Research

Working Papers  

Studies of authoritarian governance attribute their longevity and economic underperformance to the loyalty-competence trade-off: rulers sacrifice the competence of their subordinates, hiring incompetent but loyal agents. However, the theory fails to explain how many autocrats maintain longevity while simultaneously achieving economic prosperity. I endogenize loyalty and competence in a two-level principal-agent model between a ruler, a pool of subordinates, and a representative citizen with two core innovations. Policy outcomes can affect regime stability, and there is no institutionalized succession process. The model generates clear predictions about how two key elements - the threat of social protests and uncertainty regarding the policy preference of any successor ruler - shape the loyalty-competence trade-off. I then explore two extensions of the model, revealing counterintuitive findings. First, the shadow of democratization increases the set of kleptocratic regimes characterized by long tenures and poor economic performance. Second, a smooth succession, which is the process by which autocrats can transfer power to their offspring, reduces the set of kleptocratic regimes.



An important public policy question is whether increasing the public sector wage can curb corruption. With the existing theory and empiric overall ambiguous in this regard, I developed a moral hazard and adverse selection model of job application and selection with the assumption that individuals in the job market differ in their Public Service Motivation (PSM). I show that an increase in the public sector wage has two opposing effects on corruption. An incentive effect (higher wages increase the opportunity costs of engaging in corruption for individuals willing to enter the pool of potential recruits before the wage increase) and a selection effect (higher wages attract individuals with lower PSM who are more likely to engage in corruption). An increase in the public sector wage reduces (increases) corruption if and only if the incentive effect dominates (is dominated by) the selection effect. I show that if the semi-elasticity of the public labor supply, which is the percentage of the labor supply as a unit change of the public sector wage, is decreasing (increasing), the incentive effect dominates (is dominated by) the selection effect. In addition, I show that if the semi-elasticity is non-monotone, a trade-off may exist between reducing corruption and increasing the total revenue collected.


Autocrats cannot govern alone. Yet, the officials with whom they rule are also the same ones who hold the seeds of their demise. How do autocrats balance the competing objectives of forming relations of dependence with those officials and nevertheless be able to minimize any threats they can generate? I use a political agency and accountability model to analyze how an autocrat appoints officials to positions of power with the caveat that the official can use his position to mount a coup against him. After performing well, the official can call for the ruler's replacement by claiming to the ruling coalition that he is a better alternative (he is competent). The paper argues that when the threat to the autocrat's power is low, he uses a performance-based rule to appoint officials. This method acts as an imperfect screening device that helps the ruler differentiate between competent and incompetent officials. As the threat of removal increases, the ruler does not appoint on the basis of performance. He increases (respectively decreases) the likelihood of appointment when the official underperforms (respectively performs well). This strategy makes incompetent officials willing to pretend to be competent. Therefore, any good performance and calls for replacement are uninformative since the ruling coalition cannot distinguish between competent and incompetent officials. The mechanism highlighted in this paper challenges the well-known loyalty-competence trade-off by making the distinction between faithfulness and loyalty. I also show that such an instrument from the ruler produces economic inefficiency.




Interethnic political polarization contributes to violence, poor governance, inequality, and stalled development. Padro I Miquel (2007) modeled these dynamics as an interethnic "politics of fear." Using a simplified version of this model to establish a baseline scenario, we study two strategies to shift politics toward less polarizing leader behavior: quotas and mobilization of a "moderate middle" that is not beholden to coethnic appeals. Quotas are modeled as exogenously imposed and fixed (e.g., constitutional) mandates for leaders to allocate resources to out-group members. We characterize conditions necessary to sustain leaders' compliance with quotas. To study the effect of mobilizing a moderate middle, we consider a richer setting in which a segment of each ethnic group is willing to extend support for out-group leaders so as to increase the probability of that leader's survival. We characterize conditions under which mobilization of the moderate middle endogenously sustains leaders' willingness to allocate resources to out-group members, making quotas unnecessary. We use this analysis to interpret the dynamics of ethnic politics and inequality in a set of cases in sub-Saharan Africa.


Works in Progress