Making Valid Cross-County Comparisons in Criminal Justice Performance

• In Event: The Practice of Justice: New Methods and Meanings to Think About Fairness, Equity and Efficiency in the Courts

Abstract

Criminal justice in the United States is mostly dispensed at the county level. However, given thousands of counties and dozens of performance measures, it is difficult to argue that differences in performance are not due to immutable county characteristics without the ability to make convincing and valid comparisons—to compare like with like. In this study, we use data collected by Measures for Justice from over 300 counties in six states to construct a typology that will enable more valid comparisons between counties across the United States when examining the performance of criminal justice. The typology is informed by structural markers of disadvantage such as poverty levels and racial composition, as well as criminal justice-related factors such as crime rates and system resources. Results first explore the validity of intra- and interstate county comparisons of criminal justice performance using the typology. We further examine the power of the county typology to predict outcomes at the case level when accounting for case and defendant characteristics in a multilevel regression. Analytical and policy implications are discussed

Authors

• Gipsy Escobar, Measures for Justice

• Robert Hutchison, Measures for Justice