Abstract
The core tenet of American civic society is that accountable, democratic institutions promote social welfare, providing fair opportunities for mobility and protecting against arbitrary punishment. Too often, however, our institutions fall short. Academic research has long recognized that socioeconomic inequity frequently stems from the very civic pillars meant to safeguard our social fabric. From an economic perspective, the costs of inequality in our schools and criminal justice system—two key pillars of the state—are abundantly clear. Poor, Black, and Brown citizens, on average, see worse outcomes from their interactions with schools and the courts. Lower-quality schooling leads to worse career outcomes and social disaffection; harsh judicial punishments similarly jeopardize economic security. While these twin problems have long been recognized, relatively little rigorous empirical work has interrogated the structural features of these institutions that generate unequal outcomes. This dissertation considers the causes of inequities in the education and criminal justice systems, and the efficacy of potential solutions.
Chapter 1. No Spending without Representation: School Boards and the Racial Gap in Education Finance
Chapter 1 examines how school boards shape racial and ethnic inequality in public schools. I provide causal evidence that greater minority representation on school boards translates into greater investment in minority students. Focusing on California school boards, I instrument for minority (specifically, Hispanic) representation using random ballot ordering, and leverage new data from a statewide capital investment program to capture intra-district resource allocations. I show that Hispanic board members invest the marginal dollar in high-Hispanic schools within their districts. High-Hispanic schools also exhibit gains in student achievement, attributable to increased investment alongside decreased teacher churn. I conclude that enhancing minority representation on school boards could help combat racial disparities in education.
Chapter 2. Defense Attorneys and Racial Disparities in Plea Bargaining
Chapter 2 considers the interplay between defense attorney incentives and well-known racial gaps in criminal sentencing. Focusing on indigent defendants, I compare plea bargaining and sentencing outcomes for nonwhite clients assigned public defenders to those instead assigned private attorneys. To deliver causal effects, I use an instrumental variables approach that leverages idiosyncratic variation in public defender assignments, driven by caseload capacities. I show that nonwhite defendants represented by private attorneys plead guilty to charges that carry longer sentences relative to comparable nonwhite defendants represented by public defenders. These effects are not explained by differences in defendant or case characteristics, and I find no comparable effects among white defendants. In other words, prosecutors extract harsher plea deals specifically from minority defendants with private attorneys, suggesting opportunistic discrimination against minority defendants with susceptible defense counsel. Consistent with a straightforward signaling model of plea bargaining, I provide empirical evidence suggesting that public defenders who have extensive interactions with prosecutors improve nonwhite defendants' plea bargaining outcomes. Reallocating public defenders to minority defendants could thus diminish racial bias in plea bargaining and better fulfil nonwhite defendants' constitutional right to effective representation.
Chapter 3. Ignorance is Strength? Information and Incentives in Plea Bargaining
Chapter 3 (joint work with Leora Friedberg) complements Chapter 2, interrogating the efficacy of low-cost information interventions that might help vulnerable criminal defendants secure better criminal justice outcomes. This paper evaluates whether informing defendants about the adverse consequences associated with pleading guilty bridges the information gulf between prosecutors and defendants and thereby reduces plea rates. As a plausibly exogenous shock to defendants' information sets, we study the Supreme Court's 2010 decision in Padilla v. Kentucky. Padilla requires defense attorneys to advise their noncitizen clients whether or not they will face deportation by accepting a plea offer. We employ a difference-in-differences design to compare citizens and noncitizens before and after the Court's ruling. Our results show that Padilla’'s information intervention reduced guilty plea rates among noncitizen cases for which a conviction would risk immediate deportation. Conversely, we find that noncitzen cases that do not risk imminent deportation plead guilty more often following Padilla, and were more likely to be incarcerated. We interpret these results as evidence that prosecutors threaten harsher charges for defendants who turn down unfavorable but not catastrophic plea deals. Our findings highlight the limitations of existing and proposed procedural reforms to the plea bargaining process.
