Contest Theory In Research

Tournament or contest theory was introduced to research economists in 1981 when Edward P. Lazear and Sherwin Rosen published their paper on incentives in rank-order tournaments. Their research provides a theory on compensation schemes in the work environment and how they translate into incentives. Those incentives are thought to lead employees to expend resources (e.g. time, effort, investment, etc. or a combination of those), increasing their expected performance. They model a contest in which participants receive prizes relative to their final rank, regardless of the absolute difference in performance. The contestants independently and simultaneously commit to expending a level of resources, while the model accounts for the effect of a random

The contest success function (CSF) states that winning probabilities are generated by relating the contestants’ exerted resources to the sum of all resources spent in the contest. Following intuition, homogeneous participants spending the same amount of resources have equal winning probabilities. If, ceteris paribus, a contestant increases the amount of resources spent, his or her winning probability increases while the opponents’ cumulative winning probability decreases. Tullock’s CSF is completed by a measure for the discriminatory power of the contest. In a Tullock contest, the contestants face a participation constraint (PC) and an incentive compatibility constraint (ICC). The PC simply shows that the contestants’ valuation of the prize has to be positive and different from zero. The ICC of the Tullock contest states that the contestants’ expected payoff from the equilibrium level of resources is his or her maximum expected payoff at the same time. Heterogeneity between players in a Tullock contest can be modeled either with various valuations of the prize or different cost functions. For (More below!) to an effort level up-front, not knowing who (and how many) they will compete against. ys has a significant effe the simple case of two contestants, the maximization problem for resources spent depends on player heterogeneity and the discriminatory power of the contest. In that case,

before choosing an effort level) are equal for perfectly homogeneous contestants, but symmetric contests are rare in reality. According to Szymanski (2003), differences in abilities are usually modeled with varying cost functions or prize valuations. Their effects on effort vary depending on contest design. For brevity, this paragraph will focus on a one-period contest with two contestants. Notice that heterogeneity, when sufficiently large, can violate the weaker players’ PC causing him to drop out. Theoretical literature states that heterogeneity has a negative impact on effort. This can be intuitively summarized in the so-called “contamination hypothesis” or “discouragement effect”. Underdogs anticipate their low likelihood to win and consequently reduce effort to minimize costs. The favorite is aware of that and reduces effort to increase efficiency (without compromising the role as a favorite), resulting in an overall performance decrease. According to Dechenaux, Kovenock & Sheremeta (2015), the absolute effect on effort varies across the different models and contest setups. In an all-pay auction, increasing the spread between the contestants’ valuation of the prize (i.e. increased heterogeneity) has a negative effect on aggregate effort expenditure. The effort expenditure of the favorite stays constant while the weaker player chooses to decrease effort. In a Tullock contest, an increase in heterogeneity