RACE AND SENTENCING: MAKING SENSE OF THE INCONSISTENCIES

Abstract

Numerous sentencing studies have addressed the question: Are racial/ethnic minorities treated more harshly in comparison to similarly situated whites? Several authors have attempted to review this voluminous and diverse body of research using traditional qualitative narrative literature review techniques. It is my contention that these narrative reviews are of limited utility because of shortcomings inherent in qualitative literature reviews. Furthermore, typically these reviews do not focus primarily on explaining why the results of studies addressing this question diverge. To remedy these deficiencies in the literature, I utilize quantitative synthesis methods ("meta-analysis") that quantify the difference in sentencing outcomes between whites and minorities after controlling for legally relevant factors (i.e., criminal history and offense seriousness) in adult, non-capital cases processed in the United States. The meta-analytic procedure employed in this research more accurately determines whether unwarranted racial disparity exists in sentencing outcomes and estimates the magnitude of such disparity. Additionally, using coded information from each study and characteristics of the sentencing jurisdiction obtained from the U.S. Census, I attempt to explain why the findings from this body of research diverge. The results of this meta-analytic synthesis indicate that minorities were sentenced more harshly than whites. Differences in sentencing outcomes between these groups generally were statistically significant but statistically small (although not necessarily substantively small). Larger estimates of unwarranted sentencing disparity were found in analyses that examined drug offenses; larger estimates of unwarranted disparity were also found when researchers assessed sentencing outcomes relating to imprisonment decisions or discretionary decisions. Smaller estimates of unwarranted sentencing disparity were found in analyses that employed more control variables, especially those that controlled for defendant SES, utilized precise measures of key variables, or examined sentencing outcomes relating to length of incarcerative sentence. However, even when consideration was confined to those analyses employing key controls and precise measures of key variables, statistically significant but statistically small differences in sentencing outcomes persisted. These findings call into question the so-called "no discrimination thesis." Furthermore, at the structural level, unwarranted sentencing disparity did not vary in a manner consistent with conflict perspective's threat hypothesis.