Covid-19 is rampaging through the country’s jails and prisons. As of last week, the Marshall Project documented more than 9,400 cases of covid-19 in U.S. jails and prisons, with more than 140 deaths. A Twitter account tracking prison data puts the figure higher — more than 18,000 staff and prisoners infected in the 45 states providing data. The known infection rate in jails in prisons is about 2½ times higher than in the general population, and as of this week, eight of the 10 largest outbreaks in the country have been at correctional facilities.

And even these grim numbers are assuredly low. As with most other things, when it comes to coronavirus testing, prisoners are a low priority. For example, as of this week, 29 prisoners in New Jersey are confirmed to have died from the virus. But the state has tested less than 1 percent of its incarcerated population. Despite an outbreak at a jail in Miami, Florida has only tested 615 prisoners, with 208 positive. But that’s from an incarcerated population of 95,000. And the state’s Department of Corrections only began posting the data after several facilities were caught trying to hide virus-related deaths.

In the facilities that are doing more widespread testing, the figures are jaw-dropping:

Nearly a quarter of prisoners tested at the Bledsoe County, Tenn., Correctional Complex have tested positive. Overall, about 20 percent of prisoners tested in Tennessee have tested positive, yet of the 16 correctional facilities surveyed by a local TV news station, 11 had tested five or fewer prisoners. Nashville journalist Steven Hale reported that, of the state’s incarcerated population of 30,000, just under 4,000 have been tested. Two months into the outbreak, the Riverbend maximum security prison in Nashville had tested just two people.

At a rural prison in Arkansas, 680 of the 1,700 prisoners have tested positive.

— Ohio is one of the few states currently mass testing prisoners. The results are astonishing. At one facility, more than 2,200 prisoners and 174 staffers tested positive. At another, nearly 1,500 prisoners and 93 staff. As of this week, 4,072 of the 6,375 inmates and staff the state has tested are positive for the coronavirus, an infection rate of over 60 percent.

— North Carolina has tested less than 2 percent of its prisoners. But at the one facility where mass testing has been done — Nuese Correctional Institution — 460 of the 770 inmates tested positive, or nearly 60 percent.

At Lakeland prison in Michigan, the first and so far only prison in the state to test all of its inmates, 785 of the 1,403 prisoners were positive.

Harris County, Tex. has tested 500 of its approximately 3,000 incarcerated people. So far, 262 have tested positive, an infected-to-tested rate of 52 percent.

— The infection rate at the Rikers Island jail is five times that of New York City, and 30 times that of the United States. The prison’s first coronavirus death was a man who had been jailed for a technical parole violation. New York Gov. Andrew M. Cuomo (D) has refused to consider mass commutations to prevent further spread.

It isn’t much better at the federal level. In early March, after numerous warning from public health officials about the virus in jails and prisons, federal prosecutors were arguing in court filings that prisoners were more protected from covid-19 inside a prison than out. A few weeks later, the Federal Bureau of Prisons put out a statement boasting that only 10 inmates had tested positive. But that’s because the agency wasn’t testing. Five days later, the number grew to 29 prisoners and 30 staff. Then came the outbreak at the Oakdale facility in Louisiana. Prison officials there responded by simply refusing any more testing of symptomatic inmates.

As of the most recent BOP update, 1,692 federal prisoners have tested positive, as have 349 staff. The Associated Press reported this week that 70 percent of federal prisoners who have been tested have tested positive. At the same time, the Marshall Project recently reported that less than one half of one percent of federal prisoners have been tested.

At the few facilities where mass testing is taking place, infection rates are similar to those we’re seeing in state prisons. At the FCI Terminal Island in Los Angeles, for example, half the inmate population has tested positive.

So far, 33 federal prisoners are confirmed to have died from covid-19. The first confirmed death was Patrick Jones, convicted of possessing cocaine with intent to distribute in a school zone. He had 21 grams of cocaine in his apartment. The “school” his apartment was near was a junior college. He was sentenced to 27 years in prison.

The most recent confirmed death was Andrea Circle Bear, a Native American woman who died after giving birth while on a ventilator. Circle Bear was serving 26 months on drug charges. A federal judge initially approved her admission into a drug-treatment program, but that decision was rescinded after federal prosecutors argued that her drug use presented a threat to her unborn child. Instead, she was sent to a prison facility in Texas, where despite her low-level crime, pregnancy and at least one preexisting condition, she was denied release, contracted the virus and died. The Rapid City Journal reports that at her sentencing, U.S. Attorney Ron Parsons told her, “Don’t let yourself or your property get mixed up in the world of illegal drugs. It ends badly."

It didn’t have to be this way. Over the past two months, defense attorneys and advocates across the country have begged, pleaded, screamed and worked frantically to free as many people as possible, often putting their own health at risk.

In King County, Wash., an early epicenter of the outbreak, public defenders filed motions to release low-level offenders, the sick and elderly, and those nearing the ends of their sentences. One attorney told me that while she and her colleagues attended release hearings in person, prosecutors voiced their opposition over the phone to avoid exposure. In Washington, D.C., city officials assured the public that the jail was taking appropriate precautions, citing written policies. But they didn’t actually visit the jail in person to make sure those policies were being followed. The city’s public defenders did and found the obscene conditions later verified by an independent auditor. In Louisiana, a sitting judge dismissed covid-19 concerns by explaining to colleagues that drug users have bad hygiene and therefore are more of a threat to spread the virus outside the jail than behind bars.

After pressure from some members of Congress, on April 3, Attorney General William P. Barr issued a memo outlining terms for some federal prisoners to be released to home confinement. But as with the D.C. jail, those policies appear to only exist on paper. The aforementioned Marshall Project report found numerous examples of prisoners being told they would be released, only to find the release orders were subsequently rescinded. Politico reported that the Federal Bureau of Prisons was isolating prisoners before releasing them, sometimes in the same units where it was isolating prisoners showing symptoms of infection. It’s such a mess of mixed signals and bungled operations that one federal judge recently called it “Kafkaesque.”

This is what happens when a bureaucracy built on punishment is tasked with compassion. It’s like asking the dishwasher to make coffee. That isn’t what it’s for.

Since the 1980s, there have been periodic waves of stories about people mysteriously found hanging in local jails — people arrested for low-level crimes such as petty theft, or drunken driving, or mundane traffic offenses. Inevitably, these flurries of stories have come with speculation that the deaths were murders. It’s hard to imagine why someone would hang himself over a reckless driving charge.

Instead, while there have certainly been plenty of incidents of inmate deaths at the hands of jail staff, most of these deaths were suicides. These suicides themselves don’t occur in waves. They’re pretty consistent. But so often, a high-profile incident gets journalists digging, and they discover the others that have been occurring all along. Once it’s clear that the deaths were suicides after all, the outrage then dies down until the next cycle begins.

Sometimes, a few voices will make the obvious and important point: People who are arrested for low-level offenses should not be committing suicide. And that they are should tell us that there’s something wrong with our jails.

But that point never really seems to resonate with the public. Americans don’t want jailers actively killing people. But we don’t want our jails too cozy, either. We want them punitive. If that occasionally means conditions that might drive otherwise healthy people to suicide, or neglect that leads to preventable death from other causes, that’s tragic but not tragic enough to hold anyone accountable.

The covid-19 outbreaks are a symptom of the same problem, but distilled, sped up and presented in starker relief. The main difference is that unlike abuse of inmates or jailhouse suicides, this time, our neglect of the incarcerated comes with a price: As jail and prison staff get infected, they’ll bring the virus home and infect the surrounding community. But we seem to be okay with that, too.

In the early days of the outbreak, I spoke with a number of public defenders and criminal-justice-reform advocates who were hoping the mass emptying of the jails would help make the case for bail reform. Once thousands of low-level offenders were released to prevent virus spread and needless deaths, the thinking went, we’d see that keeping people behind bars because they can’t afford bail is purely punitive, and doing so does nothing to improve public safety.

But that argument rested on a faulty premise. It assumed that when confronted with the urgency of the covid-19, that upon the realization that prisons and jails by their very nature make inmates sitting ducks for a highly communicable and potentially lethal disease, judges, governors, prosecutors and other public officials would have enough regard for the health, welfare and humanity of incarcerated people to at least get low-level offenders out of harm’s way. It assumed they would give a damn.

It was a good-faith assumption, grounded in the notion that even the most retributive among us would want to prevent needless mass deaths behind bars. It was wildly optimistic.

Read more: