"I keep on thinking, but nothing happens."
- Jerome Howard
"I've always had the idea that my retirement would be the greatest
contribution to science the world has ever known."
- Julius Marx
"I'm on your side...But you're not."
- Milton Friedman
"I think my life is fuller because I realize that I don't know what I'm doing."
- Richard Feynman
My research focuses on contests using game theory, and is better known as rent-seeking models or tournaments. I apply this to marketing problems such as sales contests where asymmetries are present between employees and it is dependent upon the manager to modify the structure of the contest such that her goals will be achieved and simultaneously satisfy the goals of the employees. I currently have three working papers addressing this topic and can be found at SSRN and below.
Should a firm favor a weaker or stronger employee in a contest? Despite a widespread emphasis on rewarding the best employees,
managers continue to tolerate and even favor poor performers. Contest theory reveals that evenly matched contests are the most intense, which implies that a contest designer can maximize each player's effort by artificially boosting the underdog's chances. We apply this type of "handicapping" to a two-period repeated contest between employees, in which the only information available about their abilities is their performance in the first-period. In this setting, employees are strategic and forward looking, such that they fully anticipate the potential impact of the first-period contest result on the second period contest, and thus adjust their behaviors accordingly. The manager also incorporates these strategic behaviors of employees when determining an optimal handicapping policy. If employees' abilities are sufficiently different, favoring the first-period loser in the second period increases total effort over both periods. However, if abilities are sufficiently similar, we find the opposite result occurs: total effort increases the most in response to a handicapping strategy of favoring the first-period winner.
When a manufacturer advertises, what is the impact on retailer advertising? I analyze a contest model of advertising where total advertising by the manufacturer and by retailers determines market size, and the relative level of advertising by each retailer determines market share. If retailers are symmetric I show that there is a crowding-in effect so increased manufacturer advertising increases retail advertising. But if one retailer is stronger, then marginal increases in manufacturer advertising have a crowding-out effect on retailer advertising, while sufficiently large increases have a crowding-in effect by "jump-starting" competition between retailers for the larger market. Furthermore, asymmetric abilities in such contests can lead the weaker player to effectively drop out of the contest, thereby undermining the ability of increased prizes to increase effort by intensifying competition. More generally the model can be applied to other contests such as patent races or promotion tournaments where not just the probability of winning but also the value of winning depends on contest effort levels.
A fundamental result of contest theory is that evenly matched contests are fought most intensely, implying that a contest designer maximizes effort from each contestant by artificially boosting the chances of the underdog. Such "handicapping" of weak contestants is credited with making sports contests more exciting, increasing effort in job promotion contests, intensifying student competition to enter college, and raising revenues in auctions. But when information about ability is based on prior performance, handicapping seems to create a negative incentive effect for forward-looking contestants. In a repeated perfectly discriminating contest (i.e., an all-pay auction) we find that this is not the case -- favoring the loser of the first contest not only increases total effort in the second contest, but paradoxically also increases effort in the first contest. Although handicapping maximizes effort in the two contests, it makes the outcomes a less accurate measure of ability. When the objective is not to maximize effort but to identify the better contestant, a reverse handicapping strategy of favoring the winner of the first contest is optimal.