Modeling Sentence Length Decisions Under a Guidelines System: An Application of Quantile Regression Models

Chester L. Britt, Arizona State University West

How should sentencing disparity be assessed when decisions are constrained under a sentencing guidelines system? Recent debate over the measurement of sentence disparity under a guidelines system has focused primarily on using interaction terms in regression models to capture the nonadditive effect of offense severity and prior record on sentence length. In this paper, I propose an alternative method to assessing sentencing disparity under a guidelines system that uses quantile regression models. These models offer several advantages over traditional OLS analyses of sentence length, but more importantly, allow for a test of questions such as: Do variables, such as race or offense severity have the same effect on sentence length for the 10% of offenders receiving the most lenient punishments as they do for the 10% of offenders receiving the most severe punishments? I illustrate the application and interpretation of these models with 1994 sentencing data from Pennsylvania. The effects of the independent variables using several different quantiles (10%, 25%, 50%, 75% and 90%) differ in important ways from the effects found when using OLS. I explain how this approach may provide an interesting avenue for further methodological analysis of sentencing decisions made under a guidelines system.

(Return to Program Resources)

Updated 05/20/2006