By Erin Jacobs Valentine, Sarah Picard

Jurisdictions across the United States are implementing reforms to their pretrial systems to reduce the number of people who are held in pretrial detention—that is, who remain incarcerated in jail while they await the adjudication of their cases. As part of this effort, many jurisdictions are moving away from money bail as a primary means to encourage people to return for future court dates, and are instead implementing pretrial supervision, which requires clients to meet regularly with supervision staff members. Jurisdictions often attempt to match the intensity or frequency of supervision with a client’s assessed risk of failing to appear in court or being rearrested, for example by requiring more intensive supervision for clients who are assessed as being at a high risk. However, while different levels of pretrial supervision impose different burdens and costs on both jurisdictions and people awaiting the resolutions of their cases, there has been little systematic research into how they differ in their effectiveness in improving court appearance and arrest outcomes.

This report contributes new evidence in this area using retrospective data from cases initiated between January 2017 and June 2019 in two jurisdictions: one populous, urban metropolitan area in the western United States and a sparsely populated, rural county from the same region. The research team employed a regression discontinuity design, comparing the outcomes of people whose risk scores were just below and just above the cutoff for a level of supervision. They did so for four supervision levels: (1) no supervision, (2) low-intensity supervision that involved only check-ins with supervision staff members after court hearings, (3) medium-intensity supervision that also required one in-person meeting a month with a supervision staff member, and (4) high-intensity supervision that required three in-person meetings per month. The analysis uses a noninferiority approach, which tests whether the less intensive form of supervision is at least as effective as (that is, no worse than) the more intensive form.

The analysis found that:

Overall, lower-intensity supervision was as effective as higher-intensity supervision in helping clients to appear in court and avoid new arrests. When comparing each level of supervision with the next level in intensity, assignment to less intensive supervision led to similar outcomes as assignment to more intensive supervision.

Risk scores were strongly correlated with rearrest rates and modestly correlated with court appearance rates. Unsurprisingly, people with higher risk scores were more likely to be rearrested, and somewhat less likely to make scheduled court appearances. However, higher-intensity supervision did not mitigate this effect.

Overall, the analysis found no evidence that requiring people to meet more intensive pretrial supervision requirements improves outcomes. These findings suggest that policymakers should consider other strategies to encourage people to appear in court and avoid arrest, especially since supervision has costs, including monetary costs to jurisdictions and time and travel costs to clients. It is possible, for example, that strategies that involve service connections rather than supervision could be more effective. At the same time, the results indicate that more research on the use of pretrial supervision is needed. Because the regression discontinuity design of this study focuses on cases at particular risk levels—those near the cutoff risk scores that determine supervision intensity—it is possible that the results would differ for cases with other risk levels. For example, high-intensity supervision could have effects among very high-risk cases, a question that this analysis was not designed to address. Given that prior research suggests that both service and supervision resources are most effective when reserved for higher-risk and -need cases, studies focusing solely on outcomes among this group could be of great benefit to the field.

New York: MDRC, 2023. 77p.

By Paul Guerin

This study reviews felony court cases in the Second Judicial District Court with a Public Safety Assessment (PSA) and a pretrial detention (PTD) motion filed between July 2017 and June 2023. The dataset of 6,698 cases includes court data and jail data that is used to study the cases from the filing of the case to the court disposition. It is important to note this review includes the time of the COVID-19 pandemic. The COVID-19 pandemic likely had some impact on case filings, time to case dispositions, and jails admissions and lengths of stay. This review found that a slightly higher percent of court cases on which a preventive detention motion was filed was granted compared to denied motions. The study confirms other research that cases with higher FTA and NCA scores are more likely to have granted motions and that motions were most likely to be filed on cases with violent charges. We found 55% of closed cases had a conviction and were sentenced and that 43.5% were dismissed or nolled and so did not result in a conviction. Cases with denied preventive detention motions spent few days in the MDC regardless of their disposition. Cases with a granted motion that were eventually dismissed or nolled spent slightly more than 120 days in the MDC and a similar number of days in the court system. Dismissals and nolles occur at the case level for a variety of reasons including uncooperative witnesses, lack of probable cause, and because some cases might be refiled in the Federal court system. Various criminal justice system level reasons may also exist. This includes the volume of crime and arrests with resulting court case filings, the complexity of cases, and staffing among the various agencies. This preliminary review of preventive detention motion cases in the Second Judicial District Court is the first of its kind to report on the disposition of cases with a preventive detention motion. In the future more sophisticated and detailed analyses and reporting could occur that further detail the relationship between PSA scores, preventive detention motions and results, and court case dispositions.

Albuquerque: Center for Applied Research and Analysis, Institute for Social Research, University of New Mexico , 2024. 13p.

By Kristine Denman and Ella Siegrist  

The New Mexico Statistical Analysis Center received funding from the Bureau of Justice Statistics to complete a multi-phase study assessing New Mexico’s bail reform efforts. The current report examines the impact of bail reform in six New Mexico counties. This study first explores the use and amount of bond judges ordered as recorded in criminal court cases where conditions of release were set, using data from the Administrative Office of the Courts (AOC). The data includes cases disposed between 2015 and 2019, and consists of misdemeanor and felony cases, both pretrial and post-disposition. Second, using data from New Mexico county detention centers and the AOC, the study explores the impact of bail reform among defendants booked between 2015 and 2019 for a new felony offense. This allows us to examine the impact of bail reform on pretrial practices among felony defendants—the target of New Mexico’s constitutional amendment on bail reform. Specifically, the study examines four outcomes: pretrial detention practices, the use of bond, failure/success rates among those released pretrial; and court efficiency. By analyzing pre- and post- bail reform data, we found that the amendment has been successful in reducing the average amount of bond ordered and the frequency with which it is ordered. Judges, however, ordered temporary no-bond holds when issuing a warrant for arrest more frequently after bail reform. Overall, defendants involved in new felony cases were detained for a shorter period of time. However, this was not true across the board: a slightly greater percentage were subject to a short period of detention (rather than immediate release), and those detained during the entire pretrial period spent more time in jail post-reform. During the pretrial period, new violent offenses increased slightly by 2%; new offenses overall increased by 1%. Failures to appear were more common after bail reform, with a 5% increase, but this varied significantly by county. In general, time to case resolution decreased post-bail reform, though cases involving defendants detained the entire pretrial period took slightly longer to resolve. 

Albuquerque: New Mexico Statistical Analysis Center   2022. 57p.

By Grant Duwe and Valerie Clark

Using a sample of nearly 2,100 people incarcerated in Minnesota’s prison system, this study examined the concurrent and predictive validity of a needs and responsivity assessment system. For concurrent validity, we evaluated the relationship between the 13 needs and responsivity domains with assessed recidivism risk levels. For predictive validity, we analyzed the association between the domains and recidivism for a subsample that had been released from prison prior to 2023. The hypothesized needs domains—anti-social thinking, anti-social peers, education, employment, substance use disorder, housing/homelessness, and family/domestic—were significantly associated with assessed and observed recidivism, while most of the hypothesized responsivity domains—mental health, religiosity, motivation and learning style—were not. The results suggest self-identity is a distinct criminogenic need. Gender and racial/ethnic differences for concurrent and predictive validity were relatively minimal across the 13 domains.

St. Paul: Minnesota Department of Corrections, 2023. 33p.

By Megan T. Stevenson and Jennifer L. Doleac

Virginia adopted a risk assessment to help determine sentencing for sex offenders. It was incorporated as a one-way ratchet toward higher sentences: expanding the upper end of the sentence guidelines by up to 300 per cent. This led to a sharp increase in sentences for those convicted of sexual assault. More surprisingly, it also led to a decrease in sentences for those convicted of rape. This raises two questions: (a) why did sentencing patterns change differently across these groups, and (b) why would risk assessment lead to a reduction in sentence length? The first question is relatively easy to answer. While both groups saw an expansion in the upper end of the sentencing guidelines, only sexual assault had the floor lifted on the lower end, making leniency more costly. The second question is less straightforward. One potential explanation is that the risk assessment served as a political or moral shield that implicitly justified leniency for those in the lowest risk category. Even though the risk assessment did not change sentencing recommendations for low-risk individuals, it provided a 'second opinion' that could mitigate blame or guilt should the low-risk offender go on to reoffend. This decreased the risks of leniency and counterbalanced any increase in severity for high-risk individuals.

University of Toronto Law Journal, Volume 73, Supplement 1, 2023, pp. 59-72

By: SHAMENA ANWAR, JOHN ENGBERG, ISAAC M. OPPER, LEAH DION

The use of artificial intelligence (AI) methods to aid with decisionmaking in the criminal justice system has widely expanded in recent years with the increased use of risk assessments. Nowhere has this shift been more dramatic than in the widespread adoption of AI-enabled risk assessment tools to aid in pretrial detention decisions.

Despite the promise of pretrial risk assessment tools, the ways in which these tools have been implemented has limited potential progress. The vast majority of jurisdictions that have implemented these tools have essentially provided these risk assessment recommendations to judges in an advisory manner and generally cannot require judges to follow the recommendations when making their pretrial release decisions. Studies indicate that judges frequently ignore the recommendations of the risk assessment instrument; as a result, the adoption of these risk assessment tools has not had much impact on reducing the use of monetary bail and pretrial detention.

In this report, the authors investigate the factors that are predictive of whether judges follow risk assessment recommendations and identify the impacts to pretrial detention, public safety, and racial disparities when judges follow the recommendations more often.

RAND Research - Published Sep. 5, 2024

By Julius Yam 

Authoritarianism has significant implications for how judges should discharge their duties. How should judges committed to constitutionalism conduct themselves when under authoritarian pressure? To answer this question,the article proposes a two-step adjudicative framework, documents a variety of judicial strategies, and proposes how principles and strategies can and should be incorporated into the framework in different scenarios. The first step of the adjudicative framework involves judges identifying the ‘formal legal position’ while blindfolding themselves to extra-legal factors (such as potential authoritarian backlash). In the second step, depending on the level of risk incurred by maintaining the formal legal position, judges should lift the blindfold to check whether, and if so how, the formal legal position should be supplemented with or adjusted by judicial strategies. Through this analysis, the article offers a guide to judicial reasoning under authoritarianism 

Modern Law Review Limited.(2023) 00(0) MLR 

By Sigrid Geralde Clara van Wingerden.

The sentencing decision of the judge might be the most important decision in the criminal proceedings, not only because of the impact the punishment has on the offender, but also because the sentencing decision is a cornerstone of the legitimacy of the entire criminal justice system. Nonetheless, there still are questions about the factors judges take into account when making their sentencing decision. This study aims to improve our understanding of the sentencing decisions judges make.The developments in criminal justice practices as regards the emergence of ‘actuarial justice' have directed the focus of this study to risk-based sentencing: are offenders with a high risk of reoffending more likely to be sentenced to imprisonment and to longer prison terms than low-risk offenders? To what extent do judges take information into account on the risk-related personal characteristics of the offender, such as unemployment, ties to family or friends, or drug usage, when making their sentencing decision?Using uniquely detailed data on risk-related social circumstances of the offender, and advanced quantitative and qualitative research methods, this study provides in-depth insight into sentencing.

Leiden: Leiden University, 2014. 215p.