Goldhagen, 1996; Gross, 2001). Indeed, most of the chapters in this volume examine influences on the willingness of such perpetrator groups to accept guilt for their group's past. In contrast, in this chapter, we focus on the factors that lead historically victimized group members to assign or prescriptively desire that contemporary members of the perpetrator group experience collective guilt.

Given that collective guilt can be both accepted and assigned based on category membership alone, and does not require personal responsibility for the wrongdoing, social categorization processes should play a crucial role. We report on studies that empirically examine collective guilt assignment and forgiveness of contemporary Germans for the Holocaust among North American–Jewish people as a function of how the groups are categorized. We consider the consequences of perceiving Germans and Jews as distinct and separate groups, versus perceiving both groups as part of a single more inclusive superordinate category – that of humans.

When people in the Western World see pictures of children starving in the Third World, they may feel guilty about their own wealth and wonder about the causes of these differences. Indeed, the explanations for such social inequalities have been center stage in much social and political thought (Abernethy, 2001; Landes, 1999; Leach, Snider, & Iyer, 2002). For example, Daimond (1997) considered global differences in wealth in terms of people's natural environments. In addition to such natural causes, human behavior toward other humans has played an important role in the creation of worldwide differences in wealth (Brooks, 1999). Specifically, slavery and colonization practices have played a powerful role in increasing international inequality in wealth. Making salient such inequalities in wealth has the potential to trigger feelings of collective guilt in dominant group members. In this chapter, we are particularly interested in the consequences of ingroup identification for feelings of collective guilt.

Not all members of dominant groups experience collective guilt as a consequence of being confronted with ingroup-perpetrated immoral historical events. We argue that in order to understand why and when members of dominant groups experience collective guilt, the degree to which people identify with their national group needs to be taken into account. In this chapter, we focus mainly on the Dutch colonization of Indonesia and the resulting inequalities in wealth as a source of collective guilt among Dutch people.

National and other important social groups have histories. Reminders of group history can have important consequences for present-day emotional experience. When people consider some aspects of their group's history, they may feel collective pride and celebrate their group membership, while other aspects of their group's history may evoke collective guilt and a desire to correct the wrongs committed by the ingroup. Such emotional responses need not stem from personal participation in the events but can result when the self is categorized in terms of a shared group membership. Indeed, people's reactions to intergroup events are not usually based on having had a direct role in those events; rather such reactions will be dependent on the social identity that is contextually salient when people are reminded of their ingroup's past actions. Investigation of such group-based emotional experience is a relatively recent development within social psychology, but one that reflects a rapidly expanding interest in the interface between emotions and intergroup relations (see Cairns & Roe, 2003; Tiedens & Leach, 2004; Mackie & Smith, 2002; Pennebaker, Paez, & Rime, 1997).

In this volume, we focus on people's experience of one particular group-based emotional response – collective guilt. Collective guilt stems from the distress that group members experience when they accept that their ingroup is responsible for immoral actions that harmed another group (Branscombe, Doosje, & McGarty, 2002).