When Does Rivalry Drive Performance?

Existing rivalry research finds that rivalry relationships, based in past interactions rather than current levels of competition, can increase performance in competitive settings. We extend rivalry theory by integrating it with 1) research on performance pressure and arousal, and 2) regulatory fit theory and research on status changes, leading us to propose two important moderators of rivalry, one individual (skill) and the other situational (potential for status change). We test these ideas using data from computer programming contests involving over 10.6 million competitive encounters across 63,220 software developers (Æcoders’). We find that, on average, coders who are randomly assigned to compete against others with whom they share a rivalrous history exhibit higher performance, above and beyond the effects of current levels of competition. Importantly, however, this is moderated by 1) coders’ skill level, such that rivalry is more beneficial for more skilled coders and harmful for less-skilled coders, and 2) coders’ risk of experiencing a status change, such that coders who face a possible status loss exhibit decreased performance when competing against rivals. Thus, rivalry can harm performance under certain conditions, which is vital to understanding its role in organizations.