Relative Versus Attributable Risk: Implications for Criminal Justice Policies with Disparate Impacts Across Demographic Groups

• In Event: New and Creative Ways to Study Responses to Sanctions and Risk

Abstract

Epidemiology has long drawn a distinction between relative versus attributable risk. In the context of criminal justice policy relative risk is measured by the ratio of the proportion of a specified demographic group affected by that policy to the counterpart proportion for another demographic group. Attributable risk refers to the absolute difference in these two proportions. It thus measures the difference in the proportion of the two groups who bear the burden of or benefit from the policy in question. We argue that the importance of attributable risk is under-appreciated in policy analysis of criminal justice policy. We illustrate our arguments with data on differences across racial groups in US nationwide incarceration rates and differences across ethnic groups in stop, question, and frisk rates in New York City.

Authors

• Charles F. Manski, Northwestern University

• Daniel Nagin, Carnegie Mellon University