JSP PH.D. CANDIDATE
Punishment, politics, and democracy
I'm a Jurisprudence and Social Policy Ph.D. candidate at UC Berkeley. My academic research delves into the relationship between citizens' perceptions of justice in a democracy and the responsiveness of our representative system.
I use public opinion, administrative data, and election outcomes to connect the dots between voters' attitudes and accountability in the justice system.
My research demonstrates the effect of local elections and voters' preferences on the behavior of police officers and prosecutors, and changes in local imprisonment rates. Additionally, my research explains voting behavior in the politics of crime and justice.
LL.B, LL.M, Ph.D. (expected, 2024)
Curiosity about justice as a vocation brought me to law school, where I focused on studying jurisprudence and socio-legal theory. After a short learning period with inspiring lawyers of the Israeli National Public Defense department, I came to Berkeley to focus on the structural and psychological underpinnings of the criminal legal system.
Current Works in Progress:
Perceived Algorithmic Legitimacy in the Administrative State
Algorithmic decision-making tools are changing the operation of the Administrative State, as machine learning models and other forms of automation proliferate in the government. Administrative legitimacy, an already contentious matter, is being further questioned with concerns regarding algorithmic decision-making. Can agencies maintain or even reinforce legitimacy while relying on algorithms? A burgeoning literature puts forth an array of proposals to improve administrative accountability and legitimacy of algorithmic decision-making tools, yet the efficacy of improving public perceptions of legitimacy has not been thoroughly empirically tested. This article uses a conjoint survey experiment (N=499) to test which aspects of a governmental program that employs an algorithmic tool affect the public’s perceptions of legitimacy. Moving beyond direct human-algorithm comparisons toward more complex attributes, this study applies and contrasts two theoretical frameworks: Procedural Justice and Algorithmic Governance. Also, the study unpacks notions of fairness and efficiency to construct a more nuanced model of algorithmic legitimacy. We find that a positive effect on perceived legitimacy derives from communicating the algorithmic procedure: notice about the use of an algorithm, human involvement in algorithmic decision-making, decision explanation, and hearing by request. We also find that respondents consider the policy domain and deem the use of algorithms in the criminal justice context as less fair than other domains but not necessarily less efficient. Finally, we find that perceptions of fairness differ from perceptions of efficiency.
Intensive and Extensive Margins of Legal Reform
Decades of support for tough-on-crime policies have led the United States to lock up more people per capita than most countries worldwide. In recent years, however, voters have become less supportive of such policies. Under which conditions do voters support progressive reform? Are voters willing to support politicians who are not tough on crime? I develop a theory of prosecutor elections voting behavior using data from a recent recall election and test it on a national sample. Recently, San Francisco voters, who are seen as among the more progressive voters in the country, recalled a leading progressive prosecutor. The recall election is a rare opportunity to examine voters' revealed preferences in a setting that is becoming pivotal for the politics of crime and justice. I argue that there are political majorities in favor of reforming the intensity of the criminal legal system, not its extent; voters support reducing outcomes' harshness but not limiting the scope of prosecuted behavior. I show that support for reducing the intensity of the criminal legal system is wide and cuts across voters with different political attitudes. On the other hand, support for reducing the extensiveness of the criminal legal system is narrow. I conclude that to gain political approval, politicians who intend to end Mass Incarceration should focus on reducing the criminal legal system's intensive margin, not its extensive margin.
Why Support Crime Control Policy Reform? Group Cues and Informational Persuasion
What are the effects of racial group cues on crime control policy preferences? I develop and test a theory of how people support different policy responses to crime. Existing theories often highlight the impact of identity on support for capital punishment, minimum sentencing law, or other such policies, yet the effects of attitudes toward race and party cues are overlooked. In this paper, I show that people’s dispositions toward racial and political groups affect criminal justice policy preferences. In a conjoint design and a follow-up survey experiment, I show that the effect of racial group cues depends on the person’s reported race. People of color positively follow cues from their in-group, but respondents who reported a white identity follow out-group racial cues and distrust their in-group. Moreover, cues from people’s political out-group motivate policy opposition.
The Effect of Causal Attribution on Dangerousness and Race Perceptions
What affects perceptions of dangerousness? Assessing whether an individual poses a risk to society is a routine activity throughout the criminal legal system. Decision makers may consider professional risk assessments, but the decision is subjective and might rely on heuristics. Psychology research finds that the Fundamental Attribution Error affects perceptions of morality through the belief that the actions of others are the direct outcome of personality or character. However, less is known about its effect on perceptions of dangerousness. Across two online experiments (N = 1005 and N = 276), I examine how a putative parole candidate’s choice to attribute their crime to dispositional or situational factors affects people’s perceptions of dangerousness and racial classification. The studies show that the use of situational attribution increased perceptions of dangerousness when the participant did not accept the situational account of the candidate as valid. Moreover, attributing the causes of crime to the situation and the environment predicts perceiving the parole candidate as a person of color. Understanding what affects perceptions of dangerousness is crucial for the criminal legal system, and policy suggestions are discussed.
The Effect of DA Elections on Public Safety
What is the effect of a political struggle between a district attorney (DA) and a police department on public safety? During the tenure of Chesa Boudin as San Francisco’s DA, a political rivalry developed between the district’s prosecutors and police department. Typically partners, the DA’s office under Boudin's leadership vowed to tighten up police accountability in response to misconduct allegations. In 2022 a political campaign to recall Boudin succeeded, resulting in Boudin’s electoral defeat (June) and leave of office (July). In this paper, we estimate the effect of the police department taking sides in a political campaign. We show that there was a sharp change in the jail population after the recall and argue that it is attributed to changes in police behavior, not DA preferences: an increase in crime reports, in police stops for traffic and public order offenses, and in arrests. This paper suggests that police departments react to political pressure and might influence local elections.