Publications:
Dynamic Evaluation Design, American Economic Journal: Microeconomics, Forthcoming
A principal owns a firm, hires an agent of uncertain productivity, and designs a dynamic policy for evaluating his performance. The agent observes ongoing evaluations and decides when to quit. When not quitting, the agent is paid a wage that is linear in his expected productivity; the principal claims the residual performance. After quitting, the players secure fixed outside options. I show that equilibrium is Pareto efficient. For a broad class of performance technologies, the equilibrium wage deterministically grows with tenure. My analysis suggests that endogenous performance evaluation plays an important role in shaping careers in organizations.
Cream Skimming and Information Design in Matching Markets (with Gleb Romanyuk),
American Economic Journal: Microeconomics, 2019, 11 (2): 250-76
Short-lived buyers arrive to a platform over time and randomly match with sellers. The sellers stay at the platform and sequentially decide whether to accept incoming requests. The platform designs what buyer information the sellers observe before deciding to form a match. We show full information disclosure leads to a market failure because of excessive rejections by the sellers. If sellers are homogeneous, then coarse information policies are able to restore efficiency. If sellers are heterogeneous, then simple censorship policies are often constrained efficient as shown by a novel method of calculus of variations.
Games and Economic Behavior, 2018, 110: 330-339
We study dynamic games in which senders with state-independent payoffs communicate to a single receiver. Senders’ private information evolves according to an aperiodic and irreducible Markov chain. We prove an analog of a folk theorem — that any feasible and individually rational payoff can be approximated in a perfect Bayesian equilibrium if players are sufficiently patient. In particular, there are equilibria in which the receiver makes perfectly informed decisions in almost every period, even if no informative communication can be sustained in the stage game. We conclude that repeated interaction can overcome strategic limits of communication.
American Economic Review, 2018, 108(1): 1-48
A data buyer faces a decision problem under uncertainty. He can augment his initial private information with supplemental data from a data seller. His willingness to pay for supplemental data is determined by the quality of his initial private information. The data seller optimally offers a menu of statistical experiments. We establish the properties that any revenue-maximizing menu of experiments must satisfy. Every experiment is a non-dispersed stochastic matrix, and every menu contains a fully informative experiment. In the cases of binary states and actions, or binary types, we provide an explicit construction of the optimal menu of experiments.
Working papers:
Optimal Technology Design (with Daniel Garrett, George Georgiadis, and Balázs Szentes), October 2020, Online Appendix
This paper considers a moral hazard model with (i) a risk-neutral agent and (ii) agent limited liability. Prior to interacting with the principal, the agent designs the production technology, which is a specification of the agent's cost of generating each output distribution with support contained in [0,1] . After observing the production technology, the principal offers a payment scheme and then the agent chooses a distribution over outputs. First, we show that there is an optimal design involving only binary distributions on {0,1}; that is, the cost of any other distribution is prohibitively high. Then, we characterize the equilibrium technology defined on the binary distributions and show that the equilibrium payoff of both the principal and the agent is 1/e . A notable feature of the equilibrium is that the principal is indifferent between offering the equilibrium bonus rewarding output one and anything less than that. Finally, the analysis of the model is shown to generalize to the case where the agent is risk averse.
Disclosure and Pricing of Attributes, July 2020
A monopolist sells an object characterized by multiple attributes. A buyer can be one of many types, differing in their willingness to pay for each attribute. The seller can provide arbitrary attribute information in the form of a statistical experiment. To screen different types, the seller offers a menu of options that specify information prices, experiments, and object prices. I characterize revenue-maximizing menus. All experiments belong to a class of linear disclosure policies. An optimal menu may be nondiscriminatory and qualitatively depends on the structure of buyer heterogeneity. The analysis informs on the benefits of partial disclosure in pricing settings.