In New York City, an arraignment can be over in minutes. A person who has just been arrested—often held overnight in a cell beneath the courthouse—stands before a judge for the first time. An assistant district attorney summarizes the charges. A defense attorney, usually a public defender who may have met their client only moments earlier, argues for release. The judge reviews the file, listens briefly, and decides: release or detention.
If detention, the judge may set cash bail at whatever amount seems right, because New York law imposes no cap. There is no jury, no extended deliberation, no sentencing grid. Just one judge making a call. The consequences for this detention can be drastic: those held even briefly are more likely to lose housing, to have their children placed in foster care, and to plead to charges they might otherwise contest.
The scale of pretrial detention in New York right now makes these dynamics especially urgent. In early 2025, the Rikers Island population climbed above 7,000, the highest total since 2019 and nearly double the target needed for the city’s long-planned closure of the facility. About 84 percent of those detained are being held pretrial, meaning they have not been convicted and are legally presumed innocent.
In a research study published in the Journal of Law and Courts, my collaborator Oded Oren and I asked: Does a judge’s professional background shape these decisions, and if so, how? We analyzed nearly 70,000 New York City arraignment cases. We found a clear and substantial pattern. Judges who previously worked in law enforcement—as state or federal prosecutors or police officers—are about 4 percent more likely to order pretrial detention than judges with no prior experience in either law enforcement or legal services. When they set cash bail, the amounts run roughly 32 percent higher on average. These differences persist even after controlling for charge severity, criminal history, and a host of other factors.
More from our decarceral brainstorm
Inquest—finalist for the 2025 National Magazine Award for General Excellence & cited in The Best American Essays 2025—brings you insights from the people working to create a world without mass incarceration.
Sign up for our newsletter to get the latest in your inbox every Saturday.
Newsletter
To make the stakes concrete, we estimated what it would mean to replace a single former law enforcement judge with one who had no background in either law enforcement or legal services. Over a typical ten-year judicial term in New York City, that one substitution would mean roughly 65 fewer people detained, about 17 fewer person-years spent in jail, approximately $8.7 million less in detention costs for city taxpayers, and about $6 million less in cash bail set for defendants and their families.
And because New York City’s rotating calendar assigns cases to judges fairly randomly, this is not a story about certain cases finding certain judges. It is a story about the judges themselves.
In New York City, criminal court judges are appointed by the mayor from a pool of candidates vetted through the local political establishment. Prosecutors have long held enormous advantages in this process—they are well connected within party structures, seen as experienced in criminal law, and benefit from a culture in which being perceived as “tough on crime” is an asset and the reverse is a liability. Public defenders have long faced skepticism and even stigma when seeking judicial appointments; their advocacy on behalf of people accused of crimes is sometimes treated as a disqualification rather than a credential. The result is a lopsided bench.
Among the judges in our study, 84 had law enforcement backgrounds, while just 41 had worked in legal services. Nationally, the imbalance is well documented at the state supreme court level: about 40 percent of sitting justices have prosecutorial backgrounds, while only 10 percent are former public defenders. Trial courts—where bail decisions actually happen—draw from the same pipeline. And because state courts handle nearly all of U.S. criminal cases—including 94 percent of felony convictions—that imbalance shapes pretrial outcomes far beyond New York.
Our study also reveals a striking asymmetry: while judges with law enforcement backgrounds tend to impose harsher outcomes, those from legal services do not consistently produce more lenient ones. Why didn’t our data show a clear mirror image?
One likely reason is that the judicial selection process itself moderates defense-oriented candidates—a former public defender may face pressure to demonstrate they are not “soft on crime,” a pressure that has no equivalent for former prosecutors. Another is that our legal services category necessarily grouped criminal public defenders together with attorneys from housing, family, and other civil settings, potentially diluting a signal that might exist among criminal defenders specifically. Either way, the absence of a legal services effect does not weaken our main finding. The law enforcement effect is robust on its own.
That raises a further question: How does this effect operate? Two mechanisms could explain it. One possibility is socialization. Norms and incentives within law enforcement institutions gradually shape attitudes over time. Another is selection. People already inclined toward severity enter law enforcement in the first place, and judicial selectors then elevate them to the bench. Our data lean toward selection, though the picture is not entirely clean.
Longer law enforcement tenure did not produce a consistent increase in severity—a pattern one would expect if socialization were the main driver. Judges with more than twenty years of law enforcement experience showed modestly harsher outcomes on some measures, but the pattern was uneven; they were not consistently harsher than those with only five years of experience. This suggests that whether a judge served in law enforcement at all matters more than how long they served.
The law enforcement effect doesn’t show up evenly across different types of cases, and understanding where it concentrates helps sharpen the picture.
New York’s 2020 bail reform law made release mandatory for nearly all misdemeanors and nonviolent felonies, with exceptions that preserve some judicial discretion. For most lower-level charges, background-linked differences in detention and bail-setting largely vanish—though even here, when bail is set, former law enforcement judges impose higher amounts.
The pattern becomes more pronounced in violent felony charges, where judges retain full discretion to release, set bail, or remand. In those cases, former prosecutors and police officers are significantly more likely to detain and to set higher bail amounts. Discretion, in other words, is the opening through which a judge’s old job enters the courtroom.
A second dynamic likely reinforces this pattern in violent felony cases: reputational risk. Judges know that a released defendant who later commits a violent crime can become a headline, a career liability, a political weapon. The incentives are asymmetric—no one writes a story about the defendant who was released and went on to live a regular life. Former law enforcement judges, shaped by institutions that prize public safety above other values, may be particularly attuned to this calculus. Whatever their private inclinations, the institutional pressure tilts in one direction: when in doubt, detain.
These findings should reframe two live debates in New York.
The first is judicial selection. Earlier this year, Mayor Zohran Mamdani moved to overhaul how the city chooses its judges, appointing civil rights attorney Ali Najmi to lead the Mayor’s Advisory Committee on the Judiciary and issuing an executive order directing the committee to actively recruit public defenders and legal services lawyers—professionals who have historically been passed over. The Bronx Defenders praised the initiative. Notably, Mamdani’s first three judicial appointments were still former prosecutors drawn from a pool vetted under the previous mayor—a reminder that changing a pipeline takes time even when the political will exists.
Our research suggests that his efforts are a needed corrective. The professional composition of the bench is not a neutral administrative detail; it is a policy choice with measurable effects on detention, bail, and the communities that bear those costs. Diversifying the bench does not require new legislation, but it does require changing who gets considered. Movements working toward decarceration should treat judicial selection as a frontline issue, not an insider concern for bar associations and political operatives.
The second debate concerns detention authority. This legislative session, law enforcement groups in New York pushed the PROTECT Act, which would have allowed judges to detain people they deem likely to commit future violent crimes—a “dangerousness” standard that New York’s bail reform explicitly rejected. The bill’s supporters framed it as a commonsense public safety measure. But our findings offer a sharper reading. Expanding discretion on a bench already dominated by former prosecutors and police officers is not a neutral policy adjustment. It is, in effect, a detention expansion.
Although the bill did not pass, efforts to advance its core provisions are likely to continue. New York’s bail reform substantially limited the role of judicial discretion in pretrial detention for lower-level charges. The PROTECT Act would have reopened that door. It would have expanded judicial discretion in three ways at once. First, it would have introduced “dangerousness”—a prediction of future violence—alongside flight risk as a basis for detention. Second, it would have repealed the existing list of qualifying offenses, opening that prediction to nearly any charge. Third, it would have required courts to consult an algorithmic “risk and needs assessment,” though only in an advisory capacity, not as a binding constraint. The judge would retain the final word—exactly the point at which our findings indicate professional background bites hardest.
Our findings do not mean that any individual former prosecutor is incapable of fairness, or that background is destiny. But systems are defined by patterns, not exceptions. And the pattern here is that a bench disproportionately drawn from law enforcement produces disproportionately harsh pretrial outcomes—outcomes that fall on defendants and communities with no say in who gets appointed. Our study was only made possible because New York’s bail reform law mandated the publication of detailed arraignment records. Most jurisdictions offer nothing comparable, leaving similar patterns largely invisible.
Judicial appointments are jail policy. Until we treat them that way—until we ask who sits on the bench, how they got there, and why—we will keep reproducing punishment, decision after decision after decision.
Image source: ALEJANDRO POHLENZ / Unsplash
