meso level in work units. Focusing on the full range of anger expressions rather than solely on intra-individual anger regulation will enable exploration of the relative balance among different levels and valences of anger outcomes.
Unlike prior work, we consider outcomes of anger expressions as a product of both the nature of the anger expressions and nature of the work unit anger norms. This systematic analysis of result patterns may be a useful tool in future research that examines anger and its outcomes. Using the valence by level anger outcome framework, we explored correspondence or mismatch between work unit norms and forms of anger expression. While one would expect norms to constrain forms of expression (authentic, controlled, or silent), we suggest that a fit between norms and form of expression results in more positive outcomes. Our findings in Table 2 show, for example, that 33% of positive outcomes occurred in legitimating settings with authentic expressions and 34% with controlled expressions. When a mismatch occurred, such as authentic anger expressions in suppressed settings, only 11% of outcomes were positive. Our analysis thus shows important relationships between anger norms, type of anger expression, and outcomes. Taken together our results both extend and challenge emotional labor theory and research and suggest that we may need a broader theoretical lens on how emotion is manifest in organizations and how its expression is related to organizational performance. While emotion regulation was clearly evidenced in some of the work units we investigated, the specific suppression of negative emotion that is often associated with the idea of emotional labor (Rafaeli and Sutton, 1987) was neither observed nor was it even characterized as desirable in all of our contexts. Settings characterized by contingent norms, while recognizing that it was beneficial to regulate employee anger in front of those the organization served, also made allowances for employees to express anger in crisis circumstances or where the expression furthered the goals and values of the organization. And in contexts with norms legitimating anger expression, anger was viewed as utilitarian and an essential part of the work.
Additionally, our data suggest that in settings with strong anger regulation (to serve client needs), there may also be a tendency for spillover of these norms to other aspects of the work, such as in