Disparities in the Pretrial Process: Race, Ethnicity, and Citizenship

Abstract

Prior work establishes that Black and Latino people face harsher treatment during the pretrial phase of the justice system. Yet, the mechanisms underlying pretrial racial and ethnic disparities remain unclear. Using multiple administrative data sources from a large jurisdiction in the southeast, we examined the influence of race, ethnicity, and citizenship on judicial decisions and defendant outcomes among people booked for a felony (n = 29,735). Of particular interest was the extent to which previously observed racial and ethnic disparities may be attributable to citizenship. Our results indicate that (1) Black and Latino people receive more punitive judicial decisions and experience worse pretrial outcomes than Whites, net of citizenship status; (2) noncitizens receive more lenient judicial decisions and experience better pretrial outcomes than citizens; and (3) citizenship status interacts with ethnicity, such that among noncitizens, Latinos suffer the worst outcomes of all groups. Implications for research and theorizing are discussed.

Keywords:

• Pretrial detention
• bail
• race
• ethnicity
• citizenship

Disclosure statement

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Notes

1 It is worth mentioning that the commonly used term to describe the case process before adjudication, “pretrial,” is somewhat of a misnomer since actual trials are quite rare in the American context.

2 An alternative to secured money bail in which a person must pay money upfront to obtain release is unsecured money bail in which a person does not pay money upfront but may be required to make a monetary payment if they do not appear in court when required.

3 Unfortunately, most jurisdictions do not differentiate between different kinds of “flight” risk. But there is a big distinction to be made between willful “flight” in which a person absconds or flees the jurisdiction to avoid justice and “non-appearance” in which a person simply does not appear in court at the designated time. Gouldin (2018) argued that flight is relatively rare and that the reasons for the more common occurrence of non-appearance are usually mundane and remediable. That is, the people involved are not “risky.”

4 The common practice of calling risk assessment instruments “objective” is misleading. While the tools are “objective” in the sense that the information used in them are collected and incorporated in a systematic and reliable way, they are not necessarily “objective” in the sense that the data used are free from existing societal biases or that the values produced are necessarily valid or true indicators of risk (Robinson & Koepke, 2019).

5 Those released on money bail are required to pay a monetary amount to secure pretrial release. This can either be paid via cash or through a commercial surety agency (i.e., a bail bondsman). Those released on their own recognizance are released with no conditions beyond the obligation to return to court for any future hearings. Those released on supervised own recognizance are released pretrial, but with conditions they must satisfy to maintain their pretrial freedom (e.g., drug testing, meetings with individuals from the pretrial services department, etc.).

6 Sample selection bias is a concern when using data from a single phase of the criminal justice system. In the pretrial context, selection bias may occur if people are released from jail prior to their first appearance. To assess this possibility, we used Heckman’s (1976) two-step approach. After including the correction term in our models, we found that it did not impact our results, as such we did not include the correction term in our analyses.

7 An alternative strategy is to give OR/SOR defendants a bail amount of $0 and include them in the models. However, the decision to require money bail versus OR/SOR release and the bail amount decision are distinct but related decisions that can both result in disparities (e.g., Demuth, 2003; Demuth & Steffensmeier, 2004; Goldkamp, 1979; Ottone & Scott-Hayward, 2018). We selected our measurement strategy due to our interest in examining whether there are disparities in both the decision to set bail and the bail amounts given.

8 The booking year is not included in the tables. However, this information is available upon request.

9 In a supplemental set of analyses, we estimate the same models excluding citizenship status to establish a base effect of race/ethnicity on our outcomes (available upon request). The race/ethnicity effects were substantively similar net of citizenship, and thus our presentation and discussion of results are based on analyses that include race/ethnicity and citizenship status.