Rankings, Random Successes, and Individual Performance

Rankings have become increasingly important over the past decades and impose a sharp distinction between success and failure. In this paper we examine the effects of ranking positions and great successes on individual performance by using a rich set of data on World Cup alpine ski races for the period of 1992-2013. We apply a regression discontinuity design and exploit close races as a source of randomized treatment. Our results suggest substantial short-run effects of randomly assigned podium finishes on performance, especially for racers in the middle of the skill distribution. However, the effects are short-lived and mostly driven by individuals who miss prestigious ranks by a tiny margin. We identify media attention as the key channel for performance effects and provide empirical evidence for an increasing media bias in favor of top-ranked competitors in the last two decades. These findings highlight a serious drawback of rankings.