The Law and Economics of Prosecutorial Discretion

Abstract

My dissertation examines the impact of prosecutorial discretion on punishment outcomes in criminal courts and, specifically, racial disparities in incarceration. Using court records and survey data from state courts, it also provides insights about the mechanisms driving prosecutors' recent impacts on racial disparities. Chapter 1 evaluates how prosecutorial discretion impacts racial disparities and how this impact has evolved in North Carolina state courts between 1995 and 2019. Since prosecutors mediate between the police's initial charge and the final sentence, they can compound, pass through, or attenuate racial disparities from arrests and past cases. Leveraging discontinuities in mandatory prison laws, we find that prosecutors' charging decisions compounded racial disparities in the 1990s but came to attenuate them in recent years. This reversal is concentrated in arrests initiated by police stops, suggesting that prosecutors have increasingly come to check disparities in policing. Chapter 2 uses an original survey of 203 North Carolina prosecutors in 2020 to investigate whether prosecutors have recently checked racial disparities in policing. We find that prosecutors who report actively questioning — rather than passively accepting — police decisions at arrest tend to reduce racial disparities in their cases. Specifically, prosecutors who recall more frequently doubting the police report's accuracy reduce incarceration disparities relative to other prosecutors in their office crime-units. In addition, we find that prosecutors who report that racial disparities are driven more by racial bias than racial disparities in criminal conduct reduce disparities in their cases. Chapter 3 examines the scope of prosecutor discretion by estimating whether different prosecutors systematically produce different effects on incarceration and reoffense. Since defendants cannot reoffend while in prison, the "incapacitation effect" presents an inherent tradeoff between incarceration and reoffense. Yet, prosecutors have the discretion to moderate this tradeoff by selectively incarcerating those most likely to reoffend. Leveraging the quasi-random assignment of cases to prosecutors in North Carolina, we find that prosecutors vary in their effects on incarceration and their effects on reoffense. We further find that prosecutors vary in the degree to which they incarcerate defendants who are more likely to reoffend.