UC Riverside

This dissertation includes three essays on players' incentives to acquire or share private information in competitive environments.

In chapter 2, I study the role of one-sided private information in a two-player first-price all-pay auction where one player's valuation of prize is common knowledge while the other's is privately known. Then I study the incentives of spying in cases where spying (i) can be caught with an exogenous probability and (ii) can be fed false information after caught. I find that spying can be discouraged by increasing the chance of catching a spy, and spying may even be completely deterred when false information is fed back after catching a spy.

In chapter 3, I study the incentives to form a quid pro quo information sharing agreement between ex ante symmetric players ahead of contests. I find that a limited-membership alliance, which includes a strict subset of all players, may arise even in the presence of small organization costs while an industry-wide sharing agreement may not. Such an alliance can strictly benefit alliance members, but may benefit or hurt the outsider. Even when the outsider is hurt, a Pareto improvement is possible if transfers can be arranged between alliance members and non-alliance members.

In chapter 4, I extend the analysis of chapter 3 to a first-price auction. I find that a limited-membership information-sharing alliance can lead to Pareto improvements when players' private values take three discrete types. Furthermore, the player outside the alliance benefit strictly more than alliance players from the alliance.