Baumeister et al.'s main proposition is that for groups to perform well members need to differentiate themselves from the group. Individual differentiation can be important for group performance, but Baumeister and colleagues overestimate its importance and, at the same time, underestimate the importance of identification.

That the importance of differentiation is overestimated becomes particularly clear in the discussion of social loafing, free riding, and failure to cooperate which, presumably, occurs because individual contributions are not identifiable and group members see their contributions as dispensable rather than unique or essential. Baumeister and colleagues suggest that individual differentiation will reduce these problems. This ignores the primary reason that group members do not contribute to the group: Members act out of individual self-interest. Indeed, social loafing in task groups and noncooperation in social dilemmas, two crucial domains of group performance in Baumeister et al.'s analysis, share as underlying causes: (1) contributing to group efficiency and success is costly for the individual, (2) group members value personal self-interest more than collective success and group efficiency – they lack cooperative motivation – and (3) members can get away with selfish actions because of, for example, anonymity and lack of accountability.

Consider, as one example, the wealth of studies using a broad range of tasks, including pulling a rope (Ingham et al. Reference Ingham, Levinger, Graves and Peckham1974; Ringelmann Reference Ringelmann1913a), shouting as loudly as possible (Latané et al. Reference Latané, Williams and Harkins1979; Williams et al. Reference Williams, Harkins and Latané1981), or puffing air into a mouthpiece (Kerr & Bruun Reference Kerr and Bruun1983). Although not inherently interesting, these tasks require individual effort, and contributing thus is individually costly. This work shows that social loafing emerges only when contributions are costly (and individuals are not accountable), not when tasks are inherently interesting or personally involving (e.g., Brickner et al. Reference Brickner, Harkins and Ostrom1986; Smith et al. Reference Smith, Kerr, Markus and Stasson2001; Zaccaro Reference Zaccaro1984; see also Karau & Williams Reference Karau and Williams1993). Furthermore, loafing is reduced or eliminated when group cohesiveness is high, when members identify highly with the group, and when members operate under collectivist rather than individualistic (cultural) norms (Karau & Williams Reference Karau and Williams1993). Thus, group members who are concerned with group efficiency or collective success do not loaf, regardless whether they are identifiable, accountable, or indispensable. It is those individuals who focus on their self-interest, rather than on group goals, who differentiate too much rather than not enough, who free ride and detract from rather than contribute to the group.

Some key “pathologies of groups” thus result from group members' pursuit of self-interest rather than from a lack of individual differentiation. Because this goes unnoticed, Baumeister et al.'s analysis becomes confusing. In their analysis, anonymity can both reduce individual differentiation (as in social loafing and social dilemmas) but also increase it (as in reducing conformity), which is odd to say the least. A better way to summarize the state-of-the art may be to conjecture that anonymity (1) increases the temptation and ability to pursue self-interested goals, which results in (2) both enhanced free riding and reduced conformity. Such a conjecture would fit the well-established observation that effects of individual differentiation on group functioning and performance are strongly contingent on task type (McGrath Reference McGrath1984; Steiner Reference Steiner1972). To illustrate, we use Baumeister and colleagues' distinction between productive achievement and information use tasks. In most productive achievement tasks, group performance depends mainly on member effort, and contributions depend on the extent to which individuals make self-interest subordinate to group interest. Identification helps; differentiation hurts.

In information tasks, however, group members may have unique resources – such as information, perspectives, and ideas – and need to share them with others to ensure high group performance. Here, conformity detracts from group performance, and individual differentiation may promote group success. Even in such information tasks, however, the pivotal role of group identification cannot and should not be underestimated. In fact, the issue is not whether individuals identify or differentiate, but rather that individuals need to both identify and differentiate (as in I want to contribute my unique resources to the group). If group members act out of pure self-interest, rather than in the interest of the group, differentiation typically is detrimental to group performance (De Dreu et al. Reference De Dreu, Nijstad and van Knippenberg2008). Under these conditions, differentiation gives rise to power struggles, strategic behavior, and lying and deception. Baumeister et al. do not acknowledge these downsides to differentiation.

With regard to group creativity, for example, it has indeed been found that individual differentiation stimulates group creative performance (Goncalo & Staw Reference Goncalo and Staw2006), but this effect holds only for groups that have shared goals, and it is induced via a collective reward for creative achievement (Bechtoldt et al. Reference Bechtoldt, Choi and Nijstad2012). Likewise, voicing dissent benefits creativity and innovation in groups and organizational teams but only when group members have cooperative goals and strong identification (De Dreu Reference De Dreu2007; De Dreu & West Reference De Dreu and West2001; Nijstad et al. Reference Nijstad, Berger-Selman and De Dreu2014). Thus, even in information-use tasks, where individual differentiation may benefit group performance, strong identification and concern with group goals is a necessary precondition for groups to function and perform. Differentiation always needs to build on the foundation of shared and valued group goals, and concomitant group identification.

A more viable perspective to the one Baumeister and colleagues offer proceeds on the basis of the assumption that identification and differentiation jointly enhance or reduce group performance (De Dreu et al. Reference De Dreu, Nijstad and van Knippenberg2008; Nijstad & De Dreu Reference Nijstad and De Dreu2012). Such a perspective entails that (1) differentiation is beneficial only in more cognitive tasks such as group judgment and decision-making, creativity, or negotiation, (2) cooperative motivation (and resulting identification) is a sine qua non for the benefits of differentiation to be reaped, and (3) in such more-cognitive tasks, the benefits of differentiation materialize only when members are both required and motivated to process information thoroughly.