Balancing fairness and efficiency in policing: A mean-variance utility and data envelopment analysis approach to arrest practices in U.S. cities

Abstract

This research pioneers a novel integration of the mean-variance utility model with a two-stage Data Envelopment Analysis (DEA) framework to unravel the intricate fairness-efficiency trade-offs in law enforcement arrests. Leveraging data from 157 U S. cities, it explores how police departments navigate the dual imperatives of equity and operational efficiency amidst contextual risks such as crime rates, socioeconomic disparities, and diversity dynamics. The mean-variance utility model provides a rigorous theoretical lens to decode these complexities, while DEA quantifies the relative efficiency of departments in achieving equitable outcomes with constrained resources. The findings reveal an inverted U-shaped relationship between police department diversity and fairness-efficiency, suggesting that diversity enhances outcomes up to an optimal point, beyond which operational challenges begin to erode its benefits. Community diversity positively influences fairness-efficiency, highlighting the critical need for culturally attuned policing strategies. By bridging robust theoretical models with empirical insights, this study offers actionable strategies to optimize diversity, resource allocation, and operational frameworks for equitable policing. Groundbreaking in its approach, this research delivers practical implications for policymakers and law enforcement leaders while advancing the discourse on fairness and efficiency in public sector decision-making.