Skip to content

Locked Up

Why the movement for criminal justice is stalled

The United States’ experiment with a relentlessly punitive criminal justice system is now more than four decades old. But it’s worth remembering that things were not always this way. For most of the twentieth century, America was not too different from Canada or Europe in the relative size and stability of its prison population. It was in the middle of the 1970s that U.S. prison populations started to rise steadily. By the time the prison boom peaked about ten years ago, the United States was holding nearly 25 percent of the world’s inmates, despite being home to about 5 percent of the world’s people. And the cumulative numbers of Americans who have run afoul of the law are staggering: since 1980, U.S. authorities have made over half a billion arrests, imposed something on the order of a quarter-billion jail admissions, and admitted about twenty million people to prison.

On any given day, there are about 1.5 million people in prison, nearly 1 million on parole, and more than 3.5 million on probation. County jails hold about 750,000 people every day, and people go to jail more than ten million times each year: a majority of them pretrial detainees yet to be convicted of a crime. Each year, police around the country make over ten million arrests and our courts process about thirteen million misdemeanor cases and three million felony cases.

An awareness that we’ve gone too far has settled in. Around 2009, something started to change—or so it seemed. Trying to scale back the massive reach of our criminal justice system has become one of the few genuinely bipartisan issues in these highly polarized times. Yet despite years of hard work at the local, county, state, and federal levels, this decade-long effort has yielded underwhelming results. The total daily correctional population—the total number of people in prison or jail or on parole or probation—has dropped only by about 10 percent, with almost all of that decline coming from having fewer people on probation. Prison populations are down by about 7 percent; jail populations are down by under 6 percent.

Of course, any sort of decline after four decades of growth is something to celebrate. The gains have been small because many of the reform efforts to date have focused on low-hanging fruit—like scaling back sanctions for drug offenses or expanding parole options for those convicted of nonviolent crimes. These are all good policies, and they were the most logical places to start. Yet a decade into this reform push, it is increasingly clear that taking the easy steps is not going to get us to any fundamentally new place; it is essential that we start turning our attention to deeper and more challenging issues before the moment for reform slips away.

Just as the initial push for mass incarceration was bipartisan, and just as current reform efforts have been bipartisan, so too is the resistance—if not outright hostility—to taking more difficult steps. There are many obstacles and blind spots in criminal justice reform, but here are three of the most significant:

  • The nation has been unable to have a clear and sustained conversation about how to respond to violence.


  • We see the vast inequities in who gets punished and then push for more punishment for those who “get off easy,” rather than less punishment for those who don’t.


  • There has been very little national discussion of the powerful role public-sector correctional officer unions play—far more than any private prison firm—in defending the status quo.


A Violence Confrontation

The new criminal justice consensus has been to ease up on “nonviolent” crimes and drug offenses in particular. Liberals and conservatives alike talk about the need to release people convicted of nonviolent crimes from prison to concentrate our prison resources on those convicted of violence; you hear it in liberal states like Maryland and conservative ones like Georgia. In fact, 75 percent of conservatives and 86 percent of liberals favor reducing prison sentences for nonviolent crimes, according to a 2016 poll conducted by Vox. Yet, when asked the same question about those convicted of violent offenses, with the added detail restricting the question to those posing little risk of reoffending, 68 percent of conservatives and 55 percent of liberals were against reduced sentences.

It is not surprising that people are resistant to changing how we punish violence; they do not appreciate how much it drives mass incarceration. In that same Vox poll, 68 percent of liberals and 62 percent of conservatives agreed with the claim that “nearly half of all” people in U.S. prisons are there for drug offenses. If most people are in prison for nonviolent crimes, then it follows that we should be able to make huge strides by focusing solely on these sorts of offenses.

But the respondents in the Vox poll—who are consistent with every conversation I’ve ever had about prison growth—had the numbers wrong. The percentage of people in state prisons for drugs is not anywhere close to 50 percent; it is 15 percent. It’s true that about half of all federal prisoners are serving time for drug-related offenses, but the Feds hold only about 10 percent of all U.S. prisoners. Of that 15 percent in state prisons on drug charges, only about one-fifth of them—or 3 percent of the total state prison population—are in for possession; the rest are in for trafficking (although that is an ambiguous term that often ensnares users, not just dealers, and which does not include many kingpins).

The simple fact is that people convicted of violence are now the single largest group we hold in prisons—about 55 percent. Of course, “violence” is a tricky term to pin down; plenty of crimes we classify as “violent” might not involve much, or even any, actual physical harm. Robbery, for example, which is theft by force or threat of force, is automatically classified as a violent crime, even in cases where no one is hurt. Some states go further and count certain drug and property crimes as “violent” (although the 55 percent cited above excludes such cases). Consequently, some activists now seek to downplay the category. Yes, they say, a majority are in prison for violence, but most of it is not “real” violence.

But there’s no getting around the centrality of serious violence to prison growth. Nearly half of the people in prison for violence—over 25 percent of the total prison population—are incarcerated for murder, manslaughter, rape, or sexual assault. If we emptied the prisons of everyone except people convicted of those four categories of crime, our incarceration rate would still be higher than those of almost all European countries. And plenty of the other assaults and robberies involve serious harm as well.

This refusal to think about violence leads to some painful ironies. Almost all discussions about expanding parole eligibility, for example, restrict the debate to nonviolent crimes. Yet a recent report by the Urban Institute looking at people serving long sentences found that 94 percent of them had been convicted of serious violence; of that 94 percent, 70 percent were in for homicide and another 10 percent for sexual assault or rape. For expanded parole to have a serious impact, it has to include serious violent crimes.

The gains have been small because the reform efforts have focused on low-hanging fruit—like scaling back sanctions for drug offenses or expanding parole options for those convicted of nonviolent crimes.

There’s also a way that statistics understate the prevalence of violence. People are classified by the most serious crime for which they are convicted, not arrested. So if someone is arrested for domestic violence but ultimately pleads guilty to a lesser drug charge—if, say, the police found heroin on them during the arrest—that person shows up in the data as a “nonviolent drug case,” not a “violence case,” even if the prosecutor sought prison time, as opposed to probation or diversion, because of the violence. Thus, some of the 45 percent who are in for nonviolent crimes, including the 15 percent in for drugs, are really in for violence. We lack the data to figure out how big that fraction is, but it seems plausible that at least 60 to 65 percent of those in prison, if not more, are “really” there for violence, and a good fraction of those for serious violence.

Yet most on the left and right alike remain unwilling to confront this issue. In many cases, it is out of political fear—what politician wants to face the kind of attack George H.W. Bush’s supporters used against Michael Dukakis in 1988, in which Dukakis was blamed for the weekend furlough of convicted killer Willie Horton? Or, recall the reaction to the 2016 sentencing of Stanford swimmer Brock Turner in California after he was found guilty of sexually assaulting an unconscious woman. His supposedly “lenient” sentence triggered outrage—and culminated in the first successful recall of a California judge in more than eighty years. Such reactions suggest there remains a strong bipartisan desire to impose fairly harsh prison sentences on violent crimes, out of a misguided sense that it keeps us safe, as well as a deep-seated desire to punish, regardless of public safety.

The problem is that prison is a remarkably ineffective way to respond to crime, even violence. Prison in general has little deterrent effect, and making sentences longer has barely any impact—it’s the risk of detection and arrest, not the size of the sanction, that really matters. And while it is true that someone cannot commit a crime, at least against the general public, while in prison, a growing body of evidence indicates that the longer someone spends in the prison, the more likely they are to be rearrested, to the point of cancelling out a lot of the incapacitation effect. As reformer Danielle Sered has pointed out, if prison worked well, we would be the safest country in the world. We are not.

Moreover, there are often better, more humane ways to prevent future offenses than locking someone up in prison. We have a growing understanding of what works effectively to reduce violence, whether it is better police staffing and tactics, non-police interventions like Cure Violence, which relies on local community members to intervene in gun violence to prevent retaliation, or more transformative restorative justice options like Sered’s Common Justice program in Brooklyn. Yet these are often viewed as “alternatives” to be used when convenient and ignored in the breach. Even liberal jurisdictions still default to harsh prison sentences: California responded to the Turner sentence by passing new mandatory minimums, and the Baltimore City Council passed new gun mandatory minimums to address violence.

Meanwhile, the communities most afflicted by violence are making it increasingly clear that they themselves do not favor such punitive responses. Take Common Justice, which provides survivors of serious violence (or their families) with the option of undergoing a restorative justice response in lieu of sending defendants to prison. Over 90 percent of those given this option take it. Not necessarily because they don’t feel the urge to punish, but because they know that the person who harmed them will eventually return, and Common Justice appears to make things better while prisons only make them worse. It is a pragmatic decision on their part. Those on the left and right alike justify harshness in the name of “victims” who are increasingly telling us they want something different.

Equal Time

The bipartisan preference for punishment is more than just a response to violent crime. It is often justified by appeals to fairness and equality: when we observe that privileged defendants get better outcomes than disadvantaged ones, the reflexive response is to demand more severity. But such an attitude is what sustains mass incarceration.

Two high-profile cases—Brock Turner and Paul Manafort—provide a useful lens for examining how liberal punitiveness operates and the serious problems that it raises. As discussed above, Turner’s “six-month sentence” (the popular if inaccurate description) led to the recall of the case’s sentencing judge. Paul Manafort, of course, was the former campaign manager for President Donald Trump whose recent sentence for tax evasion and other fraud offenses amounted to seven-and-a-half years, which was roundly denounced as too lenient.

To start, it’s worth asking just how lenient those sentences were. Turner, for example, was not just sentenced to six months in jail, as most reports framed it, but rather to six months in jail and three years on probation (with the threat of fourteen years in prison hanging over his head if he faltered), lifetime registration as a sex offender, and a host of collateral consequences that flow from a felony conviction. Manafort’s seven-and-a-half years is longer than almost all prison sentences imposed for all crimes in places like the Netherlands and Germany, and could very well be a life sentence for a seventy-year-old man.

The bipartisan preference for punishment is more than just a response to violent crime.

Such sentences only feel short because we have internalized a degree of penal severity unseen in most other places. Judge Danny Chin, an Obama appointee, sentenced Bernie Madoff to 150 years in prison for financial fraud, citing the need for retribution and hoping that it would help the victims to heal; Norway, conversely, sentenced Anders Breivik to the statutory maximum of twenty-one years (with parole eligibility at ten) for his 2011 terroristic mass murder of seventy-seven people, including sixty-nine youths—a sentence many survivors and parents of his victims seem to accept as just.

Yet Americans do have legitimate concerns about the racial, gender, and class inequalities that permeate our criminal justice system. In the Turner case, the outrage was fueled in part by the awareness that we tend to under-punish sexual assault when compared to non-sexual assault, and that a less-privileged defendant surely would have received a harsher punishment. The same race and class concerns drove the anger over the Manafort sentence, with commentators immediately pointing to poor, black defendants who had committed less serious crimes but received equal or harsher sentences. And it’s true: we do punish sex offenses less, especially when we take into account the lower rate of detection and arrest, as well as the greater harassment and distrust we subject victims to. And white and wealthier defendants are treated less severely than otherwise identical defendants who are poorer or people of color.

But why is the solution to treat everyone worse? If we are concerned that sex offenses are relatively under-punished, why is the instinctive response—and it often is immediate and unthinking—to demand that sex offenses get punished more, not that non-sex crimes get punished less? If we are afraid that rich defendants get shorter sentences, why don’t we demand that poor defendants get treated better, not that the Manaforts and Madoffs of the world get treated worse? We perceive equality as only being achievable through more severity for all.

Perhaps this reflects a grim political reckoning, a sense that leniency for poor or black defendants is simply impossible, so more severity is the only path to equality. But the reactions seem more visceral than that; the desire to punish is more deeply ingrained than we would like to think, and it applies to violent and nonviolent crimes alike.

The problem is that new harsher laws will likely misfire, hitting poor and minority defendants harder—not just because they make up a vastly larger share of people in the system, but because prosecutors are more likely to use the harsher charges against them while continuing to go easy on the defendants with whom they more easily sympathize.

Moreover, the demands to get tougher on the privileged won’t do anything to serve victims (many of whom prefer less-punitive responses to crime) or improve public safety. If our goal is to reduce campus rape, then rather than impose our vengeance on the small fraction of cases that get reported and produce an arrest, we should invest our time and energy and money into upfront prevention, whether through increased policing and security staffing (which has its own costs, but ones that are less pronounced than increased incarceration), or more emphasis on addressing alcohol and drug abuse. A system of less-severe, Turner-like sentences would not reflect an indifference to sexual offenses but the exact opposite: harsher sentences do very little to address the underlying problem effectively.

In forty-three states, prisoners count as residential population for the purposes of drawing state and federal legislative districts, even though they cannot vote while in prison.

Any real transformative change in our criminal justice system is not going to come about by debating proper degrees of punitiveness, but by fundamentally rethinking how we approach criminal behavior altogether. This is what makes the recent policy proposal from Rachael Rollins, the new district attorney for Suffolk County, Massachusetts (Boston’s county), noteworthy: rather than tweaking how we punish some crimes, she is pushing to move many of them out of the criminal justice system entirely, stating that as a general matter her office will not prosecute a wide range of offenses often closely tied to poverty, homelessness, mental illness, and drug addiction (like certain low-level thefts, drug possession for personal use, or breaking into an empty building to avoid cold weather).

Rollins’s actions can be seen as an effort to broadly reframe how we think about problematic behavior. Her policy is not saying that stealing small amounts or breaking into empty buildings is OK, just that not all bad behavior requires criminal punishment. Criminal punishment is a blunt, heavy-handed tool, one that we wield too widely. And while Rollins’s approach is likely inapplicable to serious violent crime, programs like Common Justice point out that we can fundamentally rethink the role of punitiveness even for those offenses.

The Public Prison Industry

Another major mistake among reformers is to get preoccupied with the private prison industry while ignoring the more significant and consequential role played by the public sector—and especially by correctional officer unions.

The private sector is a relatively minor player in the criminal justice system, despite all the hand-wringing it induces. Mass incarceration, and mass punishment more broadly, is a public-sector failure, not a capitalist one. Which implies, in the end, that the real problem rests not with some shadowy cabal of financiers, but with us, the electorate.

In 2016, private prisons held about 8.5 percent of all prisoners nationwide, and over half those prisoners are held in just five states; at least twenty-one states have no private prisons at all. There is also little evidence indicating that states with private prisons saw faster prison population growth rates than those without them. And contrary to the conventional wisdom, there is no consistent evidence that private prisons are generally worse places than public ones.

Furthermore, private prison corporations receive only a small share of prison funding. States dedicate about $50 billion per year to running their corrections departments, and local governments spend another $30 billion on county jails. Of that $80 billion, only about $4 billion is private prison revenue; less than a billion of that goes to the private firms contracted to provide health care, and only a few billion goes to private telephone and commissary providers. And that’s revenue—profits are what the private prison firms really care about, and those come to about $400 million for the companies that run private prisons. By contrast, the wages that we pay to public correctional officers in state prisons alone are on the order of about $30 billion (or nearly 70 percent of all state prison spending), which is ten times private prison revenue and nearly one hundred times private prison profits.

Correctional officer unions thus have far more at stake than private prisons, and unlike the private firms, they cannot diversify into other areas like re-entry management, probation provision, drug testing and treating, and so on, should prisons close down. We should thus expect the guards to fight reforms that intend to cut prison capacity; they are a more potent political force than the private prison industry, and they are in every state.

Public-sector resistance to decarceration involves more than just wages. In forty-three states, prisoners count as residential population for the purposes of drawing state and federal legislative districts, even though they cannot vote while in prison. Since prisoners are disproportionately people of color from urban areas, this policy effectively transfers Democratic-leaning voters from Democratic parts of the state to more rural, more Republican parts, where they count as—and I say this intentionally—five-fifths of a person with zero-fifths of a vote. State legislators across the country depend on this “prison gerrymander” to keep their seats, introducing yet another powerful barrier to real reform. Prisons also tend to be located in more economically marginal communities—that’s why these towns are often willing, if not eager, to accept a prison—providing still more reasons for the unions and the politicians to aggressively fight prison closures.

Yet while the correctional officer unions represent a serious barrier to reform, it is important to approach the issue carefully. The failure to think about the central role wages play in driving prison costs has led to some dangerous framing for reforms. Many advocates for change—especially, but not only, on the right—frame decarceration as the chance to defund prisons in order to fund other, more effective programs. And these other programs, to be clear, almost always are better than prisons. But carelessly defunding prisons is a risky proposal, since the only way to seriously cut prison costs is to cut payroll, which means cutting staff either by termination or attrition.

Understaffed prisons, of course, are dangerous places. Inadequate staffing has contributed to deadly prison riots in South Carolina, to horrific conditions in Alabama, to violence and death in North Carolina, to name just a fraction of such cases. Thus, arguing that we need to confront the power of correctional officer unions is not just a call for stripping unions of power and cutting wages. In fact, it may be that in the short-to-medium run we need to increase wages, to make sure that prisons remain safe places even as we scale back populations. The goal, however, should be to think about how to do that in a way that provides for some sort of eventual transition away from prisons.

Moreover, while prisons are often one of the few, if not only, stable jobs in their regions, they are miserable places to work. Rates of suicidal ideation are 39 percent higher than in most professions, and the rates of PTSD may be nearly triple that of military veterans (34 percent vs. 14 percent). As many a guard has pointed out, the correctional officers are more or less doing time as well; they just get to sleep at home. Thus, figuring out a way to transition away from prisons toward other employment would not only help end our reliance on incarceration; it would improve the lives of correctional officers themselves.

It ultimately is the public sector, not the private, that is at the heart of mass incarceration.

New York State provides one of the few examples along these lines. The state struggled to close prisons, despite a prison population that has shrunk by about 25,000 people since 1999, in large part because officers fought any sort of closures; the state was dotted with near-empty prisons that were fully staffed. Governor Andrew Cuomo finally secured an agreement to close several prisons by offering the affected communities millions in aid and tax breaks to try to introduce new business. New York City mayor Bill de Blasio has taken a similar tack in his efforts to close down the city’s notorious Rikers Island jail, suggesting that any correctional officers who lose their jobs could be rehired in other capacities by the city, which could simultaneously undermine union opposition and address the very real issue of job loss.

There are also racial justice implications to consider: prisons are often framed as a form of rural white welfare because rural prisons are mostly staffed by white guards. But prison towns in many places, especially in the South, actually skew more heavily minority (and black in particular) than other similarly situated small towns. Many prison conversion programs, then, would need to simultaneously address the racist economic impact of incarceration on those incarcerated as well as the racial disparities in who works for the prisons.

During a time when unionization, including public sector unionization, is under sustained attack, targeting correctional officer unions is likely to be quite unpopular with many on the left. Yet it ultimately is the public sector, not the private, that is at the heart of mass incarceration.

Despite the growing bipartisan consensus in recent years that we are spending too much money on prisons for too little positive effect, there is no easy way to scale back the system. If we are unwilling to find better ways to cope with violence, if we are unwilling to reimagine fairness and equality in terms of shorter sentences for all, if we are unable to imagine new jobs for correctional officers, then we will be stuck with our status as the most punitive, most prison-intensive society on the planet.