U.S. jails handled more than 10 million admissions in 2016, many of which were for people who had not been convicted of a crime.1 Instead, they were held after arrest but before trial. Whether a person is sent to jail or released after an arrest is determined by a complex web of state laws; court rules; local customs; and decisions by individual law enforcement officers, prosecutors, judges, and others. The result is that on a given day, nearly 750,000 people are held in the nation’s jails, but only about a third of them have been convicted of a crime; the remaining two-thirds are awaiting trial.2
Pretrial detention can last days, months, or even years, and this practice has far-reaching effects on defendants, victims, court systems, and communities. To help policymakers understand how the public thinks the pretrial justice system ought to work, this chartbook presents the results of a nationally representative survey of 1,215 individuals conducted in 2018 by a bipartisan team of pollsters: Benenson Strategy Group and GS Strategy Group.3 The study found that large majorities of Americans:
- Supported issuing citations that order people to appear in court to face charges rather than arresting them for nonviolent crimes.
- Said pretrial detention should be prohibited for minor criminal offenses, such as trespassing and public drunkenness.
- Wanted people charged with misdemeanors or nonviolent crimes to remain in their communities while awaiting trial.
- Supported capping pretrial wait times at as little as one week from the beginning of detention.
- Preferred detention for individuals awaiting trial for violent offenses but supported releasing some such defendants after learning more about the offenses and available options for protecting victims.
- Backed pretrial release for people whose likelihood of completing the pretrial period without a new arrest is as low as 70 percent; risk assessment tools would generally categorize these individuals as being at moderate or high risk for rearrest before trial.
- Favored spending more taxpayer dollars on treatment programs and victim services and less on jail for people who pose little public safety risk.
- Expressed serious concerns about the risks that people on pretrial release would flee prosecution or commit acts of violence but were less worried about other behaviors, such as using marijuana or missing court dates.
Support for limiting pretrial use of jail extended across party lines and was consistently high among households with crime victims or members of law enforcement. This research can help inform growing efforts by state and local policymakers to improve pretrial decision-making, better protect public safety, and administer justice fairly and efficiently.
Survey participants shared a range of personal experiences related to criminal justice issues. For instance, nearly half of Americans surveyed reported that they or an immediate family member had been arrested, and roughly the same proportion said that they or an immediate family member had been the victim of a crime. Half had a close friend or family member affected by drug addiction. One in 7 reported having members of law enforcement in their households.
More than 8 in 10 Americans indicated that the justice system should treat people who are awaiting trial more like innocent citizens than criminals. Few respondents agreed with an alternative statement that arrested people are usually guilty and should be treated accordingly. This pattern held for respondents across political parties and from households with members of law enforcement and crime victims.
When police respond to violations of the law, they often make a choice: either arrest and take the person to jail or release the individual and issue a citation, also known as a summons, that orders the person to appear in court.4 A person who is booked into jail may be incarcerated for several days before seeing a judge.
For nonviolent crimes, Americans strongly favored citation rather than jail. Respondents from households with crime victims or members of law enforcement agreed that police should generally release people accused of nonviolent crimes and give them a notice to appear in court to face the charges against them.
Law enforcement officials often confront people who are experiencing mental health crises and struggling with substance misuse. When minor criminal behavior appears to be related to such behavioral health needs, most Americans favored treatment over arrest. Democrats and Republicans varied notably in their views on this question, but support for treatment was high among respondents from both law enforcement and victim households.
A large majority agreed that some offenses, such as trespassing and public drunkenness, are too minor to justify pretrial detention.
Most Americans say that people facing nonviolent or misdemeanor charges should generally be allowed to remain at home during the pretrial period. Between a quarter and a third of respondents said such individuals should always be released, while majorities said they should usually be released. Overall, 93 percent supported pretrial release with an order to appear in court for nonviolent charges.
Violent crime is a broad category, and the seriousness of offenses varies considerably. When asked whether individuals accused of violent crimes should generally be released or jailed before trial, Americans overwhelmingly preferred jail. However, when given additional options—including pretrial release with specific conditions to protect the public, such as community supervision—more Americans favored release pending trial.
For example, 58 percent of respondents supported releasing people accused of violent crimes who do not have serious criminal histories if the release is accompanied by pretrial supervision; 27 percent strongly supported this model. For low-level violent crimes, such as a bar fight, 85 percent supported release with an order requiring the defendant to stay away from the victim or with pretrial supervision, including a majority that strongly supported release.
Judges in some jurisdictions use pretrial risk assessment tools to guide their release decisions. These instruments categorize defendants based on the pretrial success rates of others with similar characteristics, such as age and criminal history. In the examples offered in the survey, the measure of success was avoiding arrest during the pretrial period. People who fall into the Example A category are often classified by these tools as being at low risk for rearrest, while those in Examples B and C are frequently moderate to high risk.5 Most Americans supported releasing people in all three example groups based on their likelihood of pretrial success.
Respondents voiced significant concern about misconduct during the pretrial period, but not every risk worried them equally. With regard to new criminal activity, 94 percent were at least somewhat concerned about a released defendant assaulting someone (with 75 percent very concerned), but only 36 percent worried about a defendant using marijuana (with 14 percent very concerned).
Survey respondents’ worries about failure to appear in court also varied depending on the circumstances. More than half were very concerned about a defendant intentionally fleeing prosecution, but just a quarter said the same about an accused person accidentally missing 59% - 31% a court hearing.
The U.S. Constitution guarantees the right to a speedy trial but does not define “speedy.” More than 8 in 10 Americans believe that people should not have to spend long periods of time in a jail cell waiting for their trial to start, and large majorities even supported caps as short as one week, except in special circumstances.
In many jurisdictions, judges set a monetary condition, commonly known as bail, which must be paid in full or part—often with the assistance of a commercial bail bond company—before the defendant can be released from jail. These conditions can result in different release or detention outcomes depending on a person’s wealth or access to resources. When asked whether the requirement to put up money helps keep communities safe, a majority of Americans said no; support for this concept was much stronger among Democrats, independents, and law enforcement and crime victim households than among Republicans.
State and local governments rely on tax dollars to fund local criminal justice systems, including the costs of detaining people in jail during the pretrial period. When presented with different options for apportioning public safety dollars, large majorities favored shifting some of the money currently spent on pretrial detention into treatment for those with substance misuse and mental health needs and into counseling and other services for crime victims.
The policies and practices that govern arrest, pretrial release, detention in jail, and trial timelines have far-reaching effects, not only on people accused of crimes but also on victims, families, and communities. This national survey provides information about how the public thinks pretrial justice should work and finds substantial support for policies and decisions that limit the use of pretrial detention. Specifically, large majorities think jail should never be used for minor offenses, that those charged with misdemeanors and nonviolent offenses should remain in their communities while awaiting trial, and that even more serious cases, including some violent charges, can be handled without detention. These positions generally hold across political parties and households with crime victims or members of law enforcement, and suggest that police, judges, and policymakers should think critically about the use of jail for those awaiting trial and should expand and prioritize pretrial release.
Benenson Strategy Group and GS Strategy Group surveyed 1,215 U.S. adults on behalf of The Pew Charitable Trusts. They conducted the interviews by landline and cellular telephone in English and Spanish using a computer-assisted telephone interviewing system from May 2 through 14, 2018. Sixty-four percent of interviews were completed on cellphones. The margin of error is plus or minus 2.8 percentage points at the 95 percent level of confidence. When accounting for design effects, the margin of error is plus or minus 4.5 percentage points at the 95 percent level of confidence. The sampling error is dependent on sample size, so the margin of error is higher for subgroups.
The sampling frame for this study is U.S. adults. For the survey, a combination of landline and cellular random digit dial samples was used to represent all adults in the continental United States with access to either a landline or cellular telephone. Numbers for the landline sample were drawn with probabilities in proportion to their share of listed telephone households from active blocks (area code + exchange + two-digit block number) that contained one or more residential directory listings. The cellular sample was not list-assisted but was drawn through a systematic sampling from dedicated wireless 100-blocks and shared service 100-blocks with no directory-listed landline numbers.
Within this probability sample, respondents were called randomly. On landlines, interviewers asked randomly to speak with “the youngest male/female, 18 years of age or older, who is now at home.” If no person of the requested gender was at home, interviewers asked to speak with the youngest adult of the opposite gender. Unlike a landline phone, a cellphone is assumed to be a personal device, so interviewers asked whether the person who answered was 18 years of age or older to determine eligibility. Quotas were applied for gender and region based on U.S. census counts to help ensure that the sample was representative of U.S. adults.6
The survey team weighted the data for age, race, ethnicity, region, and education to reflect the composition of U.S. adults based on universe counts from the U.S. census. When comparing this document with the top line, which shows the verbatim text of all questions asked and breakdowns of responses to every question, figures may vary because of rounding. For the full survey questions and top-line results, see the chartbook webpage.7
Poll respondent demographics
- 31 percent identified as politically conservative.
- 31 percent identified as politically moderate.
- 30 percent identified as politically liberal.
- 85 percent identified as a registered voter.
- Zhen Zeng, “Jail Inmates in 2016,” U.S. Department of Justice, Bureau of Justice Statistics (2018), https://www.bjs.gov/content/pub/pdf/ji16.pdf.
- Benenson Strategy Group and GS Strategy Group, “National Criminal Justice Survey” (2018). For top-line results, see http://pewtrusts.org/-/media/assets/2018/11/pspp_jails_poll_toplines.pdf.
- In some situations, including those involving violence or an immediate threat of harm, jurisdictions might mandate an arrest.
- Laura and John Arnold Foundation, “Developing a National Model for Pretrial Risk Assessment” (2013), https://www.arnoldfoundation.org/wp-content/uploads/2014/02/LJAF-research-summary_PSA-Court_4_1.pdf; Colorado Association of Pretrial Services, “The Colorado Pretrial Assessment Tool (CPAT)” (2015), http://capscolorado.org/yahoo_site_admin/assets/docs/CPAT_Manual_V21_06-29-2015.179175025.pdf; Marie VanNostrand and Kenneth J. Rose, “Pretrial Risk Assessment in Virginia,” Virginia Department of Criminal Justice Services and Virginia Community Criminal Justice Association (2009), https://www.dcjs.virginia.gov/sites/dcjs.virginia.gov/files/publications/corrections/virginia-pretrial-risk-assessment-report.pdf.
- U.S. Census Bureau, “Decennial Census of Population and Housing” (2010), https://www.census.gov/programs-surveys/decennial-census/data/datasets.2010.html.
- The Pew Charitable Trusts, “Americans Favor Expanded Pretrial Release, Limited Use of Jail” (2018), https://www.pewtrusts.org/research-and-analysis/issue-briefs/2018/11/americans-favorexpanded-pretrial-release-limited-use-of-jail.