Case Study

Arrest Trends, Policy Reform, and Racial Disparities

A comparative analysis of long-term arrest patterns and reform-era shifts across major U.S. cities.

Context: Conducted as an independent policy and data analysis project examining how policing reforms intersect with race, seasonality, and enforcement patterns in New York City, Washington, D.C., and Los Angeles.

Data analysis Policy Racial equity Time-series

Problem

Public debates about policing reforms often rely on isolated incidents or short time windows. This project asked: How do arrest trends actually change over time, and who is most affected? Without long-term, comparative analysis, policymakers risk misattributing causes or overlooking persistent inequities.

Your role

  • Led the project end-to-end: data cleaning, exploratory analysis, modeling, visualization, and policy writing.
  • Designed a reproducible workflow to compare trends across cities and reform periods.

Tools

Python (pandas) Google Colab Time-series analysis Difference-in-differences Data visualization GitHub

Process

  1. Data preparation: standardized offense categories, dates, and city labels; audited missingness and outliers.
  2. Trend analysis: computed monthly totals, rolling averages, and seasonal patterns over time.
  3. Policy mapping: aligned arrest trends to major reform milestones and enforcement changes.
  4. Modeling: applied difference-in-differences–style comparisons to test reform-associated shifts.
  5. Storytelling: translated statistical results into a clear narrative with charts and stated limitations.

Outcomes

  • Analyzed 25 years of monthly arrest data (2000–2024) across 3 major U.S. cities.
  • Processed 100,000+ arrest records to surface long-term and seasonal enforcement patterns.
  • Identified recurring summer arrest spikes and reform-era inflection points with unequal racial impacts.
  • Produced a reproducible analysis notebook and policy paper suitable for academic and practitioner audiences.

What I’d do next

  • Extend the analysis to post-2024 data and additional metropolitan areas.
  • Incorporate neighborhood-level indicators to better capture localized enforcement disparities.
  • Translate findings into an interactive dashboard for policymakers and journalists.