A Comparative Simulation Study of the Fairness and Accuracy of Predictive Policing Systems in Baltimore City

2026-01-30

AI Summary

This study investigates whether predictive policing in Baltimore exacerbates racial inequities due to historical biases and feedback loops, compared to traditional hotspot policing. The authors develop the first reproducible, city-scale agent-based simulation framework that integrates historical crime data–driven feedback mechanisms with multidimensional fairness metrics to systematically evaluate both strategies in terms of short-term effectiveness and long-term bias evolution. Findings reveal that while predictive policing demonstrates higher short-term accuracy and appears more equitable initially, it amplifies bias at a faster rate over time. Hotspot policing also exhibits significant bias, with certain scenarios in Baltimore showing disproportionate enforcement in White neighborhoods—challenging the prevailing assumption that algorithmic bias exclusively disadvantages minority groups.

Technology Category

• Philosophy and Ethics of AI: Bias, Fairness & Equity
• Machine Learning: Ethics, Bias, and Fairness
• Computer Vision: Bias, Fairness & Privacy

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Social Networks and Social Media: Fairness and bias in social network and social media analysis
• Responsible Web: Algorithmic accountability and transparency on the web

Abstract

There are ongoing discussions about predictive policing systems, such as those deployed in Los Angeles, California and Baltimore, Maryland, being unfair, for example, by exhibiting racial bias. Studies found that unfairness may be due to feedback loops and being trained on historically biased recorded data. However, comparative studies on predictive policing systems are few and are not sufficiently comprehensive. In this work, we perform a comprehensive comparative simulation study on the fairness and accuracy of predictive policing technologies in Baltimore. Our results suggest that the situation around bias in predictive policing is more complex than was previously assumed. While predictive policing exhibited bias due to feedback loops as was previously reported, we found that the traditional alternative, hot spots policing, had similar issues. Predictive policing was found to be more fair and accurate than hot spots policing in the short term, although it amplified bias faster, suggesting the potential for worse long-run behavior. In Baltimore, in some cases the bias in these systems tended toward over-policing in White neighborhoods, unlike in previous studies. Overall, this work demonstrates a methodology for city-specific evaluation and behavioral-tendency comparison of predictive policing systems, showing how such simulations can reveal inequities and long-term tendencies.

Problem

• predictive policing
• fairness
• bias
• accuracy
• comparative study

Innovation

• predictive policing
• fairness simulation
• feedback loops
• hot spots policing
• algorithmic bias