The effects of tournament horizon and the percentage of winners on social comparisons and performance in multi-period competitions

We examine the effects of two important tournament design features, tournament horizon and percentage of winners, on social comparisons and performance in a multi-period setting. Prior research has individually examined these two features, but has not examined their joint effects. Replicating prior research, we predict that a higher percentage of winners will result in better performance than a lower percentage of winners by inducing more social comparisons. We also predict that grand tournaments will result in better performance than repeated tournaments through their effects on the extent to which competitors will engage in social comparisons. We expect that relative to repeated tournaments, grand tournaments will encourage more social comparisons because the performance feedback provided to competitors will be more indicative of the likelihood of future period outcomes (i.e., winning or losing) than in a repeated tournament. We also examine the extent to which the percentage of winners moderates the strength of the predicted relations among tournament horizon, social comparisons and performance. Finally, we predict that in repeated tournaments using a higher percentage of winners will be more effective at sustaining effort in later periods but that these effects will be weaker for grand tournaments. Results from a laboratory experiment with 400 undergraduate student participants support these predictions. Moreover, we find no evidence that the percentage of winners influences the impact of tournament horizon on social comparisons or overall performance. We identify implications for theory and practice.