When a woman accused of trying to kill someone with a poisoned cheesecake appeared on a list of Rikers Island prisoners to be freed because of coronavirus risks, New York’s district attorneys banded together to strike her name. They also struck the names of two men charged in an armed robbery during which a police detective was killed by friendly fire. The pandemic poses an outsize threat to the nation’s jails and prisons, where confined populations can’t take the recommended precautions and often lack access to such basics as soap and hand sanitizer — sometimes even running water. Federal and state authorities, as a result, have begun to release thousands of inmates.
The focus, however, has been on low-level offenders. “There are some at-risk inmates who are non-violent and pose minimal likelihood of recidivism and who might be safer serving their sentences in home confinement,” Attorney General William Barr wrote tepidly late last month in a memorandum to the Bureau of Prisons director. And in California, Gov. Gavin Newsom said that his state’s efforts “will be for those nonviolent offenses, and we will do it in a very systematic way.”
The debate over which prisoners to release early and what to do with them rarely considers those charged with or convicted of violent crimes, except to declare that they should stay behind bars. “I have no interest — and I want to make this crystal clear — in releasing violent criminals from our system,” Newsom said. Gov. Tom Wolf of Pennsylvania, meanwhile, proposed a furlough policy that excluded, among others, anyone with a current or even prior conviction for a violent crime.
Such refusal to think about crimes of violence is, unfortunately, to be expected. Even a decade into a sustained push to reform the way this country deals with crime, serious conversations about how we handle violence remain almost impossible. For example, late last year, when Gov. Matt Bevin of Kentucky pardoned or commuted the sentences of hundreds of prisoners just before he left office, outrage followed; some of those prisoners had been convicted of murder and rape. Two prosecutor associations in Kentucky released a statement denouncing the releases as “arbitrary, callous,” and Senate Majority Leader Mitch McConnell (R-Ky.) called them “completely inappropriate,” considering that they included “people who were incarcerated as a result of heinous crimes.” Days later, the Pennsylvania Board of Pardons refused the petitions of 15 prisoners, some elderly, who were serving life sentences for violent crimes. These decisions, conversely and tellingly, were met largely with silence.
The attitude those incidents reveal — that people convicted of violent crimes are in a special category that deserves less compassion and harsher treatment — ignores the math, misunderstands human behavior and, perhaps most important, reflects a poor moral choice. Our draconian approach toward violent crime rests on viewing certain people, and certain groups of people, as not fully human. This has always been a pressing concern in criminal justice reform; during the pandemic, it is a matter of life and death.
It’s true that too many people are in prison in the United States for drug crimes, so initial efforts to push for reform rightly focused on drug cases, the easiest ones to build consensus around: Nobody deserves a 15-year sentence for selling 48 hydrocodone pills. But a solid majority of Americans believe that about half of all people in prison are there for drug offenses, even though the actual fraction is much lower. While about half the people held by the federal system are in for drugs, the feds hold only about 10 percent of all prisoners. In the state systems, with the remaining 90 percent, only 15 percent are in for drugs. Those serving time for violence constitute 55 percent. And that’s a floor on the fraction of people really in prison for violence. The available data classifies inmates by the most serious offense for which they were convicted. If someone is arrested for domestic violence but is found to possess heroin, he may plead guilty to just the heroin charge and appear in the data as a “nonviolent drug offender,” even though the only reason the prosecutor sought prison time for the drug charge was the uncharged violence. Moreover, that 55 percent figure relies on the narrow definition of violence used by the Bureau of Justice Statistics: homicide, rape and sexual assault, robbery, and assault. Some states define violence more broadly, including offenses ranging from purse snatching and burglary to drug dealing and embezzlement, and so they have even higher percentages of prisoners officially in for violence.
About 25 percent of the entire American prison population, or about 365,000 out of more than 1.3 million people, is serving time just for homicide, rape or sexual assault. Almost everyone with a long sentence — those who would actually benefit from commutations, whether or not a pandemic is involved — is in prison for serious violence. A recent report by the Urban Institute found that among those serving the longest sentences — the top 10th — 94 percent had been convicted of serious violence; 69 percent of those, murder. We often read about people sentenced to decades in prison for drug crimes, but we read about them because they are rare.
Even before the coronavirus reached us, it was obvious that any effort to scale back prisons would require us to contemplate a less punitive response to violence — one that doesn’t rest on misunderstandings about the nature of violent crime and those who commit it or are victimized by it.
A stack of empirical papers makes it clear, for example, that what deters crime is the certainty of punishment, not its severity — the likelihood of getting caught, not the length of the prison time later imposed. Law enforcement and other tough-on-crime groups often dispute this claim, as the Florida Sheriffs Association did recently in a report opposing reforms in part because, it claimed, cutting harsh sanctions would undermine deterrence. But the actual data is unambiguous. Commuting a sentence now for a crime committed a decade ago will not affect someone’s decision today about whether to commit a violent crime (even assuming that decision is being made coolly and calmly, which is probably not the case, given the role that drugs, passions and the harms of prior traumas, not to mention youthfulness, play in driving violent acts).
Nor does clemency hamper the goal of incapacitating criminals, which has long been an article of faith among proponents of long sentences like those imposed under “three strikes” laws, even if the crimes aren’t violent: In Ewing v. California, for example, the Supreme Court in 2003 upheld a 25-years-to-life sentence for a man convicted of stealing three golf clubs. In her opinion for the 5-4 majority, Justice Sandra Day O’Connor wrote that the sentence was “justified by the State’s public-safety interest in incapacitating and deterring recidivist felons.” But another stack of empirical papers consistently demonstrates that people both age into and out of violent behavior. For various physical, hormonal and social reasons, people become more violent in their teens and 20s but tend to desist in their 30s and 40s — which is around the age that the risks from contracting the coronavirus start to rise, making the issue all the more pressing now. How many people demanding we throw away the key on an 18-year-old also say, in their 40s, “Man, I’m not the same hotheaded person I was in my 20s”? (Answer: all of them.)
One reason older people are less likely to be involved in violence is that they are more likely to have a job and to be in a stable relationship, both of which appear to lead to less criminal behavior. So it’s not just a function of age; the more time we keep people locked up, the harder it is for them to get work and form relationships upon release. This clearly increases the risk of reoffending.
What’s more, even if long sentences were to have stronger deterrent and incapacitating effects than the research suggests, small waves of gubernatorial commutations will have minimal impact on public safety since the criminal justice system punishes only a small fraction of violent crimes. According to the 2018 National Crime Victimization Survey, about half of all nonlethal violent crimes are reported to the police. About half of all reported aggravated assaults and a third of all reported robberies and rapes result in an arrest. About one-third of people arrested and charged with violent crimes see their cases dismissed, while half are convicted of a felony. And of those convicted, only about half are sent to prison. In other words, something like 5 percent of all violent crimes lead to incarceration.
Even for homicide, which draws the most ire when it comes to clemency, the numbers deceive. The clearance rate — the number of crimes reported to or known by the police that end in an arrest — is about 60 percent for homicides; of those arrested, 13 percent see their cases dismissed, and 80 percent get a felony conviction, with 90 percent going to prison. That means that only 43 percent of all homicides result in a prison admission. Those numbers are from 2006, the last year with consistent data, but victim reporting rates were mostly lower in 2018, and police clearance rates are roughly the same, so there’s no reason to think things have improved.
Yes, policing and punishment influence crime rates, but the number of people for whom clemency matters is such a small fraction of the total number of people causing harm that the overall effect on public safety would be next to nothing. Even in times like these, with some people pushing for policies that would significantly reduce the prison population, at least temporarily, any sort of impact is likely to be slight at best (especially since those furloughed from prisons would be required to return once the pandemic has passed). Tellingly, crime is actually down in many places.
If prison growth is driven by how we punish violence, it may seem as if as a society we think mass incarceration isn’t such a bad thing after all. It’s clear that we are a deeply punitive country, but it’s time to delve into the morality of severely penalizing violence. I once heard Heather Ann Thompson, the author of “Blood in the Water,” a history of the 1971 Attica prison uprising, talk about an exercise she does with her students. She asks them to think about someone they love dearly, and then to imagine that person being horribly murdered, raped or seriously beaten. She then asks them what sort of punishment the attacker deserves, and she gets the expected harsh replies. Then, she asks, imagine that same loved person, but instead of being the victim, now they are the attacker. How do you want to punish them? Unsurprisingly, the answers change; the students don’t want to let their loved ones off the hook, but they’re far less willing to throw those lives away.
That’s because our sense of what a crime “deserves” is driven less by what that crime is than by who the perpetrator is. Race and poverty inevitably play huge roles. This is starkly apparent in the political response to New York state’s recent bail reform law, which went into effect in January. These changes denied judges the ability to rely on cash bail for almost all misdemeanors and certain nonviolent felonies. Before the new law, people charged with low-level offenses could easily find themselves locked up in county jails for days or weeks or even years — facing the loss of a job, of a home, the risk of physical or sexual assault and abuse — simply because they could not make bail. Critics and some newspapers, the New York Post in particular, insistently attacked the reforms on the grounds that they would make “us” less safe. (The reforms were curtailed a few days ago, when Gov. Andrew Cuomo threatened in the midst of the pandemic striking New York City to shut down the entire state government, including the Department of Health, unless the legislature passed a budget that included revisions to the law.)
But who is “us”? The bail reform law did not emerge spontaneously in Albany. It was a product of years of lobbying and grass-roots agitating in the very communities the critics purport to speak for. Bail reform came about because of sustained efforts in higher-crime, poorer communities of color, led by groups such as VOCAL-NY and JustLeadershipUSA — communities that feel most heavily the costs of crime and the costs of enforcement. When they push for reforms, they are saying the system causes more pain than good.
To be clear, poor communities of color aren’t monoliths; there are plenty of people in them who favor tough policies, and respect for the police appears to be near all-time highs among nonwhite Americans. Victims’-rights advocates in general reject most reforms, such as the California law that bars children 16 or younger from being tried in an adult court, even for murder. In one case in Sacramento, for example, the 15-year-old boy who murdered a local high school football star had to be tried in juvenile court, and will be released by the time he is 25; the victim’s mother, Nicole Clavo, assailed the requirement that the boy be tried as juvenile, asking, “Why should he be afforded the luxuries and opportunities that my child will never receive?”
But the “us” the critics really speak for are mostly those who live in relatively wealthy, safe areas. For them, bail reform, expanded parole or any other significant shift in how we handle criminal justice seems as if it could (slightly) increase the (already small) chance of their being victimized, with no real direct upside, since the people being needlessly stopped, detained, jailed, charged, tried, convicted and imprisoned — and suffering all the harms those interactions create — are not their family members, not their friends, not their co-workers and not even their neighbors. Why, the thinking goes, should “we” expose ourselves to risks that have no upside for “us”? This troubling logic has long challenged reform efforts, and it poisons how the criminal justice system responds to the coronavirus right now.
Once we think carefully about how we define “us,” “our” willingness to attack, say, Bevin’s leniency while being silent about the Pennsylvania board’s inaction becomes all the more troubling. The asymmetric response is so commonplace that it has a name: the “Willie Horton effect,” after a man serving a life sentence in Massachusetts who fled a furlough program in the 1980s and committed a horrible crime. The media outcry over this one bad case — despite the program’s 99 percent success rate — led Massachusetts and almost all other states to abolish or curtail furlough programs within a few years. The Horton effect persists because we don’t fully value the lives of those in prison for needlessly long sentences.
Consider this recent example: A few months ago, the Robert F. Kennedy Human Rights foundation aimed to bail out hundreds of women and children from Rikers and other New York jails. The proposal faced significant backlash from New York police and prosecutors, and it generated significant press coverage. Very few in the political class, however, seemed to support getting mothers home to their families for Thanksgiving, or getting children back to their parents. There was little concern for love or affection or friends or families, which all those inmates had but few of “us” cared to acknowledge, because the people in jails and prisons don’t actually count as people. All that mattered were the risks that the bailout posed to “us.”
Same with the cases in Kentucky and Pennsylvania, and with the foot-dragging over furloughs in the face of the coronavirus. The Bevin commutations scare “us”; the Pennsylvania rulings, which quietly sent people back to their cells to die of old age, merit barely a peep. This attitude toward crimes of violence will leave us as the world’s largest jailer for decades to come. It is an attitude that makes us no safer and reflects a willingness to throw away some lives but not others, often for the most cynical, chauvinist reasons. And in a time when the coronavirus is starting to spread rapidly through our nation’s prisons, it will prove lethal to many who are behind bars.