How Community Bail Reform Changes Who Bears The Cost Of Trust

Cash bail in the United States has a specific history. The Eighth Amendment forbids "excessive bail," but it does not guarantee bail; it presumes some version of bail exists. For most of American history, bail was set at modest amounts and was intended as a mechanism to ensure court appearance, not to preventively detain.

The shift began in the 1960s and 1970s, when the commercial bail bond industry — a uniquely American institution, illegal in most of the rest of the world — grew to dominate pretrial release. Judges increasingly set bail at amounts that defendants could not pay directly, assuming they would use a bondsman. Bondsmen charge a non-refundable ten percent fee. If a defendant pays a thousand dollars to a bondsman for a ten-thousand-dollar bail, that thousand dollars is gone regardless of whether the defendant is convicted or acquitted.

The bondsman's incentive is to post bail for people who are likely to show up to court, because failure to appear forfeits the bond. This creates a de facto screening system: the bondsman decides who is "trustworthy" enough to bond out. People without family or community resources, people with erratic addresses, people with mental illness or addiction — the people most likely to need support — are the least likely to be bonded.

The cash bail system therefore has three interlocking effects:

First, it filters pretrial release by wealth. Poor defendants stay in jail. Wealthy defendants go home.

Second, it privatizes an enormous amount of the pretrial system to the bail bond industry, which extracts roughly two billion dollars per year in non-refundable fees from defendants and their families — most of whom are poor.

Third, it produces pretrial incarceration rates that are, by international standards, shocking. The United States holds more pretrial detainees per capita than almost any other country on earth. In most European democracies, pretrial detention is rare and brief. In the United States, it is common and sometimes lasts years.

The foundational research on pretrial release and court appearance comes from decades of pretrial services data in jurisdictions that have experimented with non-monetary release.

The D.C. Pretrial Services Agency, which has operated a largely non-monetary pretrial system since the 1990s, consistently reports that about eighty-eight to ninety percent of defendants released pretrial make all court appearances. About ninety percent are not rearrested while on release. These numbers have been stable for decades and are published in annual reports.

The New Jersey Judiciary's reports on bail reform, published annually since 2017, show similar patterns. Pretrial jail population dropped forty-four percent between December 2016 and December 2017. Court appearance rates remained stable at approximately ninety-two percent. New criminal activity during pretrial release remained stable. By 2019, the pretrial jail population had stabilized at a level about thirty-five to forty percent below pre-reform levels.

A 2020 analysis by the Center for Court Innovation in New York found similar patterns after New York's January 2020 bail reform: court appearance rates in affected cases stayed at approximately ninety-two to ninety-five percent, and the rate of new arrests during pretrial release for people affected by reform was statistically indistinguishable from pre-reform rates.

A 2023 study published in Criminology and Public Policy, examining bail reforms across multiple states, concluded that there is no consistent evidence that eliminating cash bail increases crime. Some studies find slight decreases in certain categories; some find no change; none find the kind of increases that would support the "dangerous criminals freed" narrative.

The Bail Project, which has bailed out more than thirty thousand people since 2018, reports that ninety-one percent of the people they bail out return to all court dates. The Bronx Freedom Fund, operating since 2007, reports similar numbers. These are people the system had deemed untrustworthy enough to require cash bail, and who had no money. When bailed out, they showed up.

This is the empirical finding that keeps getting buried: the system's judgment about who is "trustworthy" is not based on trustworthiness. It is based on wealth. When you strip the wealth filter out, the appearance rates stay the same.

D.C. has the longest-running experience with functional elimination of cash bail. The Bail Reform Act of 1966, a federal law applying to D.C., created a strong presumption for non-monetary release. The Pretrial Services Agency was established to provide supervision, reminders, and support for defendants on pretrial release.

The D.C. system is not "no bail." It is "release with conditions." Those conditions can include: court date reminders, check-ins with pretrial services, drug testing, electronic monitoring, substance abuse treatment, mental health services, and in some cases, residence requirements. For defendants charged with serious violent offenses, detention without bail is possible under specific statutory criteria.

The annual PSA reports show that in a given year, approximately ninety-four percent of defendants are released pretrial, about six percent are detained. Of those released, about eighty-nine percent make all court appearances, about ninety-one percent are not rearrested.

D.C. is not crime-free. D.C. has crime problems like most American cities. But the evidence does not support the claim that eliminating cash bail caused or worsened those problems. Crime rates in D.C. track roughly with national trends over the decades that cash bail has been largely absent.

New Jersey's reform is the most studied recent example because it happened in a single, clear transition in January 2017. Before reform, about seventy-five percent of New Jersey pretrial jail inmates were held because they could not afford bail. After reform, cash bail essentially disappeared as a mechanism, replaced by the Public Safety Assessment (a risk assessment tool developed by the Arnold Foundation) and a menu of release conditions.

The first comprehensive evaluation, conducted by the New Jersey Administrative Office of the Courts and independent researchers, found:

- Pretrial jail population dropped forty-four percent in the first year. - Court appearance rates remained stable at about ninety-two percent. - Pretrial re-arrest rates remained stable at about fourteen percent (comparable to pre-reform rates and to national averages for pretrial populations). - Racial disparities in pretrial detention narrowed but did not disappear.

The reform survived political pressure. Unlike New York, New Jersey has not substantially rolled back its bail reform, in part because the data was clear and the law had broader political support, including from the state's Chief Justice Stuart Rabner and from law enforcement organizations that had initially opposed reform.

New York's story is different. The 2019 reform, which took effect January 2020, eliminated cash bail for most misdemeanors and non-violent felonies. The legislation was the result of years of organizing by VOCAL-NY, the Katal Center, Brooklyn Community Bail Fund, and allied organizations.

Implementation began in January 2020. By March 2020, the COVID pandemic had disrupted court operations, jail populations, and data collection, making clean pre- and post-reform comparisons difficult. By early 2021, a political backlash emerged. Tabloid coverage focused on individual cases of people released pretrial who were then arrested for new crimes. Mayor-elect Eric Adams made bail reform rollback a campaign priority.

The rollbacks came in stages — 2020, 2022, 2023 — expanding judicial discretion to set bail in additional categories of cases and adding new "dangerousness" considerations. By 2024, New York's bail system was significantly more punitive than the 2019 reform had envisioned, though still less reliant on cash bail than the pre-2019 system.

What did the data show during the backlash period? Studies by the Data Collaborative for Justice at John Jay College, the Center for Court Innovation, and the New York State Office of Court Administration generally found that bail reform was not associated with meaningful increases in overall crime, pretrial re-arrest rates, or failure to appear rates. A 2023 study in the journal Criminology found that the share of people rearrested on a violent felony while on pretrial release was statistically indistinguishable before and after reform.

The rollbacks happened anyway. This is worth sitting with. The evidence did not support the rollback narrative. The rollback happened because the narrative was politically potent, and because the beneficiaries of reform — overwhelmingly poor, disproportionately Black and Latino New Yorkers — did not have the political power to defend the reform against a coordinated tabloid-and-police-union campaign.

Why is the "bail reform causes crime" narrative so sticky, when the data consistently undercuts it?

A few reasons:

Availability heuristic. A single horrific case — a person released pretrial who commits a serious crime — is vivid, memorable, and available to the mind. The thousands of people released pretrial who commit no crimes are invisible statistics.

Counterfactual invisibility. When a person on pretrial release commits a crime, the crime is attributed to the release decision. But under the cash bail system, some of the same people would have been released (because they or someone in their family paid), and some of the same crimes would have happened. We don't count the counterfactual.

Selection bias in media coverage. Local tabloid crime coverage disproportionately features cases involving bail reform when they fit the narrative. Cases that don't fit the narrative — a person released pretrial who went home, went to work, and showed up to court — are not news.

Police union and district attorney interests. Law enforcement organizations often have institutional interests in maintaining high pretrial detention rates: it aids plea bargaining leverage, it reduces prosecutorial caseload, it produces political support from tough-on-crime constituencies. These institutions have significant media and political influence.

Racial coding. The image of the "dangerous criminal freed" is racialized in American public discourse. Research by the Sentencing Project and others has documented that media coverage of crime disproportionately features Black and Latino defendants, priming racial threat perceptions even when racially neutral policies are being debated.

Understanding the narrative problem matters because the empirical case for bail reform is, frankly, strong. The political case has to be built against a headwind of false stories.

The New Jersey reform of 2017 was the result of a coalition that included the Drug Policy Alliance, the ACLU of New Jersey, the Innocence Project, the NAACP, faith community organizations, formerly incarcerated leaders, and the families of people held on bail. The coalition worked for about five years — from initial organizing in 2012 to legislative passage in 2014 to implementation in 2017 — to build the political conditions for reform.

The work included: research and data analysis, state-by-state comparison studies, direct lobbying of legislators, media campaigns featuring the stories of people harmed by cash bail, legal advocacy in cases that illustrated the injustice of the system, and coalition-building across the left-right political spectrum (including conservative voices who objected to cash bail on liberty and cost grounds).

The New York reform of 2019 was built by a different coalition over a similar timeframe. VOCAL-NY, a grassroots organization led by people directly affected by the criminal legal system, was central. Other key players included Katal Center for Equity, Health, and Justice; Brooklyn Community Bail Fund; the Bronx Defenders; Black Lives Matter affiliated organizations; and a growing coalition of progressive prosecutors and legislators.

The common features of the coalitions that succeeded: they were led by people directly affected by the system; they paired policy advocacy with direct service (bail funds, court support, legal representation); they built media infrastructure to tell different stories than the tabloid narrative; and they played a long game, building toward a legislative moment over years of work.

The rollbacks in New York show what happens when the coalition cannot sustain the political defense of reform against a counter-campaign. This is not an argument against reform. It is an argument for building reforms deep enough and broad enough that they are politically resilient.

Exercise 1: Find out how your jurisdiction handles bail. How many people are held in your local jail pretrial? What is the average bail amount? Is there a bail fund in your area? This information is often available through county jail reports, court system annual reports, or the Vera Institute's Incarceration Trends project.

Exercise 2: Calculate the cost. If your local jail holds a thousand people pretrial, and the average daily cost per inmate is one hundred twenty dollars, the pretrial jail is costing taxpayers about forty-four million dollars per year. How much of that would be unnecessary under a reformed system?

Exercise 3: Read a pretrial services report. D.C.'s Pretrial Services Agency publishes an annual report. New Jersey's Administrative Office of the Courts does the same. Read one. Notice what the data actually shows, as opposed to what the news coverage of the data claims.

Exercise 4: Find the organizing. Who in your area is working on bail reform? Community Justice Exchange, the National Bail Fund Network, and local bail funds are good starting points. What are they asking for? What support do they need?

Exercise 5: Interrogate a news story. The next time you see a news story about a person released pretrial who committed a crime, ask: what would have happened to this person under the previous system? What is the base rate of such events? How is this story being framed, and why?

- D.C. Pretrial Services Agency. Annual Reports. 1990-2024. - New Jersey Administrative Office of the Courts. "Criminal Justice Reform Report." Annual publications, 2017-2024. - Center for Court Innovation. "Bail Reform and Public Safety in New York State." 2020, 2022. - Data Collaborative for Justice at John Jay College. Multiple reports on New York bail reform, 2020-2024. - Criminology and Public Policy. "Effects of Cash Bail Reform on Crime." 2023. - The Bail Project. Annual Impact Reports, 2018-2024. - Arnold Ventures. "Public Safety Assessment" methodology and evaluation reports. - Vera Institute of Justice. "Incarceration Trends" data project. - Sentencing Project. "Race and Punishment: Racial Perceptions of Crime and Support for Punitive Policies." 2014. - Prison Policy Initiative. "Mass Incarceration: The Whole Pie." Updated annually.

Four hundred thousand people are in jail tonight who have not been convicted of anything and who would be home if they had a bank account. This is a policy choice. It can be unmade.

The evidence from D.C., New Jersey, and to a lesser degree New York shows that eliminating or sharply reducing cash bail does not endanger the public. Appearance rates hold. Re-arrest rates hold. Jail populations drop by a third or more. Taxpayer costs drop. Families stay intact. People keep their jobs and apartments. The only loser is the commercial bail bond industry, which extracts billions of dollars per year from the poorest defendants in the country.

The premise of this book is that if every person said yes, the world we could live in would look different. Bail reform is one of the clearest tests of that premise. The saying-yes required here is not complicated. It is the recognition that the legal principle of innocent until proven guilty should mean something in practice. It is the willingness to distribute the cost of pretrial trust across the community rather than loading it onto the poorest defendants and their families.

If you are looking for a place to start, find your local bail fund. Give to it. Find the organizing coalition pushing for reform in your state. Join it. Read the actual data from jurisdictions that have reformed, so that you are not disarmed when someone repeats the tabloid story at the dinner table.

The reform happens at the scale of law. But the law changes when communities make it change.