About

Hi, I'm Nat! I am currently a Ph.D. candidate in the Department of Political Science at Columbia University, focusing on the intersection of institutions and behavior in the American criminal-legal system. My dissertation project examines how elite choices shape carceral outcomes: how police officers factor prosecutorial policy into arrest decisions, whether ever-greater volumes of digital evidence influence the speed and quality of prosecution, and the influence of selection and appointment systems on the sentencing decisions of judges. I am interested in causally-identified observational work that speaks to the everyday exercise of state power.

Before Columbia, I studied Political Science at Reed College and worked on political campaigns across the Pacific Northwest. Outside of working hours, I can often be found at one of New York's many arthouse movie theaters!

Working Papers

★ JOB MARKET PAPER ★
"Why Enforce? Police Discretion Under Non-Prosecution Policies"
Solo-authored
Event study plot: estimated effect of non-prosecution policies on arrests and summons for targeted offenses stays near zero after adoption
Fig. 2A: Enforcement of targeted offenses, by quarter relative to policy adoption.

Misdemeanor enforcement relies heavily on the discretion of individual police officers, but it is unclear what motivates these enforcement decisions. Prosecutor-led reforms of the police rely on legalistic models of policing, which predict that court outcomes help motivate enforcement. However, order-maintenance models of policing argue that arrests are motivated by more immediate concerns. I test these accounts using non-prosecution policies adopted by the District Attorneys of the Bronx (2020) and Manhattan (2022), which publicly committed to declining a range of low-level, discretionary offenses. As New York City's five boroughs share one police department, mayor, and set of laws but have separate prosecutors, I can isolate the effect of prosecutorial policy on enforcement. Using a staggered-adoption synthetic control design with precinct-level arrest and summons data, I rule out declines in enforcement larger than about 18% for offenses targeted by both policies, indicating that most enforcement continued when prosecution ended. The effect varies between precincts, but this variation is not explained by racial composition, commercial activity, gentrification, or prior enforcement intensity. These findings suggest that misdemeanor arrests give police utility beyond the deterrence effect of adjudicated punishment and that reformist prosecutors will have limited leverage over the behavior of patrol officers.

📄 Paper (PDF)📊 Slides
Revise & Resubmit, Quarterly Journal of Political Science
"Do Retention Elections Create Cycles in Criminal Sentencing? An Out-of-Sample Replication"
with Donald P. Green

Judicial retention elections are low-salience, uncompetitive events where incumbents virtually never lose. However, some articles have found that criminal court judges become significantly more punitive as these safe elections approach. Using newly available data spanning multiple election cycles, we conduct an out-of-sample replication of the most widely cited article in this literature (Huber and Gordon 2004) and find no evidence of election cycle effects.

📄 Paper (PDF)

Curriculum Vitae

📄 Full CV (PDF)

Selected highlights below; grants and works in progress are in the full CV.

Education
Columbia University
Ph.D. Political Science (Expected Summer 2027)
M.Phil. Political Science (2026)
M.A. Political Science (2024)
Reed College
B.A. Political Science (2020)
Publications

Christian Breunig and Chris Koski with Nat Rubin. "How Interests and Executives Set Public Policy in Four States." In: Means, Motives and Opportunities: How Executives and Interest Groups Set Public Policy. Cambridge University Press, 2024.

Teaching
  • Spring 2026: "American Urban Politics," TA for Carlos Vargas-Ramos
  • Fall 2025: "Voting and Political Behavior," TA for Yamil Velez
  • Spring 2025: "Judicial Politics," TA for Jeff Lax
  • Fall 2024: "Voting and American Politics," TA for Robert Erikson
  • Spring 2024: "Race and the Carceral State," TA for Andrew McCall
  • Fall 2023: "Data Science for Political Analytics," TA for Greg Wawro
Conferences & Presentations
  • 2026: American Political Science Association Conference (Boston); Midwest Political Science Association Annual Conference (Chicago); Justice and Injustice Conference (Columbia University); Political Science Graduate Student Seminar (Columbia University)
  • 2025: Midwest Political Science Association Annual Conference (Chicago); American Politics Graduate Student Workshop (Columbia University)
  • 2024: American Political Science Association Annual Conference (Philadelphia); Mini-APSA (Columbia University)
  • 2023: American Politics Graduate Student Workshop (Columbia University)