Baumeister et al. present a new, nuanced version of the old argument that “groups are bad for you” (Brown Reference Brown1988). They acknowledge that strong group identity creates cohesion and empowers members to achieve things they could never achieve alone. But, they argue, strong group identity promotes mindless conformity, removes accountability and thereby subverts reason, efficiency and morality. The sole solution is to reintroduce differentiated individual identification into groups.
Historically, such views were first articulated but the elites who observed from afar and with horror the mass disruption in the late nineteenth century (e.g., Le Bon Reference Le Bon1896/1960; for a review see Van Ginneken Reference Van Ginneken1992). Similar fears of seemingly mindless mobs still exist today (Reicher & Stott Reference Reicher and Stott2011). So it comes as no surprise that Baumeister and colleagues use crowds to epitomize “some of the worst, most vicious and destructive tendencies of groups” (sect. 2.3.2, para. 1). Nor is it surprising that they draw on Le Bonian theory and its modern derivative, deindividuation research, to buttress their argument. However, in doing so, they mischaracterise the evidence in ways that have serious implications for their overall argument.
First, the authors ignore half a century of crowd research that spans the disciplines of history, political science, anthropology, and psychology. Close analysis of the actual phenomena shows the traditional notion of mindless mobs to be a myth (e.g., Davis Reference Davis1973; McPhail Reference McPhail1991; Reicher Reference Reicher, Hogg and Tindale2001; Turner & Killian Reference Turner and Killian1957). Evidence from contemporary and historical events show that even a violent crowd typically acts in a highly patterned way that reflect collective norms, values, and beliefs. Crowds do not lack morality or ignore reason. Rather they subscribe to an alternative moral economy to that endorsed by those in authority (cf., Thompson Reference Thompson1971). Such studies also show how the act of representing this alternative morality as a lack of morality was an ideological act designed to discredit dissent and protect the status quo (Giner Reference Giner1976).
Second, the authors misrepresent research on deindividuation processes (which refers to the consequences of losing one's individual identity through becoming an anonymous part of the group). Baumeister and colleagues cite individual studies that show a relationship between anonymity and various forms of antisocial behavior. They also cite our meta-analysis, which reviews 60 studies examining these effects (Postmes & Spears Reference Postmes and Spears1998). They suggest the results show that “[t]he primary effect of deindividuation was to reduce accountability, especially in enabling people to take illicit selfish benefits (e.g., cheating, stealing)” (sect. 2.3.2, para. 2).
Notwithstanding the fact that the latest statements of deindividuation theory actually distinguish this conscious accountability route from deindividuation per se (e.g., Prentice-Dunn & Rogers Reference Prentice-Dunn, Rogers and Paulus1989), this is not what the empirical evidence shows. “Deindividuation” sometimes increases and sometimes decreases antisocial behavior. Overall, the effects are close to zero. Moreover, when groups are more antisocial, it is not the result of a loss of individual identity per se, but rather because of the specific norms of the groups with which people identify (Postmes & Spears Reference Postmes and Spears1998). Considering the amount of evidence that has been amassed on these points, we can confidently conclude that the picture of deindividuation Baumeister et al. paint is a myth. That is exactly the conclusion drawn in a prior narrative review (Diener Reference Diener and Paulus1980).
Anonymous members of groups neither lose accountability nor act selfishly. The key point is that there is a shift whereby the group rather than the individual becomes the basis of both accountability and interest. We become accountable for our performance as group members; we act for the collective gain rather than our personal benefit. The broader point here is that anonymity in the group in itself does not make us thoughtless or antisocial (indeed our meta-analysis shows anonymity increases adherence to local group norms). We deliberate on the basis of group level concerns and hence the nature of our actions will depend on group norms.
Furthermore, because groups always exist in intergroup relations, concepts such as anonymity and accountability need to be seen in relational, not absolute terms. The issue is not whether we are anonymous or accountable, but rather to whom we are anonymous and to whom we are accountable. So, in a conflict between crowd members and the police, say, whether people are more or less anonymous and accountable to the out-group (the police) or to the in-group (other crowd members) will be critical to how they act.
These multidimensional and relational dynamics have been studied widely in the SIDE tradition (Social identity model of deindividuation effects – see Klein et al. Reference Klein, Spears and Reicher2007; Reicher et al. Reference Reicher, Spears and Postmes1995; Spears & Lea Reference Spears and Lea1994). This literature shows that deindividuation conditions may influence not just whether norms influence behavior, but also which norms do so. That is, how do different forms of identifiability impact whether people feel inclined and enabled to act on the basis of in-group norms or else constrained to act within the terms of out-group norms?
So where does all this leave us in terms of Baumeister et al.'s overall argument? We contest the notion that strong social identity necessarily leads to lack of reason, lack of accountability, and hence “bad” behavior. We suggest that strong social identity does not remove rationality and morality but rather provides them with a different, a group basis (Spears Reference Spears2010). Moreover, strong social identity empowers group members to enact their own views, whatever these might be.
Put differently, a rounded survey of the evidence shows that the problems of “bad” groups do not lie in a generic “bad” group psychology but rather in specific “bad” group norms. Violent groups normatively validate violent action. Conformist groups normatively invalidate critical comment. The solution to problematic behavior of crowds and groups, then, is not to fragment or individualize the group. To do that is to disempower people and hence to neuter alternatives to the status quo. The solution to problematic behavior of crowds and groups is to challenge and change toxic group norms.
Baumeister et al. present a new, nuanced version of the old argument that “groups are bad for you” (Brown Reference Brown1988). They acknowledge that strong group identity creates cohesion and empowers members to achieve things they could never achieve alone. But, they argue, strong group identity promotes mindless conformity, removes accountability and thereby subverts reason, efficiency and morality. The sole solution is to reintroduce differentiated individual identification into groups.
Historically, such views were first articulated but the elites who observed from afar and with horror the mass disruption in the late nineteenth century (e.g., Le Bon Reference Le Bon1896/1960; for a review see Van Ginneken Reference Van Ginneken1992). Similar fears of seemingly mindless mobs still exist today (Reicher & Stott Reference Reicher and Stott2011). So it comes as no surprise that Baumeister and colleagues use crowds to epitomize “some of the worst, most vicious and destructive tendencies of groups” (sect. 2.3.2, para. 1). Nor is it surprising that they draw on Le Bonian theory and its modern derivative, deindividuation research, to buttress their argument. However, in doing so, they mischaracterise the evidence in ways that have serious implications for their overall argument.
First, the authors ignore half a century of crowd research that spans the disciplines of history, political science, anthropology, and psychology. Close analysis of the actual phenomena shows the traditional notion of mindless mobs to be a myth (e.g., Davis Reference Davis1973; McPhail Reference McPhail1991; Reicher Reference Reicher, Hogg and Tindale2001; Turner & Killian Reference Turner and Killian1957). Evidence from contemporary and historical events show that even a violent crowd typically acts in a highly patterned way that reflect collective norms, values, and beliefs. Crowds do not lack morality or ignore reason. Rather they subscribe to an alternative moral economy to that endorsed by those in authority (cf., Thompson Reference Thompson1971). Such studies also show how the act of representing this alternative morality as a lack of morality was an ideological act designed to discredit dissent and protect the status quo (Giner Reference Giner1976).
Second, the authors misrepresent research on deindividuation processes (which refers to the consequences of losing one's individual identity through becoming an anonymous part of the group). Baumeister and colleagues cite individual studies that show a relationship between anonymity and various forms of antisocial behavior. They also cite our meta-analysis, which reviews 60 studies examining these effects (Postmes & Spears Reference Postmes and Spears1998). They suggest the results show that “[t]he primary effect of deindividuation was to reduce accountability, especially in enabling people to take illicit selfish benefits (e.g., cheating, stealing)” (sect. 2.3.2, para. 2).
Notwithstanding the fact that the latest statements of deindividuation theory actually distinguish this conscious accountability route from deindividuation per se (e.g., Prentice-Dunn & Rogers Reference Prentice-Dunn, Rogers and Paulus1989), this is not what the empirical evidence shows. “Deindividuation” sometimes increases and sometimes decreases antisocial behavior. Overall, the effects are close to zero. Moreover, when groups are more antisocial, it is not the result of a loss of individual identity per se, but rather because of the specific norms of the groups with which people identify (Postmes & Spears Reference Postmes and Spears1998). Considering the amount of evidence that has been amassed on these points, we can confidently conclude that the picture of deindividuation Baumeister et al. paint is a myth. That is exactly the conclusion drawn in a prior narrative review (Diener Reference Diener and Paulus1980).
Anonymous members of groups neither lose accountability nor act selfishly. The key point is that there is a shift whereby the group rather than the individual becomes the basis of both accountability and interest. We become accountable for our performance as group members; we act for the collective gain rather than our personal benefit. The broader point here is that anonymity in the group in itself does not make us thoughtless or antisocial (indeed our meta-analysis shows anonymity increases adherence to local group norms). We deliberate on the basis of group level concerns and hence the nature of our actions will depend on group norms.
Furthermore, because groups always exist in intergroup relations, concepts such as anonymity and accountability need to be seen in relational, not absolute terms. The issue is not whether we are anonymous or accountable, but rather to whom we are anonymous and to whom we are accountable. So, in a conflict between crowd members and the police, say, whether people are more or less anonymous and accountable to the out-group (the police) or to the in-group (other crowd members) will be critical to how they act.
These multidimensional and relational dynamics have been studied widely in the SIDE tradition (Social identity model of deindividuation effects – see Klein et al. Reference Klein, Spears and Reicher2007; Reicher et al. Reference Reicher, Spears and Postmes1995; Spears & Lea Reference Spears and Lea1994). This literature shows that deindividuation conditions may influence not just whether norms influence behavior, but also which norms do so. That is, how do different forms of identifiability impact whether people feel inclined and enabled to act on the basis of in-group norms or else constrained to act within the terms of out-group norms?
So where does all this leave us in terms of Baumeister et al.'s overall argument? We contest the notion that strong social identity necessarily leads to lack of reason, lack of accountability, and hence “bad” behavior. We suggest that strong social identity does not remove rationality and morality but rather provides them with a different, a group basis (Spears Reference Spears2010). Moreover, strong social identity empowers group members to enact their own views, whatever these might be.
Put differently, a rounded survey of the evidence shows that the problems of “bad” groups do not lie in a generic “bad” group psychology but rather in specific “bad” group norms. Violent groups normatively validate violent action. Conformist groups normatively invalidate critical comment. The solution to problematic behavior of crowds and groups, then, is not to fragment or individualize the group. To do that is to disempower people and hence to neuter alternatives to the status quo. The solution to problematic behavior of crowds and groups is to challenge and change toxic group norms.