MICRO THEORY: Professor Rene Kirkegaard (University of Guelph)

Log-Supermodular Contests

A contest regulator is unlikely to know the exact primitives. Nevertheless, data from past contests may reveal which agent is the favorite and which is the underdog. The value of such information is examined by utilizing a large and unifying family of noisy contests. The defining characteristic is that the marginal return to effort exhibits a log-supermodularity property. With homogeneous technologies, robust comparative statics include collective discouragement and disparity effects. However, there is limited scope for robust regulation. Any regulation that does not exclusively benefit the underdog risks backfiring. These conclusions extend to a subset of contests with heterogeneous technologies.