U.S. flag

An official website of the United States government, Department of Justice.

Meta-Analysis of Race and Sentencing Research: Explaining the Inconsistencies

NCJ Number
Journal of Quantitative Criminology Volume: 21 Issue: 4 Dated: December 2005 Pages: 439-466
Date Published
December 2005
28 pages

This study employed a meta-analysis of 71 published and unpublished studies to investigate whether African-Americans are treated more harshly by the criminal justice system than similarly situated Whites.


An impressive body of research has sprung up around the accusation that the American criminal justice system treats African-American suspects and defendants more harshly than their White counterparts. In order to investigate this claim, the author employed a quantitative, meta-analytical strategy to systematically and comprehensively review the available research concerning the influence of race on sentencing outcomes. In all, 71 published and unpublished studies were analyzed in terms of 5 types of sentencing outcomes: (1) length of incarceration, (2) imprisonment decisions, (3) ordinal scales of sentence severity, (4) discretionary lenience, and (5) discretionary punitiveness. The analysis was performed on the logged odds-ratio using the inverse variance method. Results indicated that independent of other measured variables, African-Americans were sentenced more harshly than their similarly situated White counterparts. In terms of policy, the findings suggested that the use of structured sentencing mechanisms were associated with smaller unwarranted sentencing disparities. Future research should focus on analyzing cases at lower levels of aggregation to avoid pooling cases from several jurisdictions. Footnotes, figure, tables, references

Date Published: December 1, 2005