By Jennifer Skeem, Lina Montoya, Christopher Lowenkamp

IN THE U.S., many jurisdictions are trying to reduce incarceration by improving pretrial decision-making. The pretrial decision is either to release the defendant until the court date or keep the defendant in jail to prevent re-offending or absconding. Rates of pretrial detention can be remarkably high, particularly in the federal system. There, the majority of defendants are detained before trial, even though less than 10 percent are arrested for a new crime or fail to appear while on pretrial release (Cohen & Austin, 2018; see also Rowland, 2018). Pretrial detention has serious consequences, including an increased likelihood of conviction, a harsh sentence, future re-offending, and unemployment (Dobbie et al., 2018; Leslie & Pope, 2017; Lowenkamp, 2022; Oleson et al., 2017). These consequences, in turn, are disproportionately borne by Black defendants (Didwania, 2021; Dobbie et al., 2018; Kutateladze et al., 2014; Leslie & Pope, 2017). Based on a sample of over 337,000 defendants drawn from 80 federal districts, Didwania (2021) found that 68 percent of Black defendants were detained - - - - - pretrial, compared to 51 percent of White defendants. Increasingly, efforts to improve pretrial decision-making include the goal of reducing racial disparities. In pursuing this goal, stakeholders probably assume that personal bias is to blame—i.e., that racial disparities in pretrial detention reflect the influence of implicit racism on human decision-making, and therefore that (perhaps) diversity training for practitioners would prevent such discrimination (see Devine & Ash, 2022). The majority of Americans frame racism as an interpersonal rather than structural problem—meaning that they focus on “a few bad apples” who discriminate, rather than on laws, policies, and systems that have a disparate impact (Rucker & Richeson, 2021). But disparities can also reflect “upstream” structural forces like socioeconomic and geographic conditions that lead to racial differences in the likelihood of rearrest or failure to appear. Black defendants tend to have more serious criminal histories and other potential risk factors for poor pretrial outcomes than White defendants (Didwania, 2021; Grossman et al., 2022; Spohn, 2008). Because risk of rearrest or flight are legitimate considerations for pretrial release, disparities related to differences in risk are hard to address via pretrial reform. Efforts to address disparities that flow from these kinds of structural forces would better be directed toward approaches like well-timed and well-targeted early prevention programs. In short, understanding the extent to which structural factors play a role in racial disparities is a matter of primary concern for shaping effective solutions (see Beck & Blumstein, 2018). In this study, we use federal data to explore the association between place—in this case U.S. district and geographic region—and racial disparities in pretrial officers’ recommendations for detention. We focus on officers’ recommendations in the federal system for three reasons. First, pretrial officers play a central role in assisting federal judges with the pretrial release decision, and officers’ detention recommendations strongly predict detention itself (see below, Pretrial Recommendation Context). Second, we conducted this work with the Probation and Pretrial Services Office of the Administrative Office of the U.S. Courts, as part of their effort to reduce disparities by specifying targets for change. Third, the vastness and diversity of the federal system provide a unique opportunity to characterize the districts and regions of the U.S. where racial disparities in pretrial detention are greatest, so that they can be prioritized in problem-solving efforts. The federal system encompasses 93 districts that differ geographically, socially, and culturally—but they are governed by a common set

Federal Probation, 2022.