Exploring the Effects of Offender and County Characteristics Across Conviction Type: A Multi-Level Model

Brenda Sims Blackwell, Georgia State University
Chester L. Britt, Arizona State University West

Much of the research on the severity of sentencing decisions includes dummy variables to statistically control for the type and severity of conviction offense. More recently, some research, in an attempt to test for variation in the effects of such characteristics as the race and sex of the offender, has estimated separate models for drug and non-drug offenders. To build on prior research and more fully explicate whether individual and contextual factors vary in their impact across offenses, we perform two sets of analyses using multi-level modeling strategies. First, we test for variatio in offender and case characteristics across different types of conviction offense (violent, property, drug, other). We then add a third level to the statistical model to determine if differential effects of offender and case characteristics are contingent on the county in which the offender is sentenced. Using data from more than 50,000 cases included in the State Court Processing Statistics database, preliminary results yield differential effects of ofefender and case characteristics on sentence severity across offense type.

(Return to Program Resources)

Updated 05/20/2006