Kyung Hwan Baik

Contests with group-specific public-good prizes

We examine the equilibrium effort levels of individual players and groups in contests in which n groups compete to win a group-specific public-good prize, the individual players choose their effort levels simultaneously and independently, and the probability of winning for each group depends on the groups’ effort levels. In the basic model, we show that, in each group, only the highest-valuation players expend positive effort and the rest expend zero effort; there is underinvestment in the contest for the group as a whole. Next, in the main model in which the players are budget-constrained, we show that low-valuation players free ride on high-valuation players’ contributions, not vice versa, but the free-rider problem is “alleviated” as compared with the basic model.

Bidding for a group-specific public-good prize

Abstract We examine the equilibrium effort levels of individual players and groups in a contest in which two groups compete with each other to win a group-specific public-good prize, the players choose their effort levels simultaneously and independently, and the winning group is determined by the selection rule of all-pay auctions. We first prove nonexistence of a pure-strategy Nash equilibrium, and then construct a mixed-strategy Nash equilibrium. At the Nash equilibrium, the only active player in each group is a player whose valuation for the prize is the highest in that group; all the other players expend zero effort; and the equilibrium effort levels depend solely on two values — the highest valuation for the prize in each group.

Reexamining efficient rent-seeking in laboratory markets

Given Millner and Pratts (1989) finding of behavior inconsistent with theory, we reexamine efficient rent-seeking in experimental markets. We show that (a) when r = 3, no Nash equilibrium exists and therefore behavior has no theoretical benchmark to judge consistency, and (b) when r = 1, with a new experimental design utilizing an explicit expected payoff matrix, rent-seeking behavior is consistent with both Nash equilibrium and dissipation hypotheses.

Delegation in contests

Abstract We consider contests with two players in which each player has the option of hiring a delegate who expends her effort to win a prize on his behalf. We first show when no delegation, unilateral delegation, or bilateral delegation occurs. Next, we demonstrate: (a) Whenever a player hires a delegate, he hires a delegate whose ability exceeds his own, and (b) when bilateral delegation occurs, the player with a higher valuation hires a delegate with more ability. Finally, we establish: As compared to the model without delegation, a total effort level is less when unilateral delegation by the player with a higher valuation or bilateral delegation arises, but it is greater when unilateral delegation by the player with a lower valuation arises.

Collective rent seeking with endogenous group sizes

Abstract We examine a contest in which two groups compete to win larger shares of a private-good rent, and where group size is endogenized by allowing inter-group mobility of group members. We also relax the restriction on the range of parameter values of the sharing rules adopted by Nitzan (‘Rent-seeking with non-identical sharing rules’, Public Choice 71, 43–50, 1991, and ‘Collective rent dissipation’, Economic Journal 101, 1522–1534, 1991) and Lee (‘Endogenous sharing rules in collective-group rent-seeking’, Public Choice 85, 31–44, 1995). We find that the two groups tend to be of equal size, and that the optimal sharing rules place great emphasis on relative outlays. The rent is ‘substantially’ dissipated in a collective contest with endogenous group sizes and sharing rules.

Two-stage rent-seeking contests with carryovers *

We examine a two-stage contest in which players in twogroups compete noncooperatively to win a rent. In thefirst stage, each group selects a finalist. The twofinalists compete for the rent in the second stage.First-stage efforts are carried over to the secondstage in the sense that they are partly effective inthe second stage as well. We show the following. Inthe case of player-specific carryovers, therent-dissipation rate increases in the carryover rate.With the carryover rate equal to one, the rent isfully dissipated. In the case of group-specificcarryovers, however, the rent-dissipation rate isindependent of the carryover rate.

Competitive-Share Group Formation in Rent-Seeking Contests

We examine the possibility of forming a competitive-share group in rent-seeking contests. The competitive-share group has the following characteristics: (a) the group is treated as a single player competing with the nonmembers to win the rent, (b) the members choose their outlays noncooperatively, and (c) the members compete to take larger shares of the rent. We show that the competitive-share group is always formed voluntarily. One important consequence of such group formation is that rent dissipation is less relative to individual rent seeking — in other words, the social cost associated with rent seeking is less.

Contests with spying

Abstract Contests between players with unknown relative ability are common. Under the plausible condition of decreasing aversion to uncertain ability, a player in a contest with unknown relative ability expends more resources than in a contest with known relative ability. If spying is allowed to reduce the uncertainty about ability, then in the case of unobservable spying, the level of spying is high, the level of effort is low, and total expected payoffs are high relative to the case of observable spying. Consequently, the social cost of the contest is low with unobservable spying as compared to observable spying.

Optimal punishment when individuals may learn deviant values

This paper extends the two-period model in Polinsky & Rubinfeld (1991) by introducing the possibility of social learning of illicit gains between the second and the first period. We show that if offenses result more from social learning of illicit gains than from an inherent characteristic of born offenders, it may be desirable to punish first-time offenders as severely as repeat offenders. This sheds light on deterring property crime, like larcency or auto theft, which exhibits a high degree of social learning. This may also provide a logical foundation for supporting harsh government coercion of loosely organized crime (such as a criminal street gang) in which the social learning is likely to be prevalent and illicit gains become a public good to co-offenders, from start to finish.

Endogenous Timing in a Gaming Tournament

This paper examines the theoretical background and actual behavior in a gaming tournament with endogenous timing where a person has more incentive, structure, and time to form a strategy. The baseline treatment suggests that subgame perfection is a reasonable predictor of behavior –- subjects made 170 of 208 theoretically predicted choices of best actions, with the majority of mistakes made in timing choices by the players who did not survive the cut to the second round. Four sensitivity treatments established that the design feature that lead to more predictable behavior was time to think –- 745 of 960 correctly predicted decisions with more time versus 595 of 960 with less time. A random effects Probit model suggests that the key design feature that closed the gap between predicted and observed behavior was not necessarily the non-linear payoffs created by the tournament design, but rather that the key was providing people with more time to think about their strategy.