Challenging Aggregation: Methodological Flaws in the 'Person of Color' Variable for Criminal Justice Analysis

• In Event: Lightning Talk: Advancing Criminological Research Through Innovative Methodologies

Abstract

The methodological practice of aggregating racial and ethnic minorities under a unified "Person of Color" (POC) classification introduces substantial measurement error that significantly impacts the assessment of criminal justice outcomes. Using incident-level data from arrest and criminal case records, we calculated disparity ratios for jail and prison incarceration and violent crime arrests, comparing aggregated and disaggregated racial/ethnic categories. Analysis of 950,810 criminal cases (2014-2022) in a large urban county in the southwestern United States and 765,310 arrest records from a large metropolitan Police Department (2014-2020) demonstrates that POC aggregation systematically obscures population-specific disparities. Disaggregated data show that Black individuals are 2.7 times more likely to be incarcerated than White individuals, while Hispanic individuals are 1.2 times more likely—a substantial difference masked by aggregate measures. Similarly, in violent crime arrests, aggregating Black and Hispanic individuals into a 'POC' category concealed the extent of Black-White disparities and inflated Hispanic-White disparities. Aggregating Black and Hispanic populations under the 'People of Color' metric leads to a systematic underestimation of Black disparities and overestimation of Hispanic disparities, revealing a significant methodological limitation. These findings suggest that POC aggregation masks group-specific patterns in criminal justice outcomes, supporting the need for disaggregated analytical approaches.

Authors

• Jennifer Bourgeois, Texas Southern University

• Howard Henderson, Texas Southern University