Bail-Setting Decisions

The bail-setting decision is one of the early court decisions made in a case, and it has attracted attention from researchers studying legal decision making. When a case is adjourned (postponed), the court must decide what to do with the defendant until the next hearing of the case—basically, should the defendant be released on bail or not? The main goal of the bail decision is to ensure that the defendant appears at court for the next hearing. The bail decision also can affect later decisions in a case. Although laws govern the bail decision-making process, they are typically vague and ill defined, thus allowing courts considerable discretion. Past research on bail decision making has mostly been conducted in the United States and the United Kingdom; researchers have aimed to describe how courts make bail decisions as well as to evaluate efforts to improve bail decision making.

Because it arises each time a case is adjourned for trial, sentence, or appeal, the bail decision is one of the most frequent legal decisions made by the criminal courts. The primary goal is to ensure that the defendant surrenders to the court at the next hearing of the case and so does not abscond. A secondary goal is that the defendant does not threaten community safety (e.g., offend) while released on bail. In the United States, the court sets a monetary amount of bail. A defendant may either be required to provide a security (deposit the amount with the court) before release, which is forfeited if he or she fails to appear in court, or be released on recognizance, which is a promise to appear, so the amount is paid only if he or she fails to appear. (For a fee, bail bondsmen can act as a surety, a third party who agrees to pay the forfeited amount to the court.) Nonfinancial conditions, such as curfew and surrendering firearms, may also be applied to bail. In the United Kingdom, defendants can be bailed (released) unconditionally; bailed with nonfinancial conditions or financial conditions, such as surety or security; or denied bail and remanded into custody. Whereas in the United States the bail decision is commonly measured on a continuous scale reflecting the monetary amount at which bail is set, in the United Kingdom the decision is typically measured as categorical because financial bail is uncommon. In most jurisdictions, bail jumping (skipping bail or absconding) is an offense.

The bail decision-making process is often governed by legislation, which is periodically revised. For instance, in the United States, currently there is the Federal Bail Reform Act of 1984 (state laws vary); in the United Kingdom, there is the Bail Act of 1976. It has often been recommended that the practice of bail decision making should adhere to the principles of due process rather than crime control. Thus, there is typically a general right to bail or pretrial liberty. However, there are exceptions if the court decides that the defendant may pose a risk of absconding or offending. The laws typically recommend that the court considers several factors (e.g., the defendant’s offense, community ties, previous convictions, prior bail record, and strength of the prosecution’s case) when judging these risks and consequently making bail decisions. Beyond this, the court is afforded considerable latitude in making bail decisions in terms of how it weights and integrates these and other factors.

The bail decision can have significant negative ramifications for defendants and their families if a defendant is denied bail or cannot raise the bail amount. For instance, defendants may lose their homes, employment, contact with their families, and their reputations, as well as experience the adverse effects of custody. In addition, evidence suggests that the bail decision may influence later decisions on a case, such as the decision to plea, convict, and sentence. Here, defendants who do not get bail are more likely to plead guilty or be convicted and are also more likely to receive a custodial sentence than their bailed counterparts.

Much of the past research, as noted previously, has investigated bail decision making in the United States and the United Kingdom. Studies have been conducted by psychologists, sociologists, criminologists, and legal scholars using methodologies such as experiments involving decision makers being presented with simulated cases, interview and questionnaire surveys of decision makers, courtroom observations of bail hearings, analyses of bail records and statistics, and analyses of bail laws. While most of this body of research has aimed to describe and explain how bail decisions are made, several studies have explored efforts to improve bail decision making. Overall, the research has yielded consistent findings.

Describing and Explaining Bail Decisions

Researchers have aimed to describe and explain bail decisions in terms of the variations in decision making and the factors that influence bail decisions. Studies have documented the variation in bail decisions made across cases and across jurisdictions (courts or decision makers), as well as within jurisdictions (courts or decision makers). There are apparent disparities in how cases that vary in their extralegal characteristics, such as the defendant’s gender and race, are dealt with. In addition, different jurisdictions (courts or decision makers) may disagree on how to deal with cases that are similar. Beyond this, there is variability where the same jurisdiction (court or decision maker) is inconsistent in dealing with similar cases on different occasions.

Research has shown that bail decisions may be influenced by both legal and extralegal factors. Legal factors include the nature and seriousness of the offense the defendant is charged with, the defendant’s previous convictions, and the strength of his or her community ties. Specifically, bail is more likely to be denied or set at a high amount if the defendant is charged with a serious offense, has previous convictions, or has weak community ties. The extralegal factors that have been found to affect bail decisions include the defendant’s gender and race and the police and prosecution’s recommendations. Here, denial of bail or its high amount is more likely to be associated with the defendant being male or non-White and a recommendation to deny bail.

In addition to identifying the factors that may influence bail decision making, some psychological studies have examined how the information is processed to form a decision. Here, there is evidence to suggest that the bail decision is the result of a simple strategy where only a few factors are considered rather than a more complex strategy involving weighting and integration of several factors.

Improving Bail Decisions

Researchers and policymakers have attempted to improve bail decision making by reducing discretion and increasing the availability of relevant information. As mentioned, the law affords the court considerable discretion in how it makes bail decisions. There have been attempts to reduce variability in bail decisions and the influence of extralegal factors as well as increase the accountability, transparency, and equity of bail decisions by limiting this discretion through the introduction of more precise guidelines. For example, in the United States, bail guidelines that specify the factors that the court should use and how they should make their risk assessments have been developed and implemented in several jurisdictions since the early 1980s. John Goldkamp and colleagues have evaluated the utility of such guidelines using randomized controlled trials and pre-post analyses. They found that decisions made under guidelines differed from those made without guidelines in several respects, including that under guidelines the bail amount was lower, there was an increase in the use of nonfinancial conditions, and there was a reduction in the time in pretrial custody. However, the impact of guidelines appears to differ across jurisdictions as they are applied and used differently.

There have also been efforts to increase the effectiveness of bail decisions by improving the court’s ability to judge the defendant’s risk of absconding on bail. The idea is that low-risk defendants, such as those who have strong ties to the community and thus may be unlikely to abscond, can be appropriately released. For example, in the United States, the Manhattan Bail Project (later renamed the New York Release on Recognizances Project) involved the collection, verification, and scoring of information on a defendant’s community ties (e.g., residence, employment, and family situation) and then providing a recommendation directly to the court concerning the defendant’s suitability for bail. In a randomized controlled trial involving real cases, the Vera Institute of Justice in 1963 found that defendants in the experimental group for whom a recommendation was provided were more likely to be bailed than those in the control group for whom the recommendation was withheld. In the United Kingdom, Bail Information Schemes provide largely positive information about a defendant’s community ties to the prosecution and defense, who then can relay it to the court. In 2002, in an experiment involving simulated cases, the author found, however, that such schemes did not have a statistically significant effect on the bail decisions made.


  1. Dhami, M. K. (2002). Do bail information schemes really affect bail decisions? Howard Journal of Criminal Justice, 41, 245-262.
  2. Dhami, M. K. (2005). From discretion to disagreement: Explaining disparities in judges’ pretrial decisions. Behavioral Sciences and the Law, 23, 367-386.
  3. Dhami, M. K., & Ayton, P. (2001). Bailing and jailing the fast and frugal way. Journal of Behavioral Decision Making, 14, 141-168.
  4. Goldkamp, J. S., & Gottfredson, M. R. (1985). Policy guidelines for bail: An experiment in court reform.Philadelphia: Temple University Press.
  5. Goldkamp, J. S., Gottfredson, M. R., Jones, P. R., & Weiland, D. (1995). Personal liberty and community safety. Pretrial release in the criminal court. New York: Plenum Press.

Return to the overview of Trial Consulting in Forensic Psychology.