• Working Papers:

1. Endorsements and Referrals: Product Recommendations in Bilateral Trade(with Peter Achim and Lily Yang) 

Abstract: In this paper, we investigate monopoly pricing when the buyer can acquire information about the value of the seller's product from an independent expert. We demonstrate that the expert is advantageous to the buyer only when the cost of seeking advice is not too low. When the buyer faces a moderate cost of obtaining information from the expert, the seller ’prices out’ the expert by offering discounts to dissuade the buyer from seeking advice. This approach generates a positive surplus for the buyer. In contrast, when the buyer’s cost is low, the seller ’prices in’ the expert, capitalizing on the buyer’s easy access to information to demand a premium. We also show that the seller and intermediary may benefit from tacit collusion and direct payments in pricing-out equilibria but never in pricing-in equilibria. 

2. On the England University Dormitory Assignment Problem(with Zaifu Yang)

Abstract: In the dormitory allocation systems used by many universities in England, new students are systematically granted higher priority than senior students. Motivated by this institutional practice, we develop an assignment model that formally captures this priority structure. Our framework stands in sharp contrast to the existing literature, where students endowed with dormitories typically retain priority in the allocation process. Within this setting, we study the joint assignment of new and senior students to dormitories and introduce the Proposing and Swapping Deferred Acceptance (PSDA) mechanism to find a stable matching. We further demonstrate that, under this mechanism, it is a dominant strategy for all students to truthfully report their preferences..

3. Job Matching in Sports (with Zaifu Yang) 

Abstract: In many sports, such as football and wheelchair basketball, roster constraints between teams and players are not fixed but vary flexibly with players’ types. Motivated by this institutional feature, we develop a two-sided matching model with flexible constraints. This approach stands in sharp contrast to the existing literature, where constraints are typically assumed to be exogenously given and fixed. Within this framework, we study the problem of assigning players to clubs when feasibility depends endogenously on the composition of each club’s roster. To address this problem, we propose the Club Proposing Multiple Stages (CPMS) mechanism, which is designed to produce stable and efficient matchings under flexible constraints. We further establish that truthful preference reporting constitutes a dominant strategy for each club under the CPMS mechanism.

4. Matching with Feasibility and Priorities Constraints (Job Market Paper)

Abstract: This paper considers a matching market in which agents belong to institutions that impose feasibility and priority constraints on participation in cross-institutional reallocation. Observing a phenomenon in China, we develop a two-sided matching model between doctors and patients affiliated with hospitals that incorporates constraints. Our objective is to facilitate cross-institutional reallocation between two types of institutions while respecting the constraints they impose. By integrating individual preferences with institutional priorities, we introduce a refined notion of stability. To implement stable and feasible allocations, we propose a Multi-stage Deferred Acceptance (MDA) mechanism and establish its key theoretical properties. Moreover, we show that the MDA mechanism is strategy-proof only for certain groups of patients.

5. Matching over couples (with Zaifu Yang)

Abstract: will be added soon. 

6. Dynamic Stability Across Multiple Platforms (with Rui He)

Abstract: Many matching markets are fragmented, with agents interacting across multiple platforms or locations and adjusting their participation over time. This paper studies a dynamic one-to-one matching model with history-dependent preferences, in which agents’ preferences evolve in response to past matching outcomes and agents may switch platforms to improve their matches. We refine the concept of stability to account for both fragmentation and dynamics, distinguishing among local stability, global stability, and dynamic stability, the latter requiring convergence to a stationary matching that admits no profitable deviations over time. We propose a Men-Preceding Global Mechanism (MPGM), building on deferred acceptance, and show that when each platform implements a stable matching mechanism, MPGM guarantees the existence of a dynamically stable matching. We also establish an impossibility result: without stability at every platform, global and local stability are incompatible.