One year into New Jersey's nationally watched overhaul of its bail system, the state's pretrial jail population has dropped 20 percent as courts have all but stopped setting cash bail.
Instead, judges are deciding whether to release or detain defendants before trial after reviewing a scored assessment of their chances of not showing up for court or endangering the public.
The system, which has been held up as a model by advocacy groups who say poor offenders are too often stuck in jail because they can't afford bail, "is working as intended," said Judge Glenn A. Grant, acting administrative director of the New Jersey Courts.
"No pretrial release system can be perfect," Grant said in an interview last week. Court officials are considering how to better address domestic violence in the new system, he said. But "this one removes the unfairness of relying upon money. That is the most significant accomplishment."
The bail changes, which took effect Jan. 1, 2017, have some vocal critics. Newark Mayor Ras Baraka last week called on officials "to fix the serious flaws in bail reform" after a woman was killed in a domestic-violence incident — as did the New Jersey State Policemen's Benevolent Association. And the bail bond industry is fighting the changes in court.
But court officials say statistics show the law is playing out as hoped. Just 44 defendants in New Jersey had cash bail set last year, according to a report released last week by the judiciary.
The state is largely meeting the law's time frame for determining whether defendants should be released or detained: Courts made decisions for 81.3 percent of jailed defendants within 24 hours, and 99.5 percent within 48 hours, according to the report.
"People used to sit in jail for days — weeks," said Roseanne Scotti, New Jersey state director of the Drug Policy Alliance. "It's just an incredible achievement."
But some questions remain unanswered. While the judiciary's report broke down the numbers of defendants who were released and detained, advocates are awaiting data on the demographics of those defendants. They also want information on how often released defendants failed to appear in court.
"We're really in a situation where all eyes are on New Jersey. Everyone wants to see how reform is working, and these are critical parts of that question," said Alexander Shalom, senior staff attorney at the American Civil Liberties Union of New Jersey.
Also unclear is how the state will pay for the system going forward. In making release decisions within 48 hours, judges must consider the recommendations issued by a newly created Pretrial Services Program, which uses an analytic tool that considers a defendant's criminal and court history and assigns a numeric score to their risk of skipping court or endangering the public.
That program, which also monitors released defendants, is going to run a "significant financial deficit" by the fiscal year starting in July 2019, Grant said. He noted that the gap was anticipated. While the program costs $35 million a year, court fee increases that were implemented to pay for it only net about $22 million.
The state's 21 counties are also dealing with added costs, because prosecutors' offices have had to hire additional staff, said John Donnadio, executive director of the New Jersey Association of Counties.
While the pretrial jail population has dropped, "that hasn't translated to cost savings as of yet," Donnadio said. "It may, two or three years down the road" as jail staffs shrink through attrition, he said.
A spokesman for Gov. Murphy did not respond to a request for comment on the governor's opinion of the bail changes and whether he would include funding in his budget proposal.
New Jersey's bail changes passed with bipartisan support in 2014. In addition to replacing cash bail with the risk-assessment system, lawmakers and then-Gov. Chris Christie backed a constitutional amendment — later approved by voters — that let judges order defendants jailed until trial.
The idea was twofold: Keep poor, nonviolent defendants unable to pay even nominal amounts of bail from languishing in jail, while at the same time giving judges new power to keep potentially dangerous defendants behind bars without the option to make bail.
Of 142,663 defendants charged with crimes last year, 44,319 were issued arrest warrants. Of that group, 8,043 — about 18 percent — were ordered detained by courts, according to the report.
That detention rate varied by county: from 7.7 percent in Bergen County, up to 34.4 percent in Atlantic County, according to court statistics. Prosecutors must file motions to detain defendants; judges then decide whether to detain or release the person, and under what level of monitoring.
Shalom said that "some prosecutors' offices are seeking detention too often, and some judges are granting detention too often."
Grant said detention rates vary among counties because crime varies among counties, and prosecutors have discretion in what charges to bring.
"The public-safety assessment is just a recommendation," Grant said. "The system is designed to be run by humans."
The assessment was developed by the Laura and John Arnold Foundation, a philanthropic organization. It scores a defendant based on nine factors, including the person's age at time of arrest, whether the charge is a violent offense, whether the person has prior disorderly persons or indictable convictions, and whether the person has failed to appear in court pretrial before.
The factors are "the best predictors" of whether a person will commit a new crime or fail to return to court, according to the foundation, which analyzed 1.5 million cases across the United States to develop the analytic tool. The tool is also being used by Arizona and Kentucky and a variety of counties and cities.
While assessments are still conducted, New Jersey defendants charged with escape, murder, rape, or robbery automatically get no-release recommendations. The New Jersey Supreme Court revised the recommendation process several times last year, including so that defendants charged with certain weapons offenses also would receive no-release recommendations. The state Attorney General's Office had asked the court to make changes that flagged more defendants as high-risk.