Since New York State’s 2019 bail reform went into effect, controversy has swirled around the question of its impact on public safety—as well as its broader success in creating a more just and equitable system. The Covid-19 pandemic (which hit three months after the bail reform’s effective date), the upheaval following the killing of George Floyd, and the subsequent enactment of various police and criminal justice reforms are confounding factors that make assessing the specific effects of the 2019 bail reform particularly complex.
This paper attempts to give the public a better sense of the risks of this policy shift and the detrimental effect that the changes have had on public safety. First, I will lay out the content of the bail reform and will measure pertinent impacts on crime and re-offending rates. Then I will review changes made in the 2020 and 2022 amendments. I will look at the push for supervised release and closing Rikers Island and how those initiatives fed into the momentum behind these laws. Finally, I will propose recommendations to improve bail reform’s impact on public safety, which include:
- Allow judges to set bail, remand, release on recognizance (ROR), or conditions of release for any crime and any defendant. There should be a presumption of release for misdemeanors and nonviolent felonies, which could be rebutted by the defendant’s prior record or other factors that indicate that the defendant is a flight risk. There should be a presumption of bail, remand, or nonmonetary conditions for defendants charged with violent felonies or weapons offenses. This presumption could also be rebutted by evidence of the defendant’s roots in the community, lack of criminal record, and similar factors.
- Allow judges to remand defendants without bail if they are deemed a danger to the public or have a high probability of re-offending. This is an option available to judges in every state but New York. Allowing judges to remand defendants deemed to be a danger to society also removes the economic disparity between people for whom bail is set.
- Allow judges to change the conditions of release based on violations of the original release conditions or failures to comply with the judge’s order. Judges should be allowed to do this on any crime, not just qualifying offenses.
- Require the New York State Office of Court Administration (OCA) to collect additional data, including rearrests of defendants between plea and sentence and rearrests post-sentence, and to provide a full accounting of the total number of rearrests, not just the number of defendants rearrested.
In addition to these proposals, New York City must reconsider the latest plan to close Rikers Island and to build four community jails with a totally inadequate maximum of 3,300 jail beds. Activists and lawmakers based those plans on the assumptions of perpetual reductions in crime and a safe reduction in jail populations. In fact, those assumptions are clearly mutually exclusive.
For 27 years, from 1993 to early 2020, under the “old” bail laws and the “broken” criminal justice system, index crime in New York City steadily declined by nearly 76%. In just two years of the new bail laws and other progressive reforms, index crimes in New York City rose 36.6%. There are many reasons for the rise in crime, but as the analysis below will demonstrate, it is not coincidental that the sudden, massive increase in city crime came at precisely the same time as the release of 2,000 career criminals from city jails.
New York State’s new bail laws, enacted in 2019 and made effective at the beginning of 2020, were billed as a means to end the “mass incarceration” of the poor and minorities who were unable to post even small amounts of bail. Advocates justified the changes (outlined below) with anecdotal incidents of perceived injustices in the criminal justice system, coupled with dubious statistical analyses. In New York City in particular, the movement to end so-called mass incarceration coincided with and propelled the ancillary movement to close Rikers Island jail and to build new, community-based “justice centers” in each NYC borough to house the inmates then held on Rikers and other city jails.
Amendments to the law in 2020 and 2022 were responses to strong public disaffection with the law and the citywide increases in crime that accompanied the effective date of the 2019 reform. After almost three decades of declining crime in NYC, lasting from about 1993 through early 2020, the staggering increases in crime commenced almost precisely with the effective date of these reforms, which resulted in more than 2,000 inmates charged with nonviolent felonies and misdemeanors being released from city jails alone by January 1, 2020—the date the law went into effect.
New York City crime data show that index crimes went up 20% in the first two and a half months after the bail reform took effect, increasing by double digits in the crimes for which judges could no longer set bail: burglary +26%; car theft +68%; grand larceny +16%; and petit larceny +19%. These numbers coincided with the release of every defendant held solely on bail charged with burglary, car theft, grand larceny, and petit larceny from city jails as a result of the new law.
An analysis of Office of Court Administration (OCA) data shows that 69.7% of city defendants arraigned on felony charges between July 1, 2020, and August 31, 2020, had a prior conviction or a pending case, and 30% of those defendants got rearrested while their case was pending. If they had a prior conviction or pending case and the judge found them to be a risk of flight, 45% of them were rearrested while their case was pending. The rearrest rates for this latter group run as high as 70% for commercial burglary, robbery third degree, and grand larceny, and 79% for petit larceny—all non-bailable offenses.
Ironically, given the stated aims of bail reform advocates, the overwhelming majority of the victims of crime in NYC are black and Hispanic, and the recent crime wave is having a disproportionate effect on their neighborhoods. In 2021, blacks and Hispanics, who made up 48.5% of the city’s population, accounted for 90.7% of murder victims in the city, 96.9% of shooting victims, 73.2% of rape victims, 71.5% of robbery victims, 79.8% of felony assault victims, and 52.8% of grand larceny victims. They even constituted a disproportionate percentage of misdemeanor victims: 68.3% of criminal mischief victims, 57.9% of petit larceny victims, and 75.5% of misdemeanor assault victims.
The New York State legislature must change these laws—that is, reform the reform—by expanding judicial discretion with respect to release decisions, allowing them to impose conditions such as bail on any defendant’s release, irrespective of the charge, within certain broad guidelines; and, perhaps most important, by allowing judges to consider public safety and the defendant’s risk of re-offending when making release conditions.
The Content of the 2019 Law
On April 1, 2019, the New York State legislature passed the first version of bail reform (hereafter referred to as the 2019 law), which was signed into law as part of the state’s massive budget bill. This was a sea change in bail legislation, as it had existed for decades. Essentially, the new law set forth a limited number of criminal charges for which a judge could set bail, referring to these crimes as “qualifying offenses.” These consisted mostly of violent felony offenses, such as rape, first-degree robbery and burglary, kidnapping, sexual assault, armed robbery, and criminal possession of a weapon as a felony. Specifically excluded from the list of qualifying offenses were residential burglaries and cases where the defendant was charged with unarmed robbery while being aided by another, as well as almost all felony drug cases.
The 2019 law also prohibited judges from setting bail on almost any nonviolent felony offense. Crimes such as grand larceny, commercial burglary, perjury, criminal contempt, bail jumping, escape, setting fire to commercial properties or cars, stalking, selling or possessing even large amounts of narcotics on the street, manslaughter, criminally negligent homicide, and more than 350 other crimes were all deemed “non-bailable,” meaning that a judge could not set bail regardless of the defendant’s prior criminal record. The law also prohibited a judge from setting bail on almost any misdemeanor except second-degree criminal contempt (violating a court order) but only in domestic violence cases.
Even if the defendant were charged with a qualifying offense, the judge was required to impose the “least restrictive kind and degree of control or restriction that is necessary to secure” the defendant’s return to court when required. Alternatives to bail include releasing the defendant on his own recognizance, setting nonmonetary conditions for his release, electronic monitoring of the defendant, and remanding him without bail. Previously, judges were allowed to set bail on any crime—and could consider the defendant’s character, reputation, habits, mental condition, family ties, length of residence in the community (if any), and employment and financial resources, as well as the weight of the evidence against him, the probability of conviction, and the possible sentence. Under the 2019 law, a judge could no longer consider many of those factors. Any consideration of a defendant’s financial condition was limited to setting a bail that the defendant could make.
The court could not, under either the pre-2019 bail reform or the revisions enacted in 2020 or 2022, consider the defendant’s risk of re-offending or public safety when determining release conditions. Before the 2019 reforms, judges had enough leeway in what they could consider in setting bail to fashion a reasonable argument that a problematic defendant was a flight risk and should be held on bail—indirectly protecting the community, in the process. Defendants who had lengthy records could have bail set even if the crime they were charged with was a nonviolent felony (car theft, drugs, grand larceny) or a misdemeanor (petit larceny, assault 3, criminal contempt). Judges at arraignment would routinely review a defendant’s record and arrest charges and set bail to keep a defendant incarcerated, on the theory that the defendant put his own interests above that of society and would therefore likely not appear in court. This prevented the defendant from re-offending or harming others. Now, those are all non-bailable offenses, and defendants charged with misdemeanors and nonviolent felonies are routinely released with no bail under the new laws, regardless of past crimes.
The 2019 law also required police officers to issue Desk Appearance Tickets to defendants charged with misdemeanors or Class E felonies. These tickets are essentially summonses requiring the defendant to appear in court at a later date. There were certain exceptions to this rule—e.g., if the defendant had an open warrant, the complainant needed an order of protection, or the defendant was in such condition as to require medical or mental health care.
Judges and Jails Face Early Fallout from the Change
After the law passed, court administrators realized that on New Year’s Day 2020, when the law was scheduled to take effect, all the defendants then being held solely on bail on these non-qualifying offenses would have to be released. Experts estimated that 40% of the approximately 5,000 inmates held pretrial in city jails on bail alone, and thousands of others being held in local jails around the state, would be released because of the new law. To avoid the crush of bail reviews under the new law, the state OCA instructed judges in November 2019 to begin reevaluating the bail status of all defendants then held on bail, releasing those who were held on bail for whom bail could not be set under the new law. The judges were also told they could, at their discretion, begin implementing the law immediately, prior to its effective date. Many of them had already begun doing so in October, 2019.
On June 6, 2019, the Queens County District Attorney’s Office conducted an analysis of the county defendants awaiting trial in NYC jails who were being held solely because they could not make bail. (Excluded from this analysis were defendants being held on other non-bail holds, such as remand, out-of-state warrants, or parole and probation violations, or those defendants serving a city sentence.) There were 398 defendants held solely because they could not post bail. Table 1 shows their criminal background.
Criminal Case History of Queens County “Bailable” Defendants in Custody, June 6, 2019
|A felonies include: Murder, Kidnapping 1, Attempted Murder of a Police Officer, Arson 1, Possession of 8 ounces or sale of more than 4 ounces of narcotics|
|B felonies include: Attempted Murder, Assault 1, Burglary 1, Robbery 1, Rape 1, Kidnapping 2, Arson 2, Grand Larceny 1, Possession and Sale of Controlled Substances|
|C felonies include: Attempt to commit a B felony, Burglary 2, Robbery 2, Rape 2, Arson 3, Criminal Possession of a Weapon 2, Grand Larceny 2|
|D felonies include: Attempt to commit a C felony, Robbery 3, Burglary 3, Criminal Possession of a Weapon 3, Assault 2, Arson 4, Grand Larceny 3, Perjury, Criminal Contempt, Possession of Controlled Substances|
|E felonies include: Attempt to commit a D felony, Grand Larceny 4|
|A Misdemeanors include: Petit Larceny, Assault 3|
|B Misdemeanors include: Attempt to commit a Class A Misdemeanor|
Of all the defendants, 95% were being held on felonies, and 41% were being held on violent A and B felonies. Defendants being held on nonviolent D and E felonies generally had the worst criminal records, with an average of 13 and 16 prior arrests, respectively, and an average of 7 and 9 felony arrests. Only 20 misdemeanor defendants were being held solely on bail, and the defendants charged with Class A misdemeanors averaged 12 prior arrests, 5 of them for felonies. This was clearly a troubled lot.
A separate analysis done by the Queens DA’s Office on May 29, 2019, found that 363 defendants would have to be mandatorily released under the 2019 bail reform because bail could no longer be set on the offenses that they were charged with, and there were no other holds on them. Table 2 reports the criminal history of those 363 defendants.
Criminal Case History of Queens County Defendants Who Would Qualify for Mandatory Release, May 29, 2019
The day the 2019 bail law was passed, NYC jails held 7,822 inmates, about 5,209 of them being held pretrial. On January 1, 2020, there were 5,809 inmates—a reduction of more than 2,000 inmates (approximately 40% of the pretrial detainees) in just eight months, most occurring in the last few months of 2019. The predictions were correct. The legislature and advocates knew exactly what they were doing. The number of defendants being held on misdemeanors and nonviolent felonies fell dramatically, as shown in Table 3.
New York City Jail Population
By January 15, 2020, any defendants who were being held in city custody awaiting trial solely because they could not make bail on the following charges had been released:
- Burglary 2 or 3 (home or commercial burglary) (C or D felony)
- Robbery 2 (aided by another) (C felony)
- Grand larceny (D felony)
- Car theft (D or E felony)
- Drug possession or sale (A, B, C, D, or E felony)
- Petit larceny (A misdemeanor)
- Misdemeanor assault (A misdemeanor)
By March 15, 2020, just two and a half months into the 2019 law going into effect, and before Covid-19 affected the criminal justice system, crime in NYC had begun its rise.
Compared with the same period year-to-date in 2019, when the old bail laws were in effect, overall crime in NYC went up 20.05%. It went up in the following crime categories, shown in Table 4.
Percent Increase in Crime, Comparing YTD March 15, 2019, and YTD March 15, 2020
The only crimes to show decreases were murder (–3.2%) and rape (–11.5%), crimes for which judges could still set bail.
It should be noted that overall crime in NYC went down 1.1% in calendar year 2019, compared with 2018, with burglary down 8.9%, grand larceny down 1.4%, car theft down 0.2%, and misdemeanor assault down 0.8%. In fact, Table 5 shows the slowing decline in the rate of crime reduction throughout 2019, going from a 7.3% reduction YTD in March 2019 over 2018, to a much smaller 1.1% reduction YTD in December 2019. These numbers coincide with the enactment of the 2019 bail reform and the resultant decline in city jail populations (Table 6).
NYC Overall Monthly Crime Rate Change in 2019 over 2018
NYC Jail Population, 2019
The change in crime patterns from 2018 to 2019 was most noticeable in car theft, which declined 12.5% in April 2019, compared with the same period in 2018, yet declined just 0.2% by December 2019, compared with the same period the year before. Burglary went from a 14.5% reduction in April 2019 to an 8.9% reduction by December 2019. Grand larceny went from a 5.5% decline in April 2019 to a 1.4% decline by December 2019. And petit larceny, which declined 0.6% in April 2019, compared with the same period in 2018, actually increased 3.2% in December 2019, compared with the same period the year before. All these crimes became non-bailable under the 2019 reforms.
In mid-March 2020, Mayor Bill de Blasio began to close various portions of the city—schools, bars, restaurants, businesses, cultural facilities—and subway service became limited. By June 28, 2020, major crime overall in NYC actually decreased 2.83%, fueled almost entirely by a reduction in grand larcenies, the most numerous felony index crime, which fell 20.3%. This was largely the result of people not leaving their houses for work or shopping, thus being unavailable as crime victims. Petit larceny, mostly consisting of shopliftings, went down 7.2%—again, fueled mostly by store closings in the early stages of the pandemic. But the rise in other crimes continued: murder rose 23.1% YTD over 2019; burglary rose 45.8%; car theft rose 60.8%; and shootings rose 43.9%. These are percentage increases in these crimes not seen in decades, following nearly three decades of crime reductions in NYC.
Figure 1 and Table 7 show trends in the number of index offenses from January 1 through approximately the first two and a half months of the year, for each year 2017 through 2022. (For the data on non-index offenses, see the Appendix of this report.) To depict all the offenses on the same scale, 2019 is treated as the baseline, and offenses are reported as the percentage difference between each year and 2019. Thus, all the offenses appear as 0% for 2019.
Percent Change in Index Offenses, YTD Through Mid-March, Relative to 2019
Index Crime YTD Through Mid-March, 2017–22
The increases between 2019 and 2020 clearly show the impacts of the release of more than 2,000 inmates from city jails. After decades of constant, methodical decline in almost every crime category, crime in all the categories for which inmates had been released due to the new bail laws went up. In just the first two and a half months of 2020, according to NYPD CompStat reports, more than 3,000 more crimes had been committed, compared with the same period just a year earlier. That included 755 more robberies, 351 more felony assaults, 536 more burglaries, 1,221 more grand larcenies, and 532 more car thefts. The only thing that temporarily stopped the rise was the Covid pandemic—and, even then, victimization continued to increase when adjusted for the amount of time people spent, for example, in public or on the subways. When the economy opened up more fully in 2022, crime by more traditional measures resumed its rise, climbing from 16,707 index crimes in the first two and a half months of 2019 to 22,824 during the same period in 2022, a combined rise of 36.6% in index crimes.
Reforming the Reform: 2020 Bail Law Amendments
Partially in response to these unprecedented crime increases, as well as the accompanying press coverage and public outcry, the NYS legislature passed additional reforms in 2020 (hereafter referred to as the 2020 law). These amendments were made effective on July 2, 2020. They made several changes but kept the main body of the 2019 law in place.
Qualifying offenses were expanded to include residential burglary, escape, bail jumping, witness intimidation and witness tampering, sex trafficking, assault or arson as a hate crime, conspiracy to commit murder, money laundering in support of terrorism, Class A-I felony drug cases (sale of two ounces or more or possession of eight ounces or more of a narcotic drug), and crimes that result in the death of another person. When you consider that the legislature, in the 2020 amendments, had to add escape and bail jumping to the list of crimes for which a judge can set bail, you begin to comprehend the lack of thought that went into the 2019 law in the first place. Other than the homicide, burglary, and felony drug charges, these additional qualifying offenses represent a small number of the defendants charged in New York State.
The legislature also created new qualifying offenses to cover cases where the defendant commits a felony while on probation or parole, or where the defendant is arrested for a felony and qualifies for sentencing as a persistent felony offender. They also allowed judges to set bail on defendants charged with any felony or Class A misdemeanor “involving harm to an identifiable person or property,” where the charge arose while the defendant was released on his own recognizance or under conditions on a separate felony or a comparable Class A misdemeanor.
However, for the judge to set bail, the court must find, at the defendant’s arraignment, “reasonable cause” to believe that the defendant committed both the new crime and the one for which the defendant had been released. The mere fact that the defendant had a case pending when he was arrested for the new crime is not enough. Nor did the statute define exactly what “involving harm to an identifiable person or property” means. (This was later “clarified” in the 2022 amendments.)
Reforming the Reform of the Reform: 2022 Amendments
New York governor Kathy Hochul proposed several further amendments earlier this year—most notably, suggesting that judges should be allowed to consider public safety and the risk of re-offending when setting bail on almost any crime committed while armed with a firearm. Hochul’s proposal did lack a certain internal logic in that it allowed judges to consider public safety when someone attempted to murder another by shooting but not if the defendant used, say, a knife or subway train. It seemed more of a response to daily press reports about the gun violence plaguing New York State, and the city in particular, than any well-defined strategy to prevent or reduce crime, or seriously change the bail laws.
On April 7, 2022, the state legislature—again, as part of the annual budget bill—passed new changes to the bail law. These amendments did little to change the court’s ability to set bail on most penal law charges. The legislature added criminal possession of a firearm as a new qualifying offense but only if the defendant committed that crime while out on recognizance or released under nonmonetary release (NMR) conditions for a separate felony or misdemeanor. A check of the OCA database indicates that there were just 199 defendants arraigned on this new qualifying offense between July 1, 2020, and June 30, 2021, in NYC, and that only 46 of them had any pending case when they were arraigned. Thus, under the 2022 amended law, only 46 of them would have been eligible to have bail set.
The legislature also added two more firearms offenses to the list of qualifying offenses: criminal possession of a defaced firearm and criminal sale of a firearm to a minor. The OCA database shows that there were just 275 cases of possession of a firearm under the first offense, and no cases of sale to a minor since July 1, 2020, in NYC, out of the almost 91,000 cases arraigned. The charge of criminal possession of a defaced firearm is contained in PL Section 265.02, which actually contains 10 subdivisions. Only one, subdivision 3, is now a qualifying offense under the 2022 reforms. OCA data do not break down the charge by subdivision, so it is impossible to tell how many of the 275 arrests would be qualifying offenses.
The legislature added certain language that sounds tough. For example, it added to the list of factors that a judge may consider when deciding whether to set bail. The judge may now consider “whether the charge is alleged to have caused serious harm to an individual or group of individuals,” as well as the defendant’s violation of any order of protection issued by the court, not just family orders. Even with these additions, the judge still may set bail only on a qualifying offense, still must set the “least restrictive alternative and conditions that will reasonably assure the defendant’s return to court,” and still may not remand the defendant for public safety reasons or risk of re-offending.
Even under the 2020 and 2022 amendments to the 2019 bail reform, the fact that a defendant has a lengthy record is not sufficient for a judge to set bail, unless it is a qualifying offense. Public safety and risk of re-offending are not to be considered by the judge setting bail.
The Failure of Supervised Release
Bail “reform” as a goal has been around for years.
In 2009, NYC started a program called “supervised release” (SR). The program started in Queens County and was sponsored by the NYC Criminal Justice Agency (CJA). The stated goal of the program was to reduce the pretrial population in city jails by giving judges an alternative to setting bail, thereby releasing defendants who could not afford bail. Defendants who were charged with nonviolent felonies and who were viewed as having a substantial risk of having bail set by the court, or who were viewed by CJA as not posing a high risk of pretrial recidivism (i.e., not having more than six misdemeanor convictions, more than one prior felony conviction, or an adult violent felony conviction in the past 10 years) would be released on their own recognizance and put into the SR program. There, they would meet with a case manager, who would gather contact information on defendants and advise them of their responsibilities to keep in contact with the case manager and make all court appearances. CJA would also reach out to defendants to advise them of all court dates. The CJA case manager could make referrals for substance abuse or mental health treatment programs, but attending these was entirely voluntary on the part of the defendant.
An analysis of the CJA-run program’s activity from 2009 to 2012, conducted by CJA in 2013, unsurprisingly found the program to be a success. Of the 1,001 defendants enrolled in the program, 225, or “only” 22.5%, were rearrested while in the program, including 34% of the 16-to-19-year-olds. The rearrest rates included only the time the defendant was “in the program,” meaning up until disposition of the case, and did not include the time between plea and sentence, nor did it track the recidivism rates of defendants after sentence. At this point, there was no real effort to compare “success rates” for defendants in the SR program with similarly situated defendants who did not enter the program.
In 2013, CJA expanded the SR program to Manhattan (MSR) and created a control group, consisting of defendants who fit the criteria for the program but who were not screened for it and were released without entering the program. Defendants in the program were similar to the ones in Queens—charged with nonviolent felonies with the same exclusion of those deemed likely to re-offend. The study showed that 26% of the program defendants were rearrested while the case was pending, compared with just 23% of the control group. (Perhaps the better comparison would have been with defendants in similar circumstances who were held on bail—who, of course, got rearrested at a 0% rate.) Failure-to-appear rates for MSR defendants were lower than for the control group, but rearrest rates for MSR defendants were higher.
Between 2013 and 2016, the population of city jails went from 11,696 to 9,758, a reduction of almost 17%. In the same period, adult arrests went down from 292,581 to 249,446, a decline of almost 15%, continuing a trend of lower jail population as arrests and crime decreased.
In May 2019, the Office of the Queens DA conducted an analysis of the program. It followed the Queens SR defendants during the period between disposition and sentence and for a five-year period thereafter. The office found that the defendants enrolled in 2013 had a rearrest rate of 75% (Figure 2), with an average of five arrests each by May 29, 2019. Of all the defendants enrolled in the program between 2013 and 2019, they had a rearrest rate of 51%, with an average of 3.4 arrests each. Some 44% of the rearrests were for felonies. This included rearrests both while enrolled in the program and after completion of the program.
Percent of Supervised Release Participants Rearrested by May 29, 2019
The clear indication is that this program as a rehabilitation model was a failure. Graduates of the program continued their criminal activity at a rate of 75% over five years. Yet the program was the model for the nonmonetary conditions of release in the 2019 bail reform law, and the later failures of the NMR defendants were easily predictable from the failure of SR defendants.
The Rikers Factor
In her State of the City address on February 10, 2016, New York City Council Speaker Melissa Mark-Viverito called for the creation of an independent commission that would seek ways to reduce the population of city jails so that Rikers Island could be closed.
The effort to reduce the population of city jails was now in high gear. The Independent Commission on New York City Criminal Justice and Incarceration Reform, chaired by former New York State Court of Appeals chief judge Jonathan Lippman (hereafter, the Lippman Commission) was created to study the issue. The Lippman Commission claimed to be open-minded. As Judge Lippman stated in the introduction to the report:
We entered the process with no predetermined judgment. I asked the members of the Commission—law enforcement officials, business leaders, judges, academics, and community activists alike—to look at the justice system with a fresh set of eyes. We let the facts be our guide as we examined both the successes and the failures of recent years.
At the same time, the commission took its mandate from Speaker Mark-Viverito, whose State of the City speech was approvingly noted:
In her 2016 State of the City address, New York City Council Speaker Melissa Mark-Viverito called for fundamental criminal justice reform. Titling her speech “More Justice,” Mark-Viverito announced the creation of an independent commission to explore “how we can get the population of Rikers [Island] to be so small that the dream of shutting it down becomes a reality.”
The Lippman Commission’s mandate was clear—not whether to close Rikers but how to close Rikers. In March 2017, after a yearlong study, the Lippman Commission made a series of recommendations to reduce the size of the city’s jail population to 5,500 from nearly 10,000 through, among other things, bail reform. The report excoriated the conditions on Rikers. Repeatedly calling it a “penal colony,” “inhumane,” “torture island,” and “the land that time forgot,” the commission stated that it was urgent that Rikers be closed. It was a “stain on our city,” the commission wrote.
Armed with the Lippman report, legislators and advocacy groups intensified their campaign to close Rikers. Mayor de Blasio came forward with a plan to reduce the jail population to 6,000, close Rikers, and build four new jails, one in each borough except Staten Island. Because of the need to house upward of 1,500 prisoners in each jail, and the limited building footprints available near the courthouses, the jails in the city’s original plan were skyscrapers, including one 39 stories tall in Brooklyn and another 45 stories tall in Manhattan. The community opposition to this plan was so intense that Mayor de Blasio soon reduced the size of the proposed jails to a maximum capacity of 3,300, in order to reduce the size of the buildings. At the time when the New York City Council approved these jails, in October 2019, the population of the city jail system stood at approximately 7,000.
The mayor and the Lippman Commission were banking heavily on the continued reduction in crime, as well as various alternatives to pretrial incarceration and bail, as a means of reducing the city’s jail population to their proposed levels. They relied on the city’s then nearly three-decade trend in crime reduction, which would reduce overall crime by almost 76% between 1993 and 2020.
The election of several progressive politicians to city and state offices—as well as an active movement seeking to “end mass incarceration,” close Rikers Island, and address the racial imbalance in the jail population vis-à-vis the city’s racial makeup—further propelled the movement to reduce the jail population.
Misrepresenting Data Doesn’t Make New Yorkers Safer
As part of the 2019 bail reform law, the state legislature required the NYS Unified Court System, in conjunction with the NYS Division of Criminal Justice Services, to collect data on bail reform and make them available on its public website for researchers to analyze the effects of the legislation. The information is readily available to the public.
The data cover cases arraigned from January 1, 2020, when the 2019 bail law took effect, to June 30, 2021. The site tracks those cases to the latest date for which they have available statistics, which is September 17, 2021, at the time of writing. The site contains excellent tools that allow researchers to select data points and date ranges to examine how cases have been handled under the new bail regime.
As noted above, the 2019 bail law was changed in certain respects, effective July 2, 2020, so the law currently in effect covers those cases arraigned after that date. Changes to the law made in the 2022 budget did not become effective until May 2022, and cases arraigned under it are thus not covered in the data. In looking at these data, it is important to note several things:
- The data cover, at the time of writing, only cases arraigned from January 1, 2020, to June 30, 2021.
- The data cover activity on those cases only as far as September 17, 2021, at the time of writing. In other words, any dispositions, rearrests, or court activity after that date are not included in the data.
- The data cover only rearrests of defendants while their cases are pending. That means that after a defendant’s case is disposed, by plea or dismissal or otherwise, any rearrests are not included in the data. Arrests occurring between plea and sentence, or after sentence, for example, will not be included. In the original OCA release of data, all rearrests were included, resulting in some 16,000 additional rearrests. These arrests were later removed from the data, as the law required tracking arrests only while the case was pending.
- The data include only one arrest per defendant—the data are designed to tell us what percentage of defendants were rearrested, not how many times they were rearrested. Thus, a defendant who gets arrested five times while his case is pending is counted as only one rearrest. This, when coupled with no. 3 above, means that overall rearrests of released defendants will be undercounted.
There are several ways to analyze these data. Since the analysis here deals with NYC crime statistics, it is limited to NYC cases. Because there are so many different conclusions based on these numbers already published, it is useful to look at what each claim is based on, before proceeding.
State Senator Andrea Stewart-Cousins and other bail reform advocates claim that “2% of defendants get rearrested for a violent felony while their case is pending.” To arrive at that statistic, Stewart-Cousins took the number of all defendants who were rearrested for violent felonies in NYS during the 18-month period (6,533) and divided it by the total number of defendants arrested (284,096), arriving at 2% (it is actually 2.3%). So that statistic is technically true.
Among the 284,096 people arraigned in NYS during that period, 72% of them were arraigned for misdemeanors; for about a third of these defendants, it was their first arrest. The 284,096 number also includes 44,112 defendants, mostly upstate, for whom there is no rearrest information. When you take those 44,112 defendants out of the equation, the violent felony arrest rate increases to 2.7%. But even that 2.7% result ignores the 17,479 defendants rearrested for nonviolent felonies and the 23,939 defendants rearrested for misdemeanors while their original cases were still pending. When they are added in, the rearrest rate becomes 20%. Thus, one in five of all the defendants arrested in NYS for any crime are rearrested while their case is pending.
The 2% number is often stated inaccurately in headlines and in conversation as “2% of the defendants released on bail get rearrested,” without one or both of the qualifiers “for a violent felony,” or “while their case is pending.” As a percentage, it is useless as far as analyzing the impact of bail reform. But in the real world, it still means that at least 6,533 serious violent crimes were committed by people released on bail, and an additional 41,418 nonviolent felonies and misdemeanors were committed by people from the same pool.
The 2% statistic cited by Stewart-Cousins also treats all arrests from January 1, 2020, to June 30, 2021, equally in terms of the rearrest data, but the data on rearrests end on September 17, 2021, for all defendants. In other words, a defendant arrested on January 1, 2020, has almost 21 months to get rearrested and have those data recorded, while a defendant arrested in June 2021 has only about a three-month period in which the rearrest will be recorded. Therefore, the rearrest numbers in the data will be larger for the defendants arrested in January 2020 than for the defendants arrested 18 months later, in June 2021. Including all defendants in the calculation, regardless of the amount of time they were at risk of rearrest, artificially lowers the overall rearrest rates.
Similarly misleading is a study by New York City Comptroller Brad Lander’s Office. Lander cites an analysis published by the New York City Criminal Justice Agency, which gives monthly figures of rearrests of people out awaiting trial. Lander uses that report to claim that only 1% of people awaiting trial get rearrested for a violent felony each month. Again, it is easy to drop the qualifiers: “awaiting trial,” “violent felony,” “each month.” He also claims that the 1% figure has remained unchanged since bail reform. But there is another way to describe the data. If you look at the numbers in the very same analysis before bail reform, people awaiting trial committed an average of 227 violent felonies per month in the first four months of 2019 (pre-reform) and 228 per month in the first four months of 2020. Those numbers rose to 267 per month in 2021 and 297 per month in the first four months of 2022. In reality, the number of violent felonies committed by people out on bail each month went from 227 in 2019 to 297 in 2022, an increase of over 30%. It is all in the packaging of the numbers. You could just as easily have said that only 1/30th of 1% of people awaiting trial get arrested each day, or 1/720th of 1% get arrested each hour. Comptroller Lander, like Senator Stewart-Cousins, ignores the nonviolent felonies and misdemeanors committed by people on bail, as if those crimes are meaningless in the lives of everyday New Yorkers.
A Dramatic Decrease in Public Safety
For the purposes of this analysis, unless otherwise noted, I will use the data for defendants arraigned from July 1, 2020, through August 31, 2020, covering the process of their cases through September 17, 2021, the latest OCA data at the time of writing. This will give us data on the new (2020) bail law’s first two months and will give us an insight into the rearrest rates of those defendants arraigned during that period, allowing a reasonable amount of time to gauge their post-arraignment behavior. Notably, these rearrest numbers count arrests only prior to the disposition of the underlying case and only one arrest per defendant.
First, we look at all felony arraignments in that period in NYC (Figure 3). There were 3,680 felony arraignments in NYC between July 1, 2020, and August 31, 2020. Of that group, 981 (27%) of them were rearrested before their case was disposed, 5.6% of them for a violent felony.
Defendants Rearrested While Awaiting Felony Trial in NYC, July 1–August 31, 2020
When we look at the felony arraignments in Figure 4, we see that of the 3,680 defendants arraigned, 2,564 (69.7%) had a prior conviction or pending case when they were arraigned. Of those 2,564 defendants, 805 (31%) were rearrested before their case ended, and 177 of them (6.9%) were rearrested for a violent felony.
Defendants with Prior Conviction or Pending Case Arrested While Case Is Pending in NYC, July 1–August 31, 2020
When the 2020 bail law was passed, the legislature created a new category of NMR, modeled after CJA’s Supervised Release program referred to above. If the defendant was to be released with nonmonetary conditions, the judge could set other conditions on the defendant’s release. Some of these conditions included requiring the defendant to be in contact with a pretrial services agency; abide by travel restrictions; not possess a firearm; maintain employment, housing, or education; and refrain from associating with certain persons. The court could also refer the defendant to a pretrial services agency for placement in “mandatory” counseling or treatment. The court could require the defendant to abide by these conditions of NMR only if it found on the record that the defendant “poses a risk of flight to avoid prosecution.”
Under pre-2019 law, defendants deemed a flight risk were more likely to have bail set, and those who also had a pending case or prior convictions were even more likely to have the court set bail. Judges also had the discretion to set bail on any crime. Under bail reform, judges cannot set bail on defendants with extensive prior criminal records unless the new crime is a qualifying offense. They cannot even set nonmonetary conditions of release on such defendants unless they find the defendant to be a flight risk. Of the 2,564 felony defendants arraigned in our sample who had a prior conviction or pending case, 492 of them were also deemed a flight risk by the judge and were released on nonmonetary conditions, illustrated in Figure 5. Eventually, 248 of them (50%) were rearrested during the pendency of their case, 43 (8.7%) of them for a violent felony.
Flight Risk Defendants’ Rearrest Status After Nonmonetary Release in NYC, July 1–August 31, 2020
In Table 8 below, the latest rearrest data at the time of writing are broken down by crime and criminal record. The first column shows the current arraignment charge, the second column shows the number of defendants arrested for that crime in the time frame, and the third column shows the percentage of those defendants who had a prior conviction or pending case at the time of their arraignment, ranging from a low of 34% for DWI to a high of 90% for Burglary 3 (commercial burglaries). The fourth and fifth columns show the rearrest rates for defendants with no prior arrests or pending charges versus those with prior convictions or pending charges at the time of their arrest. The rearrest rates are substantially higher for every crime where the defendant has priors.
Rearrest Rates by Offense and Prior Criminal Record in NYC, July 1–August 31, 2020
The chart (Figure 6) below illustrates the rearrest rates for defendants with prior convictions or pending cases compared with those with no priors, grouped by arraignment charge.
Rearrest Rate by Offense and Prior Criminal Record
Table 9 and following chart (Figure 7) show the rearrest rate for defendants with criminal convictions or pending cases by release decision at arraignment, i.e., what conditions of release the judge set at the defendant’s arraignment on the pending charge. Defendants released on NMR have substantially higher rearrest rates than those released on recognizance (ROR) or where bail was set. This should not really be surprising, as to be released on NMR, the judge had to have found that the defendant was a flight risk, and, since the defendants all have criminal records, they would logically be expected to be at a higher rate of re-offending. They did not disappoint.
Percentage of Defendants with Records Rearrested While Case Pending by Release Decision in NYC, July 1–August 31, 2020
Rearrest Rates by Offense and Bail Status
Ironically, the NMR program has turned out to be quite a reliable predictor of future criminal activity—or, at least, of future captured criminal activity. The overall numbers are straightforward. Of all defendants arrested for felonies in NYC, 70% have a prior conviction or pending case at the time of their arrest. More than a quarter of all felony defendants (27%) will be arrested before their case is disposed of. Almost a third of the felony defendants who have a prior conviction or pending case when they are arrested will be arrested again while their case is pending. And more than 50% of those defendants placed in the NMR program will be rearrested before their case ends. All these rearrest numbers must be kept in context. They do not capture the full extent of the alleged criminal activity of those defendants released after arraignment. They count rearrests only while the case is pending and count only one arrest per defendant. In addition, criminals do not commit crimes because they think they will get caught. They commit crimes because most of the time they get away with it. NYPD’s clearance rate in the first quarter of 2022 on burglary is 36.9%, robbery is 49.4%, grand larceny is 17.2%, and grand larceny auto is 12.8%. Both the recorded number of rearrests and rate of rearrests are lower than in reality, as is the number of crimes actually committed by these defendants released under bail reform.
Conclusions and Recommendations
The NYC movement to “end cash bail” was designed to end “mass incarceration” in a city that had one of the lowest incarceration rates of any major city in America. In 2014, NYC’s incarceration rate was 194 per 100,000 people, compared with Los Angeles (263), Chicago (281), Houston (294), Dallas (368), and Philadelphia (810). The movement was designed to end racial disparities in the NYC jail population vis-à-vis the makeup of the general population, but the percentage of black inmates in the NYC jail system has risen slightly since the 2019 law went into effect, from 55.3% on April 1, 2019, to 58.8% on June 13, 2022. To look at it differently, there are 1,025 fewer black inmates in city jails since the 2019 bail reform law passed. But in the first two full years (2020 and 2021) since the reforms took effect, measures of public safety for black residents in NYC have declined. For instance, the proportion of murder and nonnegligent manslaughter victims in NYC who are black increased from 56.6% in 2019 to 65% in 2020 and 67% in 2021. Although black New Yorkers made up only 24% of the city’s population in 2020, they were the victims of 74% of shootings.
The efforts of the legislature in the past two “reforms” of the 2019 law were clearly designed to respond to political pressures and deal with specific crimes that were garnering intense press attention. The amendments included so many conditions to the crimes that were made qualifying offenses that the reforms are ultimately meaningless. This Whac-A-Mole corrective process will do little to solve the structural damage done to the criminal justice system by these laws.
The overly restrictive nature of the New York bail reforms does not allow judges to deal with the myriad of criminal activity that comes before them. Defendants with extensive prior criminal records charged with misdemeanors could be more of a danger than defendants charged with serious felonies on their first arrest. But bail could be set on the latter, not the former. To say that a judge can never set bail on, or remand, a commercial burglary defendant with a prior felony conviction and 13 prior misdemeanor convictions, or a shoplifting defendant with 30 prior misdemeanor convictions, or a defendant with three prior felony assault convictions now charged with misdemeanor assault is entirely illogical (as we have seen, those examples are not unlikely exaggerations). There are simply too many permutations to expect the state legislature to cover them all. The legislature must enact prudent, sensible bail reform that allows judges the flexibility to tailor the proper pretrial conditions that serve the public’s interest as well as the defendant’s.
Real reform would include legislation that makes the following changes:
- Allow judges to set bail, remand, ROR, or conditions of release for any crime and any defendant. There should be a presumption of release for misdemeanors and nonviolent felonies, but such presumption could be rebutted by the defendant’s prior record, pending cases, the nature of the crime, and other factors that indicate that the defendant is a flight risk. Similarly, there should be a presumption of bail, remand, or nonmonetary conditions for defendants charged with violent felonies or weapons offenses. This presumption could also be rebutted by evidence of the defendant’s roots in the community, lack of criminal record, and similar factors.
- Allow judges to remand defendants without bail if they are deemed a danger to the public or at a high probability of re-offending. This is an option available to judges in 49 of the 50 states and the federal government. Allowing judges to remand defendants deemed to be a danger to society removes the economic disparity between people for whom bail is set.
- Allow judges to change the conditions of release, including more restrictive release conditions, higher bail, or remand if the defendant violates any of the original release conditions or fails to comply with the judge’s order. Judges should be allowed to do this on any crime, not just qualifying offenses. Currently, judges cannot remand defendants or set bail if they fail to comply with court-ordered conditions of release on a non-qualifying offense.
- Require OCA to collect and publish additional data, including rearrests of defendants between plea and sentence, rearrests post-sentence, and to provide a full accounting of the total number of rearrests, not just the number of defendants rearrested.
In addition to these proposals, NYC must take a hard look at the latest plan to close Rikers Island and to build four community jails with a totally inadequate maximum of 3,300 jail beds. Lawmakers and activists based this plan on dual assumptions of perpetual reductions in crime and a safe reduction in jail populations. Neither of those assumptions is a reality, and, in fact, they have proved to be mutually exclusive. New Yorkers must conduct bail reform in a way that allows our prosecutors, defense attorneys, and courts to deal effectively with the city’s rising crime rate. Under the “old” bail laws and the “broken” criminal justice system, index crime in NYC was steadily reduced over 27 years, from 1993 to early 2020, by 76%. In just two years of the new bail laws and other progressive reforms, index crimes in NYC rose 36.6%. There are many reasons for the rise, but as the analysis above demonstrates, it is not a coincidence that the sudden increase in city crime came at precisely the same time as the release of 2,000 career criminals from city jails.
Figure 1-A shows trends in the number of non-index offenses from January 1 through approximately the first two and a half months of each year, 2017 through 2022. To depict all the offenses on the same scale, 2019 is treated as the baseline, and offenses are reported as the percentage difference between each year and 2019. Thus, all the offenses appear as 0% for 2019.
Non-Index Offenses, YTD Through Mid-March, Relative to 2019
CompStat Numbers on Non-Index Crimes YTD Through Mid-March
About the Author
Jim Quinn is a former executive assistant district attorney in the Queens County District Attorney’s Office, where he served for 42 years.
Special thanks to Robert VerBruggen, fellow at the Manhattan Institute, for his contribution in preparing figures for this report.
Photo by Dmitry Vorobyev/iStock