From [HERE] This review found evidence in the body of research discussed above that prosecutorial decision making is associated with racial and ethnic disparities in case outcomes. Most of the 34 studies reviewed here suggest that defendants’ or victims’ race directly or indirectly influence case outcomes, even when a host of other legal and extra-legal factors are taken into account. This is not to suggest, however, that the research shows that prosecutors always treat blacks and Latinos or Latinas more punitively: while a greater number of studies found that minority defendants are more likely to be prosecuted, held in pretrial detention, and to receive other harsh treatment, researchers also found proof of prosecutors treating white defendants more harshly for certain offenses and at certain discretion points. In other words, the research findings vary noticeably by the type of data and analyses used.

When drawing conclusions from the characterizations of researchers’ findings presented in this review, readers should consider the following four caveats. First, studies that did not find any differences by race and ethnicity should not be used as evidence for the absence of any discriminatory practices, because data and analytical limitations may have contributed to the absence of findings. Second, many of the studies were based on limited sample sizes in a single or a few jurisdictions that yielded insufficient evidence for supportable generalizations of findings. Third, some studies that suggest that minorities are treated more leniently should be also viewed with an eye toward whether more lenient treatment of blacks and Latinos may stem from a general devaluation of their communities or corrective action by prosecutors in response to aggressive or improper policing practices. Fourth, it is also possible that studies finding differences by race and ethnicity were more likely to be published, and thus are overrepresented in this review.50 Finally, while we made every effort to accurately summarize findings from the 34 studies in this review, we strongly recommend reading the studies in their entirety as they appeared in peer-reviewed journals (see References, p.18).