Publications
[1] An Equilibrium Labor Market Model with Internal and External Referrals (with Youze Lang), International Economic Review, 65(2), 2024, 655-692;
Based on the SCE Job Search Survey 2013-2017, about 40% of workers find their jobs through referrals from friends, relatives and acquaintances. We make a distinction between two types of referrals based on whether the referrer works at the hiring firm (internal referrals) or not (external referrals). Interestingly, jobs found through internal referrals pay more, while jobs found through external referrals pay less, than those found through formal methods. An equilibrium labor market model is then built by introducing an incentive-compatible mechanism through which workers can share job opening information with each other, into the classical on-the-job search model. A non-degenerate wage distribution arises in equilibrium with a wage premium for internal referrals, and a wage penalty for external referrals. When calibrated, our model can capture some salient features (particularly pertaining to referrals) of the U.S. labor market.
[2] On the Pure Theory of Wage Dispersion (with Cheng Wang), Review of Economic Dynamics, 47, 2023, 246-277;
We study an equilibrium model of the labor market with identical firms and homogeneous workers, and with search and on-the-job search. Jobs are dynamic contracts that allow firms to match the worker’s outside offers or let the job be terminated. For a non-degenerate distribution of wage offers to arise in the environment, it is necessary and sufficient that (i) there is a positive cost of job turnover, in terminating an existing job or posting a new one; and (ii) there is asymmetric information regarding the worker’s outside offers. The model is calibrated to the U.S. labor market to match observed worker flows and the observed mean-min ratio in wages earned. The calibrated model predicts a unimodal distribution for both wages offered and wages earned.
[3] Optimal CEO Turnover (with Cheng Wang), Journal of Economic Theory, 203, 2022, 105475;
We study a dynamic principal-agent/firm-CEO relationship that is subject simultaneously to moral hazard, limited commitment, and shocks to the CEO’s market value. Termination is used as (a) an incentive instrument to punish the CEO for bad performance, (b) a cost minimization device that uses the CEO’s outside value as an external means for compensation, and (c) as a means for replacing the incumbent CEO with a less expensive new CEO. Termination occurs after the CEO receives either a sufficiently high or a sufficiently low outside value. The model generates both voluntary and involuntary/forced turnovers, and counter-offers occur on the equilibrium path. The model is calibrated to the U.S. data to capture the key observed features of CEO pay and turnover. It shows that increased moral hazard offers an explanation for the observed increase in the level and variance of CEO compensation, as well as the increase in forced CEO turnover.
[4] Optimal Self-Enforcement and Termination (with Cheng Wang), Journal of Economic Dynamics and Control, 101, 2019, 161-186;
We study a principal-agent problem where the agent receives a stochastic outside opportunity (offer) each period, and he cannot commit to the ongoing contractual relationship. Termination, which is costly, allows the principal to go back to an external market to hire a new agent. The principal responds strategically to the agent’s outside offers, choosing optimally between retaining the agent, which must be self-enforced, and terminating him, which then ends the current contract. The model generates both voluntary and involuntary terminations, and dynamics and stationarity outcomes that are of interest especially for understanding employment relationships with on-the-job search.
[5] Equilibrium Matching and Termination (with Cheng Wang), Journal of Monetary Economics, 76, 2015, 208-229;
In an equilibrium model of the labor market with moral hazard, jobs are dynamic contracts, job separations are terminations of optimal dynamic contracts. Transitions from unemployment to new jobs are modeled as a process of random matching and Nash bargaining. Nonemployed workers make consumption and saving decisions as in a standard growth model, as well as whether or not to participate in the labor market. The stationary equilibrium is characterized. The model is then calibrated to the U.S. labor market to study quantitatively the worker flows and distributions, the compensation dynamics, and the effects of UI system.
[6] Outside Opportunities and Termination (with Cheng Wang), Games and Economic Behavior, 91, 2015, 207-228.
We study a dynamic principal–agent relationship where a stochastic outside opportunity (offer) arises each period for the risk averse agent. Termination is costly, but it allows the agent to pursue the available outside opportunity and the principal to return to an external market to hire a new agent. The principal acts strategically with respect to the agent’s outside offers and we show that it is optimal to terminate the current relationship if and only if the agent’s outside offer is above a cutoff level. The optimal contract generates both voluntary and involuntary terminations. Severance compensation arises endogenously and we show that it may be paid only in involuntary terminations. Conditional on termination, the size of the optimal severance compensation depends positively on the agent’s current compensation, but negatively on his outside offer. The optimal contract dictates an inverted-U relationship between compensation and the probability of termination.
Working Papers
[1] The Blockchain Surplus (with Cheng Wang), October 2024, Under Review;
[2] Unemployment Volatility and Wage Dispersion with Optimal Incentive Contracts (with Jie Luo), October 2024, Under Review;
[3] Wage Rigidity in an Incentive Contract with Loss Aversion (with Jie Luo), available soon;
[4] Wage Dispersion and Captial Misallocation (with Weichao Zhu), available soon;
[5] Repeated Moral Hazard with Private Evaluation: Leniency Bias (with Youze Lang);
[6] Repeated Moral Hazard with Private Evaluation: Why the Agent's Mixed Strategies Matter.
Working Progress
[1] Credit Shock Propagation and Housing Price Swing in China: the Role of Refinancing (with Kevin X.D. Huang, Lei Ning and Jingbo Wang);
[2] Networks and Referrals (with Youze Lang).