Pretrial Misconduct Research Articles

Algorithmic Recommendations When the Stakes Are High: Evidence from Judicial Elections

We ask whether increased public scrutiny leads to the more effective use of predictive algorithms. We focus on the context of bail, where judges face heightened public scrutiny during competitive partisan elections. We find that judges up for reelection are much more likely to follow the algorithmic recommendation to detain high-risk defendants just before an election. However, release decisions return to normal shortly after the election, and there is little change in pretrial misconduct rates, indicating that heightened public scrutiny, at least through competitive partisan elections, will not lead to the more effective use of predictive algorithms in bail.

Standing and Pre-trial Misconduct: Hypocrisy, ‘Separation’, Inconsistent Blame, and Frustration

Existing justifications for exclusionary rules and stays of proceedings in response to pre-trial wrongdoing by police officers and prosecutors are often thought to be counter-productive or disproportionate in their consequences. This article begins to explore whether the concept of standing to blame can provide a fresh justification for such responses. It focuses on a vice related to standing—hypocrisy—and a related vice concerning inconsistent blame. It takes seriously the point that criminal justice agencies, although all part of the State, are in real terms separated from each other, and analyses the so-called separation thesis (or theses). It concludes that hypocrisy and inconsistent blame arguments could plausibly justify exclusion and stays only in relation to lower-level offending, and even there only indirectly. This is in the sense that exclusion and stays are expressions of judicial frustration with other bodies for their failure to take pre-trial wrongdoing seriously.

Measuring Racial Discrimination in Bail Decisions

We develop new quasi-experimental tools to measure disparate impact, regardless of its source, in the context of bail decisions. We show that omitted variables bias in pretrial release rate comparisons can be purged by using the quasi-random assignment of judges to estimate average pretrial misconduct risk by race. We find that two-thirds of the release rate disparity between White and Black defendants in New York City is due to the disparate impact of release decisions. We then develop a hierarchical marginal treatment effect model to study the drivers of disparate impact, finding evidence of both racial bias and statistical discrimination. (JEL J15, K42)

Differential Prediction by Race in IRAS-PAT Assessments: An Application of Debiasing Strategies

There remain serious concerns about the potential for pretrial risk assessments to exacerbate racial disparities. Yet, current evidence on differential prediction in pretrial risk assessments is limited. The present investigation tests for differential prediction by race as an indication of bias in Indiana Risk Assessment System–Pretrial Assessment Tool (IRAS-PAT) assessments. Using pooled data drawn from a five-county IRAS-PAT validation, which included 689 Black and 2,850 White defendants, we primarily used a hierarchical regression-based approach to test between-group differences in the slopes of regression lines. Where slope-based bias was present, differential prediction was reevaluated once algorithmic corrections were applied. Findings showed IRAS-PAT assessments produced less accurate predictions of pretrial misconduct for Black defendants relative to White defendants. Only one debiasing strategy—which accounted for item-level differences across subgroups—corrected differential prediction. Debiasing strategies can mitigate differential prediction but may have limited utility for local jurisdictions under current legal frameworks.

Bail, jail, and pretrial misconduct: The influence of prosecutors

Bail, jail, and pretrial misconduct: The influence of prosecutors

The US Pretrial System: Balancing Individual Rights and Public Interests

In this article, we review a growing empirical literature on the effectiveness and fairness of the US pretrial system and discuss its policy implications. Despite the importance of this stage of the criminal legal process, researchers have only recently begun to explore how the pretrial system balances individual rights and public interests. We describe the empirical challenges that have prevented progress in this area and how recent work has made use of new data sources and quasi-experimental approaches to credibly estimate both the individual harms (such as loss of employment or government assistance) and public benefits (such as preventing non-appearance at court and new crimes) of cash bail and pretrial detention. These new data and approaches show that the current pretrial system imposes substantial short-and long-term economic harms on detained defendants in terms of lost earnings and government assistance, while providing little in the way of decreased criminal activity for the public interest. Non-appearances at court do significantly decrease for detained defendants, but the magnitudes cannot justify the economic harms to individuals observed in the data. A second set of studies shows that that the costs of cash bail and pretrial detention are disproportionately borne by Black and Hispanic individuals, giving rise to large and unfair racial differences in cash bail and detention that cannot be explained by underlying differences in pretrial misconduct risk. We then turn to policy implications and describe areas of future work that would enable a deeper understanding of what drives these undesirable outcomes.

Use of Risk Assessments in Pretrial Supervision Decision-Making and Associated Outcomes

Despite the growing use of both pretrial risk assessment and supervision as pretrial reform strategies, there has been limited investigation on the effectiveness of risk assessment-informed pretrial supervision. We conducted a multi-site, retrospective investigation in 1,505 pretrial defendants from four local jurisdictions to examine risk assessment-informed pretrial supervision decisions and associated pretrial misconduct outcomes. Our findings showed pretrial supervision decisions were generally consistent with structured guidelines and defendant risk classifications. Use of bond and electronic monitoring had little impact on pretrial misconduct. However, more frequent pretrial monitoring was associated with higher rates of pretrial misconduct across all risk levels. Reducing supervision conditions and monitoring for low-risk defendants, in particular, may help reduce rates of misconduct in pretrial populations.

Measuring Racial Discrimination in Algorithms

There is growing concern that the rise of algorithmic decision-making can lead to discrimination against legally protected groups, but measuring such algorithmic discrimination is often hampered by a fundamental selection challenge. We develop new quasi-experimental tools to overcome this challenge and measure algorithmic discrimination in the setting of pretrial bail decisions. We first show that the selection challenge reduces to the challenge of measuring four moments: the mean latent qualification of white and Black individuals and the race-specific covariance between qualification and the algorithm’s treatment recommendation. We then show how these four moments can be estimated by extrapolating quasi-experimental variation across as-good-as-randomly assigned decision-makers. Estimates from New York City show that a sophisticated machine learning algorithm discriminates against Black defendants, even though defendant race and ethnicity are not included in the training data. The algorithm recommends releasing white defendants before trial at an 8 percentage point (11 percent) higher rate than Black defendants with identical potential for pretrial misconduct, with this unwarranted disparity explaining 77 percent of the observed racial disparity in algorithmic recommendations. We find a similar level of algorithmic discrimination with regression-based recommendations, using a model inspired by a widely used pretrial risk assessment tool. Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

Effects of pretrial risk assessments on release decisions and misconduct outcomes relative to practice as usual

PurposeWe conducted a multi-site, quasi-experimental investigation of the effects of pretrial risk assessments on pretrial release decisions and misconduct outcomes relative to practice as usual. MethodsUsing a multiple non-equivalent comparison group design, we matched 2631 pretrial defendants who received a risk assessment during a 1-year pilot period to two comparison groups of defendants who did not receive a risk assessment and were processed in the same year (n = 1580) or in the year prior to the pilot period (n = 3185). Weighted multilevel regression analyses were conducted separately for each comparison to examine effects of the pilot risk assessment condition on pretrial release and pretrial misconduct outcomes. ResultsRelative to comparison groups, defendants with risk assessments were more likely to receive non-financial release. When risk assessment-guided decisions adhered to structured guidelines, defendants with risk assessments had higher rates of pretrial release and spent less time in pretrial detention. Risk assessments were associated with slightly higher rates of non-violent and new re-arrests, but not failure to appear, relative to comparison conditions. ConclusionsPretrial risk assessments can facilitate non-financial release, though with a potentially higher rate of pretrial re-arrest. Structured guidelines may help maximize pretrial release while minimizing misconduct.

Five-County Validation of the Indiana Risk Assessment System – Pretrial Assessment Tool (IRAS-PAT) using a Local Validation Approach

Local jurisdictions are increasingly using pretrial risk assessment instruments to assess risk of pretrial misconduct and inform release decisions. We adopted a local validation approach to examine the predictive validity of Indiana Risk Assessment System – Pretrial Assessment Tool (IRAS-PAT) assessments in 3,739 unique pretrial defendants across five Indiana counties. Jail, court, and pretrial risk assessment records were matched within each jurisdiction to examine pretrial misconduct outcomes (i.e., any arrest, any new arrest, and any failure to appear) during the case processing period. Area Under the Curve (AUC) estimates showed good-to-excellent levels of predictive accuracy for total scores for all outcomes (AUC Range: 0.67-0.72). Multivariable models showed defendants assessed at High (OR Range: 5.42-8.62) and Moderate (OR Range: 2.56-3.08) risk had higher rates of pretrial misconduct relative to those assessed at Low risk. Findings provide strong evidence for the predictive accuracy of IRAS-PAT assessments overall, though some item-level considerations are noted.

Racial Bias in Bail Decisions*

Abstract This article develops a new test for identifying racial bias in the context of bail decisions—a high-stakes setting with large disparities between white and black defendants. We motivate our analysis using Becker’s model of racial bias, which predicts that rates of pretrial misconduct will be identical for marginal white and marginal black defendants if bail judges are racially unbiased. In contrast, marginal white defendants will have higher rates of misconduct than marginal black defendants if bail judges are racially biased, whether that bias is driven by racial animus, inaccurate racial stereotypes, or any other form of bias. To test the model, we use the release tendencies of quasi-randomly assigned bail judges to identify the relevant race-specific misconduct rates. Estimates from Miami and Philadelphia show that bail judges are racially biased against black defendants, with substantially more racial bias among both inexperienced and part-time judges. We find suggestive evidence that this racial bias is driven by bail judges relying on inaccurate stereotypes that exaggerate the relative danger of releasing black defendants.

Does GPS supervision of intimate partner violence defendants reduce pretrial misconduct? Evidence from a quasi-experimental study

This research examines the effect global positioning system (GPS) technology supervision has on pretrial misconduct for defendants facing intimate partner violence charges. Drawing on data from one pretrial services division, a retrospective quasi-experimental design was constructed to examine failure to appear to court, failure to appear to meetings with pretrial services, and rearrest outcomes between defendants ordered to pretrial GPS supervision and a comparison group of defendants ordered to pretrial supervision without the use of monitoring technology. Cox regression models were used to assess differences between quasi-experimental conditions. To enhance internal validity and mitigate model dependence, we utilized and compared results across four counterfactual comparison groups (propensity score matching, Mahalanobis distance matching, inverse probability of treatment weighting, and marginal mean weighting through stratification). Pretrial GPS supervision was no more or less effective than traditional, non-technology based pretrial supervision in reducing the risk of failure to appear to court or the risk of rearrest. GPS supervision did reduce the risk of failing to appear to meetings with pretrial services staff. The results suggest that GPS supervision may hold untapped case management benefits for pretrial probation officers, a pragmatic focus that may be overshadowed by efforts to mitigate the risk of pretrial misconduct. Further, the results contribute to ongoing discussions on bail reform, pretrial practice, and the movement to reduce local jail populations. Although the cost savings are not entirely clear, relatively higher risk defendants can be managed in the community and produce outcomes that are comparable to other defendants. The results also call into question the ability of matching procedures to construct appropriate counterfactuals in an era where risk assessment informs criminal justice decision-making. Weighting techniques outperformed matching strategies.

Predicting pretrial misconduct with drug tests of arrestees: Evidence from eight settings

Identifying defendants at high risk of pretrial misconduct is a major problem for the judiciary. Currently, some have argued that testing arrestees for recent drug use is one way to distinguish between those who will and those who will not commit pretrial misconduct. The research reported here questions whether the incremental predictive power resulting from drug testing always improves predictions of pretrial misconduct. Using survival analysis to study time until rearrest and a probit model to analyze the occurrence of a failure to appear, we show that urine test results have no consistent power to predict pretrial misconduct after accounting for defendant's criminal records, community ties, and other factors commonly known by the court. These results are based on our analysis of eight data sets from different locales, time periods, and age groups.

Pretrial Drug-Testing Experiments in Milwaukee and Prince George's County: The Context of Implementation

This article describes the results of field experiments involving pretrial drug-testing demonstration projects undertaken in Milwaukee, Wisconsin, and Prince George's County, Maryland, in 1988 and 1989. As part of a six-site effort sponsored by the Bureau of Justice Assistance of the U. S. Department of Justice, the demonstration projects in these two sites sought to implement drug-testing programs at the pre-bail and pretrial release stages of the criminal process to serve as a means for deterring pretrial misconduct (i.e., failure to appear in court and rearrest for crimes during pretrial release). The underlying hypotheses of the projects tested by the field experiments were (a) that pre-bail drug testing would contribute more powerful predictive data that would allow pretrial release decisionmakers better to assess the risks posed by defendants being considered for release (over and above the power of other existing information) and (b) that pretrial drug monitoring during pretrial release would deter defendants from crime and flight better than release without drug monitoring. The experimental results reported found no evidence to support either claim. However, the article argues that experimental findings cannot be meaningfully interpreted in isolation from knowledge of the implementation context in the sites. In fact, the demonstration programs in each site experienced challenging obstacles in implementing the programs as envisaged. The need to take into account the difficulties posed by the challenges of implementing new programs in demonstration sites does not, however, undermine the strength of the findings reported. Rather, the implementation difficulties should, in themselves, be seen as important findings concerning the feasibility of such programs.

Drug Tests and the Prediction of Pretrial Misconduct: Findings and Policy Issues

Growing concern about the impact of drug abuse and the proliferation of drug-related caseloads in urban courts has fueled interest in urine screening and monitoring programs to detect drug use among pretrial defendants. Research on the predictive utility of urine tests in New York City and elsewhere suggests that this information does not improve the ability to classify high-risk offenders, nor does urine monitoring appear to reduce pretrial misconduct under supervised release programs. Coupled with technical and process concerns surrounding drug tests, the evidence thus far suggests caution in the adoption of pretrial drug-testing programs.

Drug Testing and Pretrial Misconduct: An Experiment on the Specific Deterrent Effects of Drug Monitoring Defendants on Pretrial Release

Recent attempts to improve the validity of prediction models of pretrial misconduct have incorporated the defendant's drug use, as indicated by drug test results. A related and important issue concerns the specific deterrent effect of drug testing the released pretrial population. Does drug monitoring of released defendants alter their drug use and chances for pretrial misconduct? The authors use experimental data from Pima County and Maricopa County, Arizona, where defendants were randomly assigned to drug monitoring and to nonmonitoring groups to explore this issue. Their findings show that in two Pima County samples, there was only a slight reduction in the rate of pretrial rearrest, and there were no differences for failure to appear at trial. In Maricopa County, their first sample shows no difference in the rate of pretrial misconduct between the monitored and nonmonitored groups. The second sample shows the monitored group to have a higher rate of pretrial failure, contrary to expectations. The authors conclude their discussion by noting the research and the policy implications these results hold for future pretrial drug testing efforts.

Pretrial Urine-Testing in the District of Columbia: Its Usefulness for Risk Classification and as a “Signaling Device” for Release Risk

Systematic urine-test data from arrestees in Washington, D.C. improved risk classification of arrestees with regard to pretrial rearrest, failure-to-appear and overall pretrial misconduct; urine testing of defendants released before trial was an effective “signaling” mechanism of post-release risk. Successful participants in the pretrial urine-testing program had rates of pretrial rearrest, failure-to-appear and overall pretrial misconduct that were about one-half the rates for individuals who failed to show up for urine testing or who dropped out after only a few tests.

Drug use and pretrial misconduct in New York city

Is recent drug use significantly associated with pretrial misconduct? Does consideration of recent drug use enhance risk classification among a sample of persons who have time free pending the disposition of their cases? Using data on arrestees in Manhattan, this paper examines these issues and some related questions. To measure recent drug use, urine samples were collected from persons shortly after their arrest and tested for four drugs: heroin, cocaine, PCP, and methadone. Two measures of pretrial misconduct are considered: whether a defendant fails to appear for a scheduled court date (FTA) and whether a defendant is rearrested prior to case disposition. Censored probit models are used to estimate the statistical association between drug test results and pretrial misconduct. Results show that drug test results are significantly associated with pretrial misconduct over and above the information typically available to judges at the time release decisions are made. Some implications of these findings for pretrial decision making are discussed.

An Evaluation of Federal Pretrial Services Agencies' Impact On Pretrial Decisions and Outcomes

In 1966 Congress created demonstration pretrial services agencies in ten federal district courts to aid judges in complying with the Bail Reform Act of 1966. This article used secondary analysis of data collected as part of the demonstration project to determine its impact on pretrial release decisions and defendant conduct. The results suggest that the agencies did create changes in the type of release conditions through changes in judicial decision making. Furthermore, these changes did not create significant increases in the rates of pretrial misconduct as feared by some. The final discussion focuses on the relationship between the agencies and the earlier Bail Reform Act of 1966.

Pretrial Release in Federal Courts

Pretrial release is the legal process by which defendants who are accused of criminal offenses are released from detention pending trial. Evaluators are frequently concerned with the qualitative and quantuative importance of defendant and offense variables in the determination of which defendants gain release and, for those who are released, which defendants will engage in pretrial misconduct. This article presents a method for deriving unbiased parameter estimates, using maximum likelihood techniques m a structural equation paradigm, for the release decision and indicates how these techniques can be extended to the analysis of pretrial misconduct.