Experimentation and Signal Dampening in Agency--The Ratchet Effect Revisited
Thomas D. Jeitschko
Michigan State University - Department of Economics
Leonard J. Mirman
affiliation not provided to SSRN; University of Virginia - Department of Economics
Egas M. Salgueiro
Universidade de Aveiro, S.A.G.E.I.
The dynamics of incentive contracts under asymmetric information have long been an important topic in economics. We address this topic in this paper by considering a stochastic, two-period principal-agent relationship, in which the true state of the world can take on two possible values and is the same in each period. We study contracts that are short term, so that after the first period the principal designs a second contract, taking the information available about the state of the world at that stage into account--that is, the standard framework in which the "ratchet effect" occurs.
Two significant changes emerge when compared to deterministic environments: First, if production is sufficiently noisy, a fully separating first period contract is optimal. The second change is that, unlike the deterministic setting where the high type's target is fixed over time, in a stochastic environment, the high type's target is ratcheted upward in the course of the interaction. This is the result of two opposing incentives of the principal: First, the principal experiments in order to increase the flow of information; and second, the principal attempts to dampen the first period signal to reduce up-front payments.
Number of Pages in PDF File: 25
JEL Classification: D8, L5, H57
Date posted: March 8, 1999