New York Needs More Than 3,300 Jail Beds

In public policy, there is often a yawning gap between expectation and reality. So it is with the closure of New York City’s jail facilities on Rikers Island.

Five years ago, when it voted to shutter Rikers and replace it with four borough-based jails, the City Council claimed a hard-fought win for safe, humane and effective incarceration. Fast forward to today, and the situation is much messier. Higher interest rates mean construction costs have skyrocketed, putting the new jails over budget and behind timeline. Mayor Eric Adams has repeatedly raised concerns about the viability of closing Rikers by the 2027 deadline. And the Department of Correction remains embattled as its jail system moves ever closer to federal receivership.

Amid this chaos, one key question remains chronically underaddressed. As envisioned by the City Council, the borough jails will have a maximum capacity of 3,300 inmates — an 80% reduction relative to Rikers, which at maximum capacity could hold 15,000 people. This limited capacity is, as I argued last year in a Manhattan Institute report, the greatest challenge to the safety and efficacy of New York’s jail system, on Rikers or off of it. 3,300 beds is simply not enough space to safely detain all of the people New York needs to hold each day.

The 3,300 figure, the product of years of political wrangling, is unprecedentedly small. At almost no point in its nearly century-long history has Rikers been home to that few detainees. The city jail population did not even reach that low level in the spring of 2020, when the first version of bail reform and COVID-induced policy shifts drove it down to just 3,800 — still 15% higher than the borough jails’ maximum capacity.

Jail capacity matters because, no matter how you slice it, some people need to be detained pretrial. That includes serious flight risks and — almost everyone, including judges on the bench who were among the Lippman Commission that led the push to close Rikers, agrees — serious, violent offenders who pose a risk to public safety. These two categories impose a hard floor on the number of people for whom jail beds are necessary.

Just a glance at recent population statistics shows how hard it will be to get under this floor. As of the end of November 2023, there were roughly 6,000 people incarcerated on Rikers Island. About 1,600 of these faced homicide charges, 1,400 were charged with robbery and burglary, 750 with assault and 500 with gun charges. If every other offender were released — the drug peddlers, the petty thieves, the rapists — Rikers would still be housing 30 percent more people than the borough-based jails can fit. If we assume that some of the nonserious and/or nonviolent offenders are flight risks, that disparity gets even bigger.

As I show in the report, these constraints pose a persistent problem. At almost no point in the six years between 2016 and 2022 could the Rikers population have plausibly been reduced by category such that it was below 3,300. The sole exception was in the depths of COVID. If a global pandemic is required to operate the borough jails at their usual capacity, they will overfill fast.

Some have contended that various technical changes — or divine favor — will make the baseline level of detainable offending low enough for the borough jails’ capacity. In a July 2021 report, for example, the Lippman Commission and Center for Court Innovation (now called the Center for Justice Innovation) argued that the average daily population (ADP) could be brought within the range of 2,700 to 3,150. 

Their estimate, though, relies on too-rosy assumptions. They expected, for example, a large reduction in ADP when the COVID-19 backlog was cleared; four years since the onset of the pandemic, it is safe to say this is not happening. A similar predicted decline attributable to parole reform did not appear. Faster case processing may help, but the Commission projected that applying the case processing standards pioneered in Brooklyn citywide would reduce ADP by only about 500 people — not nearly enough to bring the population under the 3,300 threshold.

In reality, the only hope to substantially reduce the city’s jail population is for crime to not only return to its pre-2020 levels, but to fall substantially below it. The multidecade decline in crime in New York was a remarkable achievement, and advocates are not totally misguided in imagining it might be able to continue — indeed, I have argued as much in these pages. But if New York’s jails have the capacity only to address crime at historically unprecedented lows, then any deviation from that level — either by fault of policy or nature — will break the system. Given the cyclical nature of crime, a policy premised on things remaining good forever is doomed to fail. 

One group that used to understand this, in fact, was the Lippman Commission. In its original call for closing Rikers, the Commission envisioned a system of 5,500 beds, including jails in all five boroughs. It was only in the ensuing process of getting a plan by the City Planning Commission and Council that the De Blasio administration retreated, dropping a Staten Island jail and—just days before the final City Council vote—slashing capacity to 3,300.

Questions about the borough jails’ capacity are not idle criticisms. If Rikers is shuttered and the city has more people to detail than beds to put them in, one of two situations occurs. Alleged offenders over the 3,300 threshold can be forced to double up, leading to overcrowding — a dangerous condition that can violate the Eighth Amendment of the U.S. Constitution. Or, the city can actively release violent and flight-risk offenders, which would mean that the jails are not serving their basic function. In either case, the city would be failing in its basic duties to some of its citizens.

There are options for expanding New York’s jail capacity while still closing Rikers, although none of them is easy. The city could reopen the Lincoln and Bayview correctional facilities in Manhattan, netting about 700 beds (Bayview, damaged by Hurricane Sandy, would likely need serious repairs). It could also try to acquire one of the 900-bed facilities the state sold off: Arthur Kill, on Staten Island, now owned by a production studio; and Fulton, in the Bronx, now controlled by the Osborne Association. Alternatively, the city could construct additional, smaller borough jails. One on Long Island might make sense. So would erecting a small facility adjacent to the Bronx Hall of Justice — the approach originally advocated by then-Borough President Rubén Díaz Jr.

Additionally, New York detainees could be “boarded out” to other counties’ jails, a legal (albeit rare) practice in New York State. Jails on Long Island and in Westchester County have ample capacity and could probably absorb extra population in times of strain. On the other hand, part of the goal of the borough jails was to house detainees in their own communities, enabling easy access to their families, lawyers and support networks. Sending them out of the city entirely is worse on this dimension than sending them to Rikers.

Realistically, not all of these options are on the table. However, if the city were to refurbish Lincoln and Bayview; construct two small additional jails in the Bronx and on Staten Island; and agree to board out post-trial detainees to Long Island or Westchester, it would likely grow its capacity to about 5,300 beds. That’s just shy of the 5,500 originally recommended by the Lippman Commission and north of the De Blasio administration’s original target of 5,000.

Of course, after 2027, there will remain a lot of unused jail capacity closer to the city — on Rikers Island itself. While some of the site’s 10 facilities are in disrepair, the jail can in theory hold nearly 15,000 people. If just 1 in 15 beds is habitable, the city could increase its post-transition bed capacity by 30%. Using some slice of Rikers as a supplement to the borough jails would be a relatively cost-efficient and effective solution to the basic problem of capacity. That such an approach is not under serious discussion reflects how efforts to close Rikers are driven at least as much by ideological aversion to the location — and to incarceration in general — as they are by a genuine concern for effective detention. 

Safe, humane and effective jails are essential for the nation’s biggest city to have. But jails aren’t safe and humane if they’re overcrowded, and they aren’t effective if violent criminals aren’t detained for lack of space. The city does not have to leave Rikers open — though refusing to use any of it seems more ideological than practical. But it does need to have a plan for where to send people who can’t fit in the new borough jails. 3,300 beds won’t cut it.