Peer Reviews of Economics Publications 
Editorial Board
Susan Athey
Ted Bergstrom
Michele Boldrin
V.V. Chari
Drew Fudenberg
David K. Levine
Thomas R. Palfrey
Wolfgang Pesendorfer
Matt Rabin
Ariel Rubinstein
Jose A. Scheinkman
Jean Tirole
ISSN 1558-4682
Volume 4 - March 05, 2002 Previous Next
1. Vasiliki Skreta Sequentially Optimal Mechanisms
You do not renounce selling a good just because the first round of bargaining failed. The literature on auctions under incomplete information assumes you would. If the first round fails, you committ to renounce selling the good forever. Skreta's paper looks at sequential mechanism design without this kind of committment. When designing today's mechanism for selling the good, you cannot commit to tomorrow's mechanism. Hence, the revelation principle cannot be applied. A characterization of the optimal dynamic incentive scheme for two-period problems without committment is provided. After characterizing the seller's problem for arbitrary agent types, the author shows that, in sequential bilateral bargaining, the optimal mechanism is to post a price each period.
original link cached copy reviewed by Michele Boldrin on 08/04/10
2. Jeffrey Ely and Jusso Valimaki Bad Reputation
This paper constructs a striking example of a game played by a long-run player against a sequence of short-run opponents. When the long-run player is known to be rational, then regardless of the player's discount factor these is an equilibrium that achieves the highest feasible payoff, while introducing a particular “bad” commitment type lowers the equilibrium payoff of a patient long-run player. Moreover, holding fixed the probability of the bad type, the equilibrium payoff of a patient long-run player is lower than its payoff in a one-time interaction.
original link cached copy reviewed by Drew Fudenberg on 03/28/02
3. Attila Ambrus Coalitional Rationalizability
Suppose that whenever it is of mutual interest for a group of players to avoid certain strategies, the members of the group will make an implicit agreement not to play them. This leads to an iterative procedure of restricting players' beliefs and action choices; the strategies that remain are called coalitionally rationalizable. In contrast to coalitional solution concepts based on the notion of Nash equilibrium, the set of coalitionally rationalizable strategies is always nonempty.
original link cached copy reviewed by Drew Fudenberg on 03/28/02
4. Ernst Maug and Bilge Yilmaz Two-Class Voting: A Mechanism for Conflict Resolution?
A group of agents must vote on a proposed policy. Agents have private information about the merits of the policy. The paper compares a simple voting rule with a "two-class" voting system. A simple voting rule requires k votes for the policy to be implemented. The two-class system partitions the agents into two groups and specifies a simple voting rule for each group. The policy is implemented if it is approved by both groups. The paper shows that two-class-voting aggregates more information if agents have sufficiently diverse preferences.
original link cached copy reviewed by Wolfgang Pesendorfer on 05/03/02
5. Itzhak Gilboa and David Schmeidler Inductive Inference: An Axiomatic Approach
An agent must rank the likelihood of eventualities based on a memory of past cases. For each memory, the agent is assumed to have a complete ranking. The paper provides axioms that yield the following representation: a weight is assigned to each case-eventuality pair and eventualities are ranked according to the sum of their weights (summed over all cases in memory). The key axiom asserts that if for two disjoint memories x is deemed more likely than y then the same ranking holds for the combined memory.
original link cached copy reviewed by Wolfgang Pesendorfer on 05/10/02