This website uses cookies.

By clicking the "Accept" button or continuing to browse our site, you agree to first-party and session-only cookies being stored on your device to enhance site navigation and analyze site performance and traffic. For more information on our use of cookies, please see our Privacy Policy.

  1. Home
  2. ASSA Annual Meeting
  3. 2022
  4. Preliminary Program
  5. Models of Experimentation and Reputation
« Back to Results

Models of Experimentation and Reputation

Paper Session

Friday, Jan. 7, 2022 10:00 AM - 12:00 PM (EST)

Hosted By: Econometric Society
  • Chair: Simon Board, University of California-Los Angeles

Reputation Building under Observational Learning

Harry Pei
,
Northwestern University

Abstract

A patient seller interacts with a sequence of myopic consumers. Each consumer decides whether to trust the seller after observing a bounded number of the seller's past actions and some or all previous consumers' choices. With positive probability, the seller is a type that commits to play his Stackelberg action. I show that consumers' ability to observe other consumers' choices can lead to reputation failures: There exist equilibria where every consumer imitates her predecessor with high probability and the patient seller receives his minmax payoff. Furthermore, the seller receives his minmax payoff in all equilibria where consumers do not trust him in the first period and do not trust him when the worst action profile occurred in the period before. In an extension where every consumer observes all previous consumers' choices and an unboundedly informative private signal about the seller's current-period action, the seller receives at least his Stackelberg payoff in all equilibria.

Early-Career Discrimination: Spiraling or Self-Correcting?

Arjada Bardhi
,
Duke University
Yingni Guo
,
Northwestern University
Bruno Strulovici
,
Northwestern University

Abstract

Do workers from social groups with comparable productivity distributions obtain comparable lifetime earnings? We study how a small amount of early-career discrimination propagates over time when workers’ productivity is revealed through employment. Breakdown learning environments that track on-the-job failures grant a disproportionately large advantage to marginally more favored groups, whereas breakthrough learning environments that track successes guarantee comparable earnings to groups of comparable productivity. This discrepancy persists with large labor markets, flexible wages, inconclusive signals, and misspecified employer beliefs. Allowing for investment in productivity exacerbates inequality between groups under breakdown learning.

Incentives and Performance With Optimal Money Management Contracts

Stefano Pegoraro
,
University of Notre Dame

Abstract

I characterize the dynamics of incentives in an optimal contract with investment delegation, moral hazard, and uncertainty about the agent's productivity. The principal increases the agent's incentives after good performance in order to delegate more capital to an agent with higher perceived productivity, thus implementing a convex pay-for-performance scheme. Moreover, the principal commits to reduce the agent's future incentives in order to mitigate ex-ante investment distortions. Methodologically, I provide a duality-based strategy to overcome technical challenges common to continuous-time contracting models with state variables. I also derive a sufficient condition to verify the validity of the first-order approach.

Searching for "Arms": Experimentation with Endogenous Consideration Sets

Daniel Fershtman
,
Tel Aviv University
Alessandro Pavan
,
Northwestern University

Abstract

We study the problem of a decision-maker sequentially choosing between exploring existing alternatives in her consideration set and searching for new ones. We characterize the optimal policy and identify implications for search and exploration dynamics. When the search technology is stationary or improves over time, search is equivalent to replacement. With deteriorating technologies, alternatives are revisited after search is launched and each expansion is treated as if it were the last one. We apply the model to the administration of medical treatments, clinical trials toward regulatory approval, Weitzman (1979)’s Pandora’s boxes problem, and online consumer search.
JEL Classifications
  • D8 - Information, Knowledge, and Uncertainty
  • L1 - Market Structure, Firm Strategy, and Market Performance