Contextual Disparities in Sentencing: A Hierarchical Analysis of Variations in Sentencing Severity Across Pennsylvania Courts

Brian Daniel Johnson, Pennsylvania State University
Jeffery T. Ulmer, Pennsylvania State University

Using recent data from the Pennsylvania Commission on Sentencing (PCS), this study investigates the impact of various contextual factors on sentencing outcomes across county-level courts in Pennsylvania. It examines the separate and joint effects of three subsets of theoretically relevant factors -- courtroom contexts, county crime characteristics, and county population demographics -- on both the likelihood and length of incarceration. We draw from various theoretical perspectives to introduce hypotheses about the relative influence of individual and contextual factors (as well as specific cross-level interactions) on sentencing outcomes. To examine these predictions, we employ Hierarchical Linear Modeling (HLM) techniques (Bryk & Raudenbush, 1992), designed to account for the nested nature of contextual sentencing data. Preliminary results indicate 1) significant variation in sentencing severity across counties 2) significant variation in the effects of different sentencing factors across counties, and 3) significant influences from contextual factors on mean sentencing severity across counties. These findings suggest that sentencing severity varies considerably across county-level courts, although only a small portion of this between-county variation is accounted for by our contextual variables.

(Return to Program Resources)

Updated 05/20/2006