Job Market Paper
Publications
Positive and Negative Sorting in Team Contests, with Qiang Fu, Zenan Wu, and Hanyao Zhang [from my undergraduate work]
Journal of Industrial Economics, January 2024. [SSRN]
Working Papers
Revealed Information, with Laura Doval, Ran Eilat, and Tianhao Liu, July 2025. [slides]
Collective Sampling: An Ex Ante Perspective, December 2024. [arxiv]
an earlier version that contains an application on dynamic persuasion; presented at 2025 ESWC
Pitfall of Precision in Noisy Signaling, with Shuhua Si, September 2025.
Peer Information in Mechanism Design without Transfers, with Yutong Zhang. [a preliminary draft]
Work in Progress
Dynamic Contract Design under Falsification, with Tianhao Liu. [slides]
Short summary: We study a dynamic moral hazard problem where the agent can costly manipulate the monitoring outcomes. Surprisingly, manipulation can be beneficial for the principal: by inducing the agent to both work and manipulate, the principal can prolong the relationship and improve efficiency relative to the benchmark without manipulation.
Credible Contracts and Performance Evaluations (draft available upon request)
presented at 2024 NAWMES
Abstract: This paper studies the credible design of contracts with limited commitment. The principal jointly designs the evaluation system and the payment scheme, subjective to evaluation costs. With limited commitment, the principal can deviate from her promised evaluation design and misreport the evaluation outcome, as long as the agent is not able to detect her misbehavior using his private data. I introduce a novel credibility requirement for contracts, extending the notion in Akbarpour and Li (2020): a contract is credible if the principal has no incentives to deviate in any undetectable way. In credible contracts, the principal gives lower payments when she knows more about the agent's private data and she designs the evaluation system to optimally learn what data the agent has. From the ex ante perspective, the principal always benefits from the agent having more private data. In optimal credible contracts, the principal looks for hard evidence on shirking-indicative data, and credibly punishes the agent upon finding of such evidence. When shirking-indicative data is too rare on the path, the principal prefers to use perfectly informative evaluations to establish credibility.