Sentencing Versus Prosecutorial Discretion: the Application of a New Disparity Measure

Abstract

While the term disparity arises naturally and is used frequently in discussions of discretion and fairness in criminal processing, little attention has been given to formulating precise quantitative indicators of it. This article illustrates the application of a recently developed technique for measuring disparity. Prosecutorial and sentencing decisions governing the dispositions of 2,234 convicted felons in a single jurisdiction were used to esti mate the amount of variability in sanction severity attributable to (1) "legit imate" factors of offense seriousness and prior record, (2) sentencing discre tion, and (3) prosecutorial discretion. Approximately one-half of the total variability in sanctions was at tributable to offense seriousness and prior record. Under the assumption that all of the remaining, unexplained variation could be attributed to sentencing and plea bargaining discretion, it was found that sentencing decisions ac counted for a substantially greater proportion of variability than did plea bar gaining decisions. Implications for the development of valid disparity measures are discussed.