Paulo Matos

How does the implementation of police discipline shape deterrence? I study this question in New York City, where an independent civilian board recommends disciplinary penalties but the NYPD Commissioner retains full authority over final decisions. This institutional structure allows officers to observe both the recommendation and the final disciplinary outcome. I exploit the introduction of the NYC Disciplinary Matrix, which reduced the civilian board's discretion by introducing a clearer mapping between misconduct and recommended penalties, while leaving the Commissioner's discretion unchanged. I find that disciplinary recommendations reduce new allegation rates by 23\% overall, with stronger effects after the reform, reaching 37\%. Conditional on a recommendation, final discipline reduces new allegation rates by 67\%, but this additional effect is concentrated in the pre-reform period and disappears after the matrix. The evidence suggests that recommendations place officers on alert in anticipation of the Commissioner's decision in both policy regimes, while the matrix strengthens deterrence by making lower-level prohibited behaviors and their consequences clearer to officers. At the same time, final decisions become less deterrent after the reform because penalties become more lenient. These findings highlight how institutional design shapes the deterrent effects of police accountability systems.

I develop a game-theoretic model in which a civilian oversight agency and a police chief jointly determine whether to discipline an officer found guilty of misconduct. Decisions depend on the severity of the allegation and the compliance cost of discipline. The agency observes the severity of the allegation but not the compliance cost, while the chief privately observes the compliance cost and may incur a review cost to acquire case-specific information. I show that an increase in the review cost generates two opposing forces shaping the agency’s recommendation behavior. An overruling effect encourages recommendations for discipline, as a higher review cost makes it less likely that the chief reviews and overturns the agency’s recommendation. In contrast, a double-checking effect discourages recommendations, since a higher review cost limits the chief’s ability to review and confirm cases, depriving the agency of valuable information. The interaction of these forces yields a non-monotonic relationship between the review cost and the probability that the agency recommends discipline in equilibrium. Consequently, the institutional framework that maximizes accountability depends on the review cost: a traditional system without civilian oversight is optimal when the review cost is low; an oversight agency that accounts for mitigating factors performs best when the review cost is moderate; and an agency that ignores such factors becomes optimal when the review cost is high.

We exploit an exogenous government policy in El Salvador that reallocated several high-ranking gang leaders from maximum-security prisons to lower-security facilities between 2012 and 2015, combined with highly detailed administrative incarceration data spanning more than a decade. Our key finding is that inmates exposed to these gang leaders within the first days of a new cell assignment exhibit higher probabilities of future gang-related recidivism, but not of non-gang-related crimes. We identify the transmission of criminal capital from gang leaders to other inmates as the main mechanism driving the increase in recidivism. Moreover, we find that the exposure effect is amplified among inmates with prior gang-related offenses—such as homicide, gang affiliation, or extortion—but is not affected by the pre-existing gang composition of the cell.

How does the strategic interaction of inmates within cells influence their future recidivism, and what are the implications for optimal incarceration policies? I address this question by developing a game-theoretic model in which inmates assigned to the same cell first decide whether to cooperate and, subsequently, whether to reoffend after release. The benefits and costs of cooperation are endogenous: cooperation allows inmates to expand their criminal networks through knowledge transfer but also exposes their future criminal enterprises to their partners through information leakage. As a result, recidivism decisions become strategic substitutes between cellmates in equilibrium. The model yields three main findings. First, policies aimed at reducing recidivism—such as improving inmates’ economic conditions or increasing the cost of reoffending—are more effective when cooperation occurs. Second, inmates do not always find it optimal to cooperate; when they do, enhancing legitimate economic opportunities can reduce incentives for criminal knowledge transmission. Finally, I characterize the conditions under which it is optimal, in terms of minimizing expected recidivism rates, to segregate or mix high-skilled and low-skilled criminals.