, the action of one agent -selection of treatment version -may affect the actions of another agent, with the resulting strategic interference complicating the evaluation of agents. An incentive-compatible experiment design is one with an equilibrium where each agent selects its natural action, which

In this paper, we take a novel approach at service differentiation in packet-switched networks. Existing mechanisms for delay control and differentiation typically require some form of explicit resource allocation and consequently, admission control. We propose incentive-compatible differentiated

is true in the long-run. The optimal contract is given by a combination of both aid instruments. Incentive-compatible contracts require that less patient and less benevolent recipient governments receive more grants but less concessional loans.

This paper considers nonlinear corporation income tax schedules as a means of ensuring incentive compatible investment, when rms have private information about their total factor productivity level. The tax schedule species an income tax rate conditional on the corporation's taxable income

Abstract We initiate the study of incentives in a general machine learning framework. We focus on a game-theoretic regression learning setting where private information is elicited from multiple agents with different, possibly conflicting, views on how to label the points of an input space

An incentive compatible reputation mechanism for ubiquitous computing environments

In the presence of self-interested parties, mechanism designers typically aim to achieve their goals (or socialchoice functions) in an equilibrium. In this paper, we study the cost of such equilibrium requirements in terms of communication, a problem that was recently raised by Fadel and Segal [14]. While a certain amount of information x needs to be communicated just for computing the outcome of a certain social-choice function, an additional amount of communication may be required for computing the equilibrium-supporting prices (even if such prices are known to exist). Our main result shows that the total communication needed for this task can be greater than x by a factor linear in the number of players n, i.e., n · x. This is the first known lower bound for this problem. In fact, we show that this result holds even in single-parameter domains (under the common assumption that losing players pay zero). On the positive side, we show that certain classic economic objectives, namely, single-item auctions and public-good mechanisms, only entail a small overhead. Finally, we explore the communication overhead in welfare-maximization domains, and initiate the study of the overhead of computing payments that lie in the core of coalitional games. 1

Financial support from the Ford Foundation is gratefully acknowledged. To be published in a conference issue of the

anonymous referees, a Co-Editor and numerous conference and seminar participants for valuable comments.

choice are called incentivecompatible when cooperative behavior is optimal for all parties. The most well-known examples of incentivecompatible mechanisms are auctions. However, the party that receives the auction revenue has an incentive to manipulate the outcome to increase the revenue. For example, a