Research Statement [PDF]
Working papers
1. Elite Persistence, Power Struggles and Coalition Dynamics [PDF]
Ronald Coase Workshop of Institutional Analysis 2017
Why do social and political hierarchies often prove extremely difficult to eradicate? This paper studies the persistence of elites in coalition dynamics where players use their power to eliminate others and split resources. Our innovation is to allow players to give away power when they are about to be expelled, thus endogenously creating inequality in equilibrium power structure. We characterize the stable structures in general and study the robustness of equilibrium. Interestingly, perturbations on equilibrium structures exhibit the Matthew effect: power often ends up more concentrated to a few elite members regardless of initial perturbation making the elite stronger or weaker. We use our model to explain why revolutions that aim to install social equality quickly reproduce the same type of hierarchies that the revolutionaries sought to destroy. We also discuss how elite persistence is influenced by economic development, external threats and international cooperation.
2. The Economic Motives of Foot-binding (with Lingwei Wu) [PDF]
The Northeast Universities Development Consortium Conference 2016, Economic History Association Annual Meeting 2016, The 4th International Symposium on Quantitative History 2016
How can economic incentives affect cultural practices? As a painful custom persisted in historical China, foot-binding targeted girls whose feet were systematically reshaped during early childhood. This paper models foot-binding as a pre-marital investment in the marriage market, investigating its emergence, diffusion, and decadence. Foot-binding emerged with a gender-biased shock in the elite recruitment system that made men's quality distribution more heterogeneous than women's. Foot-binding was then taken up as a tool to compete for better grooms and thrived as the elite recruitment system became a major channel for upward mobility. Foot-binding adoption in working class women also exhibited distinctive regional variation since it impeded heavy labor: it was highly prevalent in regions where women specialized in sedentary labor, and was less popular in regions where women specialized in farmland work. Both qualitative evidence and empirical analysis with the Republican China archives are consistent with our theoretical predictions.
3. Bureaucratic Control, Information Management, and Governmental Responsiveness (with Feng Yang) [PDF]
Midwest Political Science Association Annual Conference 2015, The 5th International Workshop on Economic Analysis of Institutions
When a superior officer and a subordinate jointly determine a noisy performance, and the superior cares about his reputation among the audience outside the organization, promotion becomes a signaling tool for the superior. The superior can strategically postpone promoting a good subordinate to shift blame and enhance his own reputation. Moreover, the superior has more incentives to shirk when having the tools of strategic promotion to manipulate reputation, which decreases governmental responsiveness. We provide supportive evidence using data on provincial official turnovers in China.
Research in progress
4. Monitoring Dynamics in Fighting Corruption
When an anti-corruption investigator can take costly efforts to investigate both current and past actions of a potentially corrupt agent, he "signs a deal with the devil": he may turn a blind eye to corruption behavior in early periods, allowing the agent to build up punishable wealth, then use the threat of confiscation to enforce future good actions.
5. Investment with News Arrival
When multiple firms compete to launch a new product whose market potential is unknown, firms may wait and acquire costly news to learn. We show that when the market only accommodates one product, the best firm always launches new product first. We also solve for the optimal launching time. When there are room for multiple products, the best firm may adopt a "wait and see" strategy: allowing others to test the product in the market, and entering later when the market potential is proven.
6. The Promotion Club (with Shuo Chen and Zhitao Zhu)
We consider a tournament-like incentive scheme called the "promotion club": Principal selects, from n agents, m top-performing ones into a club, and then promote one from within a club based fully on principal's idiosyncratic preference ("mindset"). Intuitively, m=1 indicates a tournament, and m=n, cronyism. We show that a proper (m>1) promotion club often helps alleviate the loyalty-competency trade-off, for two reasons: 1. effort features an inverted-U shape with club size; 2. larger club size makes it easier to recruit a potential crony. However, principal always tends to oversize the club, leading to effort distortions.
We further show that the existence of some publicly known crony/well-connected agents may lead to two equilibria: a shirk equilibrium when club size is large and all agents shirk; and a work equilibrium where club size is small and well-connected agents may work even harder than in a tournament. We use promotion data of Chinese governments at different levels to test our model predictions. Our theory and empirical evidence helps to explain how a regime with centralized personnel control (e.g. China) manages to incentivize subordinates while maintaining high degrees of discretionary power at the top.
Working papers
1. Elite Persistence, Power Struggles and Coalition Dynamics [PDF]
Ronald Coase Workshop of Institutional Analysis 2017
Why do social and political hierarchies often prove extremely difficult to eradicate? This paper studies the persistence of elites in coalition dynamics where players use their power to eliminate others and split resources. Our innovation is to allow players to give away power when they are about to be expelled, thus endogenously creating inequality in equilibrium power structure. We characterize the stable structures in general and study the robustness of equilibrium. Interestingly, perturbations on equilibrium structures exhibit the Matthew effect: power often ends up more concentrated to a few elite members regardless of initial perturbation making the elite stronger or weaker. We use our model to explain why revolutions that aim to install social equality quickly reproduce the same type of hierarchies that the revolutionaries sought to destroy. We also discuss how elite persistence is influenced by economic development, external threats and international cooperation.
2. The Economic Motives of Foot-binding (with Lingwei Wu) [PDF]
The Northeast Universities Development Consortium Conference 2016, Economic History Association Annual Meeting 2016, The 4th International Symposium on Quantitative History 2016
How can economic incentives affect cultural practices? As a painful custom persisted in historical China, foot-binding targeted girls whose feet were systematically reshaped during early childhood. This paper models foot-binding as a pre-marital investment in the marriage market, investigating its emergence, diffusion, and decadence. Foot-binding emerged with a gender-biased shock in the elite recruitment system that made men's quality distribution more heterogeneous than women's. Foot-binding was then taken up as a tool to compete for better grooms and thrived as the elite recruitment system became a major channel for upward mobility. Foot-binding adoption in working class women also exhibited distinctive regional variation since it impeded heavy labor: it was highly prevalent in regions where women specialized in sedentary labor, and was less popular in regions where women specialized in farmland work. Both qualitative evidence and empirical analysis with the Republican China archives are consistent with our theoretical predictions.
3. Bureaucratic Control, Information Management, and Governmental Responsiveness (with Feng Yang) [PDF]
Midwest Political Science Association Annual Conference 2015, The 5th International Workshop on Economic Analysis of Institutions
When a superior officer and a subordinate jointly determine a noisy performance, and the superior cares about his reputation among the audience outside the organization, promotion becomes a signaling tool for the superior. The superior can strategically postpone promoting a good subordinate to shift blame and enhance his own reputation. Moreover, the superior has more incentives to shirk when having the tools of strategic promotion to manipulate reputation, which decreases governmental responsiveness. We provide supportive evidence using data on provincial official turnovers in China.
Research in progress
4. Monitoring Dynamics in Fighting Corruption
When an anti-corruption investigator can take costly efforts to investigate both current and past actions of a potentially corrupt agent, he "signs a deal with the devil": he may turn a blind eye to corruption behavior in early periods, allowing the agent to build up punishable wealth, then use the threat of confiscation to enforce future good actions.
5. Investment with News Arrival
When multiple firms compete to launch a new product whose market potential is unknown, firms may wait and acquire costly news to learn. We show that when the market only accommodates one product, the best firm always launches new product first. We also solve for the optimal launching time. When there are room for multiple products, the best firm may adopt a "wait and see" strategy: allowing others to test the product in the market, and entering later when the market potential is proven.
6. The Promotion Club (with Shuo Chen and Zhitao Zhu)
We consider a tournament-like incentive scheme called the "promotion club": Principal selects, from n agents, m top-performing ones into a club, and then promote one from within a club based fully on principal's idiosyncratic preference ("mindset"). Intuitively, m=1 indicates a tournament, and m=n, cronyism. We show that a proper (m>1) promotion club often helps alleviate the loyalty-competency trade-off, for two reasons: 1. effort features an inverted-U shape with club size; 2. larger club size makes it easier to recruit a potential crony. However, principal always tends to oversize the club, leading to effort distortions.
We further show that the existence of some publicly known crony/well-connected agents may lead to two equilibria: a shirk equilibrium when club size is large and all agents shirk; and a work equilibrium where club size is small and well-connected agents may work even harder than in a tournament. We use promotion data of Chinese governments at different levels to test our model predictions. Our theory and empirical evidence helps to explain how a regime with centralized personnel control (e.g. China) manages to incentivize subordinates while maintaining high degrees of discretionary power at the top.