Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
In many situations, agents take risks by choosing an action that increases their performance immediately, but that potentially leads to a large loss. The current paper studies how such risk-taking behavior depends on the level of competition that the agents face. We study a tournament model and we find that more intense competition, measured by the number of competitors as well as their relative standing, induces agents to take higher risks. We use a rich panel data set on professional biathlon competitions as well as survey data from professional biathletes to confirm the model predictions. Finally, we discuss implications for organizational decision-making.