Do Financial Incentives Affect the Quality of Expert Performance? Evidence from the Racetrack

Abstract

Does the quality of performance by experts respond to financial incentives? Or as some psychologists argue are experts primarily motivated by more intrinsic consideration such as professional pride? I provide some evidence on this question by examining the relationship between horse race outcomes and the level of race prize money. If financial incentives are important then races with low prize money are more likely to see some trainers exert less than full effort thereby upsetting the calculations of race bettors. By contrast races with high prize money are less likely to be affected by unobservable variation in trainer effort so bettor odds should then be a more reliable predictor of race outcomes. In a sample of 30426 horse races I find evidence consistent with this story: average bettor payoffs in a variety of betting pools are strongly negatively related to race prize money and the probability of a bettor-favourite horse succeeding is strictly increasing in the amount of prize money at stake. These results continue to hold when I exclude low-information races from the sample thereby suggesting that prize money is not acting as a proxy for the quantity of information publicly available to bettors. As a group horse trainers apparently tailor the quality of their performance to the potential size of their payout from clients.