‘Arresting’ report hides harms of bail reform

John Jay claims recidivism is down, but numbers tell true tale

Anew report from John Jay College of Criminal Justice’s Data Collaborative for Justice screams the headline that recidivism rates have dropped thanks to New York’s 2020 bail-reform laws.

Claiming to compare apples to apples, DCJ measured people arrested pre-reform in New York City who had bail set (“2019 group”) to people arrested for the same types of crimes post-reform for whom bail could no longer be set (“2020 group”).

Among the successes? “Only” 43.8% of the people released under the new bail laws got re-arrested within two years of their initial arrest, compared with 50% of people who had bail set under the old law.

But this triumph is entirely undercut by larger realities.

COVID-era distortions

First reality: The drastic drop in total arrests in 2020 and 2021. This drop — not bail reform’s effect — caused the lower re-arrest rates for defendants released on bail.

Dramatically, New York City arrests fell from 214,000 in 2019 to 140,000 in 2020, rising slightly to 155,000 in 2021. The 2020 and 2021 declines were due in large measure to COVID, even as crime went up.

Yet in 2022, despite 55,000 more reported crimes than in 2019, there were still 24,000 fewer arrests. This means that defendants in the “2019 group,” who had to avoid arrests in 2019, had a far higher chance of being re-arrested than the “2020 group,” who roamed the streets during the less arrest-risky 2020 and 2021.

Second reality: As DCJ acknowledges in its report, the data available to them only included prosecuted re-arrests in their recidivism rates. But because prosecutors declined to prosecute a far higher percentage of arrests in 2020 and 2021 than in 2019, the tallied re-arrests for the “2020 group” will be artificially lower.

Further, in 2019 most misdemeanor and nonviolent-felony defendants were brought directly to court to be arraigned and thus would be considered prosecuted for statistical purposes. But under bail reform, more misdemeanors and nonviolent felonies were handled with desk appearance tickets, where the defendant is given a summons to appear in court at a later date.

According to Division of Criminal Justice Services stats, 17.1% of summonsed defendants fail to appear for their DAT, so those arrests are also not counted because defendants aren’t arraigned. For summonsed defendants with a prior record (whose data are especially relevant to a study intended to measure recidivism), the number who don’t show up jumps to 31%.

The final reality: Clearance rates (how often the NYPD resolves crimes) are substantially down from 2019.

What do the full data still say? The lower re-arrest rates post-reform reflect changes in policing and prosecution, and COVID, not some miraculous drop in actual crime due to bail reform.

Playing word games

Damningly, DCJ conflates recidivism with re-arrest. Recidivism indicates an individual has returned to criminal activity. Rearrest means the individual was caught for that criminal activity (and also prosecuted in the context of DCJ’s study).

In an era with vanishing chances of being caught, it is rank advocacy — rather than scholarship — to claim the lower re-arrest rate is “unmistakable” proof that bail laws reduced recidivism. Yet DCJ does exactly that — in a report that, no surprise, was partially written, edited and guided by the Brennan Center for Justice and Arnold Ventures, two progressive pro-bail-reform groups.

Was the report at least reviewed or edited by any prosecutors’ organizations or bail-reform sceptics pre-publication? One doubts it would have made it to publication.

Even setting aside the ignored realities regarding re-arrest data, the report still indicates (if you dig through its many subsections) that recidivism among defendants with prior criminal records has actually risen!

Yes, re-arrest rates for released defendants with prior violent-felony arrests grew from 61.9% to 72.4% under bail reform. And rearrest rates for released defendants with any pending cases swelled from 62.8% to 68.8%. Even DCJ had to acknowledge (hidden on page 44) that “future legislation or policy might make fewer [emphasis added] ‘high risk’ individuals (e.g., people with a prior violent felony) subject to mandatory release, allowing for wider judicial discretion in considering bail.”

DCJ notes on this page that defendants who had a pending case or prior violent-felony arrest were significantly more likely to get re-arrested.

‘Increased rates’

Persevere to the end of the “Summary and Conclusions” to learn that counter to all headlines: Bail reform “increased recidivism for people with substantial recent criminal histories.” Or see page 38: “For people charged with burglary in the second degree, pretrial release was associated with increased rates of felony re-arrest.”

DCJ’s raw numbers rebut its own argument even further! In its prereform group, 3,510 defendants had bail set. Of those, 50% (1,755) got re-arrested. But there were fully 12,350 in the post-reform group — and 5,409, or 43.8%, of them got rearrested. More than triple the rearrests in real numbers!

Other Office of Court Administration data, not in this report, show that under bail reform, 79% of defendants with a pending case or prior conviction who are charged with petit larceny and then released on non-monetary conditions get re-arrested while their case is pending.

For residential burglaries, the re-arrest rate is 62%, commercial burglaries 70%, grand larceny 70%, robbery third degree 70%.

DCJ — just like our state Legislature — simply refuses to see the forest for the blindingly apparent trees. We have more crime in New York City because we have more criminals on the streets. We have more criminals on the streets because of bail reform. Why is this so hard to understand?

Let’s stop playing around with misleading data and fix the law!