Journal of Criminal Justice, Vol. 26, No. 6, pp. 513–523, 1998 Copyright © 1998 Elsevier Science Ltd Printed in the USA. All rights reserved 0047-2352/98 $–see front matter PII S0047-2352(98)00028-2 513 RACE AND SENTENCING: A META-ANALYSIS OF CONFLICTING EMPIRICAL RESEARCH RESULTS Travis C. Pratt Division of Criminal Justice University of Cincinnati Cincinnati, Ohio 45220 ABSTRACT This paper employs meta-analytic methodology to evaluate the research results of a number of sepa- rate race and sentencing studies in the context of three competing theoretical frameworks (the differen- tial involvement, direct-impact, and interactionist perspectives), and the empirical evidence supporting each. The data indicate that although the effect size of race on sentencing is not statistically significant at this level of aggregation, its relationship to sentencing is significantly conditioned by the measure- ment techniques employed by researchers. Specifically, this study shows that certain methods of classify- ing racial groups may mask the true effect of race on sentencing decisions. The implications of this find- ing for future research are discussed. © 1998 Elsevier Science Ltd INTRODUCTION There are currently 1,471 African American inmates in prisons and jails per 100,000 African American U.S. residents, compared to 207 Cau- casian inmates per 100,000 Caucasian residents. The rate of incarceration of African Americans, therefore, is seven times that of Whites (Beck and Gilliard, 1995). One existing problem, however, is that research attempting to explain this phenomenon in terms of racial sentencing disparity is contradictory and inconclusive. Researchers have conducted numerous em- pirical studies in an effort to discover an expla- nation for the disproportionate representation of racial and ethnic minorities in our nation’s cor- rectional institutions. Studies examining the im- pact of race on the sentencing of criminal of- fenders has yielded mixed results—some studies have found that race plays a significant role in sentencing wholly independent of all le- gally relevant factors; some have found that it does not; while others have found that race im- pacts sentencing decisions through its interac- tion with other variables. Any given study regarding the relationship between race and sentencing generally tends to support one of three theoretical perspectives which will be referred to throughout this paper as: (1) the dif- ferential involvement perspective; (2) the direct- impact perspective; or (3) the interactionist per- spective.