Fairness Is More Than Algorithms: Racial Disparities in Time-to-Recidivism

TL;DR

The paper addresses how racial disparities in recidivism arise from both algorithmic risk predictions and broader socioeconomic contexts over time. It introduces a multi-stage causal framework and uses survival analysis to test counterfactual racial parity, supported by a formal theorem and lemma. Empirically, COMPAS data show no short-term racial differences within the same risk group but reveal long-term disparities in low-risk individuals, highlighting the role of cumulative structural factors. The work emphasizes policy implications beyond algorithmic fixes and provides a general framework applicable to other domains where algorithmic decisions interact with social inequality.

Abstract

Racial disparities in recidivism remain a persistent challenge within the criminal justice system, increasingly exacerbated by the adoption of algorithmic risk assessment tools. Past works have primarily focused on bias induced by these tools, treating recidivism as a binary outcome. Limited attention has been given to non-algorithmic factors (including socioeconomic ones) in driving racial disparities from a systemic perspective. To that end, this work presents a multi-stage causal framework to investigate the advent and extent of disparities by considering time-to-recidivism rather than a simple binary outcome. The framework captures interactions among races, the algorithm, and contextual factors. This work introduces the notion of counterfactual racial disparity and offers a formal test using survival analysis that can be conducted with observational data to assess if differences in recidivism arise from algorithmic bias, contextual factors, or their interplay. In particular, it is formally established that if sufficient statistical evidence for differences across racial groups is observed, it would support rejecting the null hypothesis that non-algorithmic factors (including socioeconomic ones) do not affect recidivism. An empirical study applying this framework to the COMPAS dataset reveals that short-term recidivism patterns do not exhibit racial disparities when controlling for risk scores. However, statistically significant disparities emerge with longer follow-up periods, particularly for low-risk groups. This suggests that factors beyond algorithmic scores, possibly structural disparities in housing, employment, and social support, may accumulate and exacerbate recidivism risks over time. This underscores the need for policy interventions extending beyond algorithmic improvements to address broader influences on recidivism trajectories.

Figures (5)

• Figure 1: A causal DAG corresponding to the multi-stage recidivism process.
• Figure 2: Survival analysis of recidivism patterns across racial groups and COMPAS recidivism risk groups. The subplots display survival curves and statistical significance analysis: (a) survival curves for Caucasian defendants, (b) survival curves for African-American defendants, and (c) corresponding p-values from log-rank tests over time. Gray ($p > 0.1$) indicates insufficient evidence of racial differences, light pink ($0.05 < p \leq 0.1$) indicates marginal differences, and red ($p \leq 0.05$) indicates significant differences.
• Figure 3: Survival analysis of recidivism patterns across racial groups and COMPAS violent recidivism risk groups. The subplots display survival curves and statistical significance analysis: (a) survival curves for Caucasian defendants, (b) survival curves for African-American defendants, and (c) corresponding p-values from log-rank tests over time. Gray ($p > 0.1$) indicates insufficient evidence of racial differences, light pink ($0.05 < p \leq 0.1$) indicates marginal differences, and red ($p \leq 0.05$) indicates significant differences.
• Figure 4: Survival analysis of recidivism patterns across racial scores and COMPAS recidivism risk groups. The subplots display survival curves and statistical significance analysis: (a) survival curves for Caucasian defendants, (b) survival curves for African-American defendants, and (c) corresponding p-values from log-rank tests over time. Gray ($p > 0.1$) indicates insufficient evidence of racial differences, light pink ($0.05 < p \leq 0.1$) indicates marginal differences, and red ($p \leq 0.05$) indicates significant differences.
• Figure 5: Survival analysis of recidivism patterns across racial scores and COMPAS violent recidivism risk scoress. The subplots display survival curves and statistical significance analysis: (a) survival curves for Caucasian defendants, (b) survival curves for African-American defendants, and (c) corresponding p-values from log-rank tests over time. Gray ($p > 0.1$) indicates insufficient evidence of racial differences, light pink ($0.05 < p \leq 0.1$) indicates marginal differences, and red ($p \leq 0.05$) indicates significant differences.

Theorems & Definitions (4)

• Definition 1: Counterfactual Racial Parity
• Theorem 1
• Corollary 1
• Lemma 1