Essays on experimental investigation of lottery contests

Roman M Sheremeta, Purdue University

Abstract

A contest is a situation in which individuals or groups expend costly resources while competing to win a specific prize. The variety of economic situations that can be described as contests has attracted enormous attention from economic theorists. Despite the extensive theoretical research of contests, very little empirical research has been done to evaluate the theory. This dissertation uses experimental methods to provide empirical investigation of different aspects of contest theory. The dissertation consists of four independent essays. The first essay experimentally compares the performance of four simultaneous lottery contests: a grand contest, two multiple prize settings (equal and unequal prizes), and a contest which consists of two sub-contests. Consistent with the theory, the grand contest generates the highest effort levels among all simultaneous contests. In multi-prize settings, equal prizes produce lower efforts than unequal prizes. The results also support the argument that joint contests generate higher efforts than an equivalent number of sub-contests. The second essay experimentally studies a two-stage elimination contest and compares its performance with a one-stage contest. Contrary to the theory, the two-stage contest generates higher total effort expenditures than the equivalent one-stage contest. The third essay investigates the performance of a two-stage elimination contest with effort carryover. Experimental results support all major theoretical predictions: the first stage effort and the total effort expenditures increase in the carryover rate, and the second stage effort decreases in the carryover rate. Consistent with other experimental studies, there is significant over-dissipation of efforts relative to the equilibrium prediction in all contests. The first essay argues that this over-dissipation can be partially explained by strong endowment size effects. Subjects who receive bigger endowments tend to over-dissipate, while subjects who receive smaller endowments tend to under-dissipate. This behavior is consistent with the predictions of a quantal response equilibrium. The second and third essays provide evidence that winning is a component in a subject's utility and that non-monetary utility of winning is an important factor to explain over-dissipation in contests. The final essay investigates contests between groups. Each group has one strong player, with a higher valuation for the prize, and two weak players, with lower valuations. In contests where individual efforts are perfect substitutes, both strong and weak players expend significantly higher efforts than predicted by theory. In best-shot contests, where group performance depends on the best performer within the group, most of the effort is expended by strong players while weak players free-ride. In weakest-link contests, where group performance depends on the worst performer within the group, there is almost no free-riding and all players expend similar positive efforts.

Degree

Ph.D.

Advisors

Cason, Purdue University.

Subject Area

Economics

COinS