The Case Against Cashless Bail Starts With the Victim, Not the Defendant

Here is a thought that should not be controversial: before a judge decides whether to release a man accused of a violent crime, he ought to be allowed to ask whether the man is violent. In some places, for some charges, he was not allowed. The statute listed what he might consider, and for the cases it covered the list did not include the safety of the woman who called 911.

That is not the caricature of a reform law. It is, in the plainest terms, how New York's bail statute worked for the first years it was in force. I want to look at that law, at Illinois's later model, at the best case for reform, and then at what I think the right answer is. The argument starts where all criminal law ought to start: with the person who was hurt.

Defining the terms

Pretrial detention means holding an accused person in jail while his case is pending, before any conviction. It is the most serious thing the state does to someone presumed innocent, and so it needs justification.

Money bail is a system in which a judge sets an amount the defendant must pay or post, directly or through a bondsman, to be released until trial.

Cashless bail is a loose label for laws that eliminate or sharply limit money bail. They differ widely. Some replace it with release on conditions and give judges a way to hold the dangerous after a hearing. Others restrict what judges may consider.

Preventive detention is detention based in part on predicted danger, not only on the risk that the defendant will not return to court.

The whole argument turns on one question: what is pretrial detention for? Two answers are possible. One is that it exists only to make sure the accused shows up for trial. The other is that it also exists to protect the public in the meantime. I will argue that the second answer is the right one, that the Founding tradition and the Supreme Court agree, and that laws built on the first answer get the purpose backwards. I trace that history in Excessive Bail Shall Not Be Required, and What the Founders Meant by It; the short version is that the Constitution does not require money bail, and also does not forbid a judge from holding a dangerous defendant after a hearing.

New York, 2019

In April 2019 New York's legislature passed, as part of its budget, a package of criminal justice changes that took effect on January 1, 2020. The bail provisions did several things. They ended money bail and pretrial detention for most misdemeanors and many nonviolent felonies, and they required release, with or without conditions, in those cases.

The feature of the law that stood out was a rule about what a judge could weigh. In most states and under federal law, a judge setting pretrial conditions looks at two things: the risk that the defendant will flee and the risk that he will endanger others. New York's statute, for the cases it covered, directed the judge to consider only the risk of flight, and to choose only the control or restriction needed to reasonably assure return to court. The danger the defendant posed was not a statutory factor. A judge who thought a defendant dangerous had few tools to act on it, unless the charge was on a list of qualifying offenses for which bail remained available.

That is the point I want to stress, because it is specific and checkable. The objection is not that New York ended cash bail. The objection is that it narrowed the question a judge may ask to something that has nothing to do with the thing the public most wants him to ask.

The legislature then amended the law. In 2020 it made changes that kept the core of the 2019 law, and in 2022 it made more gun crimes bail-eligible and gave judges more discretion to consider factors other than flight risk. The state budget changes signed in 2023 amended the law again. Today the statute directs the court to weigh, among other things, a defendant's record, firearm history and whether the charge caused serious harm. Read the history of the amendments as what it is: the legislature, repeatedly, widening what judges may weigh. When the lawmakers who wrote a law keep revising the same provision, it is fair to ask whether the original design left something out.

Let me be fair about what I am not claiming. I am not telling you that the 2019 law produced a specific increase in crime. That is an empirical question that depends on data I will not quote from memory, and researchers disagree about how to separate the law's effects from everything else that was happening in 2020 and 2021. Read the competing studies and the state's own reports before you hold an opinion on that. My claim is about the design: a law that tells a judge to ignore danger is badly built whatever the data show, and the amendments show the legislature agreed.

Illinois, 2023

Illinois went further, with a different design. The state's Pretrial Fairness Act, part of the 2021 criminal justice package known as the SAFE-T Act, abolished money bail altogether. After litigation, in which the Illinois Supreme Court upheld the law in July 2023 by a 5-2 vote, it took effect on September 18, 2023. Illinois is widely described as the first state to end money bail entirely.

The Illinois model differs from New York's in a way that matters. Defendants are presumed to be released. But the prosecution may petition to detain a defendant charged with a listed, detainable offense, and the judge holds a hearing. To order detention the court must find, by clear and convincing evidence, that the defendant committed a detainable offense, that he poses a real and present threat to the safety of a person or the community, or a high likelihood of willful flight, and that no condition or combination of conditions can mitigate that threat. The list of detainable offenses includes forcible felonies and other serious crimes.

Note what that is. It is Salerno's model. It is the federal model in 18 U.S.C. 3142, which I discussed in the bail history piece. It is a hearing, with counsel, with a burden on the state, with a written finding. That is the kind of detention power the Supreme Court approved in United States v. Salerno in 1987.

So Illinois should not be lumped together with New York as if they were the same policy. The Illinois law keeps a judge's ability to consider danger and routes it through a hearing. The criticisms I would make of it are different: that the list of detainable offenses may leave out conduct that deserves a hearing, that the standard of proof is demanding, and that it moves quickly, so the state must be ready. Those are fights about the details of a defensible model. I raise them because intellectual honesty requires distinguishing a bad design from a debatable one. The Illinois design is in the second category.

The best case for reform

Now the steelman, in the words its defenders would accept.

Money bail, the reformers say, sorts defendants by wealth, not by risk. A person charged with a low-level offense who cannot post a few hundred dollars sits in jail for days or weeks. In that time he may lose his job, miss rent, lose custody arrangements, and then plead guilty to something he might have beaten, because the plea gets him home. A defendant with money and the same charge goes free. The system detains the poor for being poor, which no one defending equal justice can accept. And the commercial bail bond industry makes a profit from that arrangement, taking a nonrefundable fee from the families of people who have not been convicted.

There is real force in all of this, and conservatives should concede it. The Founding tradition I described does not endorse a price list. Stack v. Boyle says that bail must be fitted to the individual and the purpose. A system that detains a person for want of a small sum, though he poses no danger and no flight risk, offends the principle that liberty is the norm. The people who made this argument were right about the facts of many local systems, and some of the most honest work on bail reform has come from people on the right who believe in limited government and in not locking up people the government has no good reason to hold.

Why the reform argument does not reach the conclusion it wants

Here is where I part company, in numbered form.

One: the reformers' own premise is about fitting the detention to the risk. Their strongest point is that detention should depend on risk, not wealth. Follow that and it cuts both ways. If a defendant of modest means who is no threat should go home, then a defendant of any means who is a threat should not. A rule that says "release everyone, whatever the danger" does not fix the arbitrariness of money bail; it exchanges one arbitrary rule for another.

Two: a statute that forbids the judge to weigh danger removes the very question the law exists to answer. Pretrial detention is justified by the state's interest in protecting people from harm and in securing the trial. The Supreme Court called the interest in preventing crime by arrestees "legitimate and compelling." If the legislature takes one of those two interests off the table, the judge cannot do his job.

Three: the harm of a wrong release falls on someone who was not at the hearing. The defendant has a lawyer, a hearing, a presumption of innocence, and a judge. The potential victim, a spouse, a witness, a neighbor, has none of those. When an error is made on the side of release, the cost is paid by a person who had no seat at the table. When an error is made on the side of detention, the defendant's lawyer can ask for review, and the system has a remedy. The asymmetry matters. It is the reason that crime victims' rights, a theme in The Victims' Rights Movement Began With a 1982 Presidential Task Force, matter in this debate.

Four: witnesses and victims in cases that depend on testimony are the people most exposed by a release they cannot contest. A defendant in a domestic violence or gang case who is released the same day may be in a position to intimidate the one person whose testimony will convict him. A statute with no way to consider that has given up part of the protection that makes testifying possible. Without witnesses, the justice system has no cases.

Five: the fairness problem has a narrower fix. The problem the reformers identified was that people who are neither dangerous nor likely to flee are held because they cannot pay. The right remedy is to presume release for those defendants, and to let the court detain those who are dangerous or likely to flee, after a hearing. That solves the injustice without the cost.

What the data can and cannot settle

A fair post has to say what the numbers allow. Researchers have studied bail reform in several jurisdictions, and the findings are mixed: some find little effect on measured crime and on failure to appear, and others find specific problems, such as a rise in rearrests of released defendants in some categories. I will not cite a particular finding from memory, because the studies differ in design, the places differ, and the period overlapped with the pandemic and with other large changes in public safety. My advice is the same advice I give in Seven Things Crime Statistics Cannot Tell You, and Two Things They Can: do not accept a one-year change as proof, in either direction.

What the data cannot settle is the point that does not depend on data at all. Whether a law lets the judge ask about danger is a question of design, and it has an answer on principle. A system built to protect the public cannot be indifferent to the public's safety.

A better answer

If I were writing the pretrial law of a state, I would build it this way.

1. A presumption of release for low-level, nonviolent offenses, with the least restrictive conditions needed to secure appearance, so that no one sits in a cell because he is poor.
2. A prompt detention hearing for defined serious offenses, with the state bearing the burden, a clear and convincing standard, defense counsel, and a written finding of the reasons.
3. A judge who may weigh both flight and danger, including a defendant's record, the use of a weapon, and the safety of the victim and witnesses.
4. Support for appearance, such as reminders and transportation help, which some studies suggest reduce missed court dates and which cost far less than a jail bed.
5. Public reporting, by the court system, of how many defendants were released, how many were detained, how many failed to appear and how many were rearrested, so the people can judge the policy by results.

That structure takes the reformers' valid point seriously and the public's valid point seriously, and it fits the Founding tradition of a right to release with a power to hold. It also puts accountability on the person who bears it, an elected or appointed judge, whose decisions are recorded and reviewable. For the comparison of models, see Three Ways to Decide Who Waits for Trial in Jail, Side by Side.

The moral center

Put the statistics down for a moment. Somewhere in your state this week a person who was threatened will go to a courthouse and ask whether the man who threatened her is coming home tonight. The answer ought to depend on what he did, what he has done before, and what a neutral judge, after hearing from both sides, believes he will do next. It should not depend on a rule that took the question away. The Founders trusted accountable judges with that decision. We should too, and we should hold them to account when they get it wrong.

The case in brief

1. Pretrial detention serves two purposes, assuring appearance and protecting the public, and a law that removes the second gets the purpose backwards.
2. New York's 2019 law, effective January 2020, directed judges in many cases to weigh only flight risk, and the legislature kept amending it afterward to widen judges' discretion.
3. Illinois's Pretrial Fairness Act, effective September 2023, ended money bail but kept a detention hearing with a clear and convincing standard, which is closer to the model the Supreme Court approved in Salerno.
4. The reformers are right that poor defendants should not sit in jail for want of a few hundred dollars, but the cure is individualized risk assessment, not a ban on asking about danger.
5. The best design presumes release for low-level cases, holds a prompt hearing for serious ones, lets the judge weigh danger and flight, and reports the results publicly.

What you can do: look up your state's pretrial release statute on your legislature's website, check whether it lets a judge consider danger, and write your state legislator a short letter about what you found.