Life support for ailing hypotheses: Modes of summarizing the evidence for racial discrimination in sentencing.

Abstract

A recent review concluded that the weight of available evidence contradicts a hypothesis of widespread overt racial discrimination in sentencing in the United States. Yet brief summaries of this body of research, found in textbooks, monographs, and the literature review sections of journal articles, commonly conveyed the opposite impression, sustaining an image of extensive support for the hypothesis. These summaries are examined to determine how this misleading impression was conveyed. Five common practices are identified: 1. Selective Citation: A biased selection of studies were reviewed or cited. 2. Letting the Evidence Speak for Itself: Overrepresentation of blacks in prison or among the executed relative to their share of the population was noted, with readers left to draw their own conclusions. 3. The Mixed Bag: Impressive lists of studies supposedly supporting a discrimination hypothesis were padded, by lumping together studies concerning a variety of criminal justice processes besides sentencing, and various legally irrelevant defendant traits besides race. 4. Research Democracy-All Studies are Created Equal: Equal weight was implicitly given to all studies regardless of methodological rigor, in a field where it was the least rigorous studies which were most likely to support the discrimination hypothesis. 5. Magnanimous Neutrality: The available evidence was described as mixed or ambiguous, which, while technically accurate, was a misleading description of evidence which largely contradicted the discrimination hypothesis.