Contingent reward versus punishment and compliance behavior: the mediating role of affective attitude and the moderating role of operational capabilities of artificial intelligence

Previous research has long focused on the positive effect of contingent reward on employees’ psychological states and outcomes. Recent studies show that contingent punishment can also positively impact employees’ outcomes. Whether contingent reward can improve employees’ compliance behavior more than punishment can, has not been adequately studied. Exploring this issue can help researchers understand contingent reward and punishment from the perspective of the relationship norms between leaders and employees. Drawing on the planned behavior theory and cognitive appraisal theory, we developed and tested a model specifying why and how contingent reward (relative to contingent punishment) increase employees’ compliance behavior. Using a scenario-based experiment across two-stage surveys of 309 participants, we found that contingent reward (relative to contingent punishment) could influence compliance behavior by increasing and decreasing distinct types of affective attitudes (i.e., self-esteem and anxiety). Moreover, the perceived operational capabilities of artificial intelligence (AI) moderated the effect of contingent reward on self-esteem and anxiety. High perceived operational capabilities of AI enhanced the positive effect of contingent reward (relative to contingent punishment) on self-esteem and its negative effect on anxiety as well. This research yields innovative insights for the improvement of compliance behavior.