interactions among employees or between employees and supervisors. Unfortunately, this spillover effect produces dysfunctional consequences for the organization. As our data show in Table 2, suppressing norms tend to generate more negative individual outcomes from all forms of anger expression (38%) than settings with either contingent (30.5%) or legitimating norms (20.5%—see Table 2). If anger norms in these settings tend to encourage only silent expression of anger among coworkers, positive organizational benefits of authentic anger will be attenuated. If employees are socialized (in their professions and/or within the work unit) to curtail anger expression, they may be deprived of the requisite skills to handle anger-provoking situations that may occur in the context of their interactions with other employees in the work unit or the larger organization—leading to poor work unit performance.
To differentiate the various combinations of work unit norms, forms of expression and patterns of outcomes associated with anger in work units, we need a more encompassing theory that recognizes the value of both anger regulation and anger utilization in the service of organizational objectives. Focusing on a range of display rules offers a more comprehensive model to enable sufficient differentiation of work unit contexts. Further, within work units, theory needs to recognize that display rules may need to vary according to the nature of the task being undertaken. For tasks such as patient care and employee relations, cultivation of the ability to regulate anger is clearly desirable. Organizational decision making and problem-solving tasks may require a controlled form of anger to ensure that relevant data about organizations is not stifled to prevent individual hurt feelings at the expense of favorable work unit outcomes (see Edmondson, 1999; Maitlis & Ozcelik, 2004). And tasks requiring advocacy or rapid, honest, motivational feedback may warrant the use of legitimate anger expression targeted at the task, not individuals.
In Figure 2, we draw our various findings together into a model of the factors predicting the outcomes of anger expression in organizations. We identify four factors that influence whether and how individuals will express their anger at work: culture, socialization, personal experience and work unit anger norms. The model also reflects our findings concerning the interaction effects of work unit anger