Most thinking about crime is highly informed by passion and instinct, whether it is the bitter desire to see punishment inflicted on those who have done wrong, or sympathetic fellow-feeling with the accused and imprisoned. The vengeful part of the human soul says: If You Can’t Do The Time, Don’t Do The Crime. The merciful part says: Tear Down The Walls, Everyone Must Be Free. (When in doubt, it’s usually best to defer to the merciful part. The merciful can make poor decisions that hurt people, but they never commit atrocities.)

This conflict between impulses was obvious in a recent controversy over the sentencing of Larry Nassar, the Michigan State doctor who sexually abused hundreds of young gymnasts over multiple decades. One would struggle to find a less sympathetic individual than Nassar. Instead of contrition, he showed brazen arrogance, writing a letter to the judge insisting that it would be too painful for him to listen to the victims’ testimonies, dismissing the girls as “wom[e]n scorned.” (Why Nassar’s lawyer ever let him send such a self-damaging letter remains a mystery.) So it was quite obvious why Nassar’s judge, Rosemarie Aquilina, displayed a certain relish in passing his sentence. Aquilina declared that it was a “privilege” to sentence him, that she was signing his “death warrant.” She even suggested that, were it not for the Constitution’s prohibition on cruel and unusual punishment, she’d happily let some especially cruel and unusual things happen to him:

Our Constitution does not allow for cruel and unusual punishment. If it did … I would allow some or many people to do to him what he did to others.

Aquilina earned praise from some for her staunch support of Nassar’s victims. But when we look closely at her actual words, they show more than just compassion for those who were hurt. They also show a desire for payback. For “many people” to “do to [Nassar] what he did to others” means that Nassar would be repeatedly raped. And Aquilina said that the only thing keeping her from sentencing Nassar to be raped was the fact that she was legally prohibited from doing so, rather than because there would be a moral problem with such a sentence.

Actually, we should note that Aquilina may have come closer than she thought to imposing her ideal of justice: rape and sexual victimization are ubiquitous in America’s prisons, and as  a child sex abuser with no prior time inside, Nassar has a non-trivial chance of being brutally assaulted exactly as the judge hoped. Precisely because Americans do not particularly care when monstrous people get “what’s coming to them,” prison rape is not an issue that especially stirs the national conscience. (It’s even played for laughs, e.g., “don’t drop the soap.”)

Once we realize the implications of Aquilina’s remarks, they become troubling, because a humane justice system does not include judges who openly fantasize about watching horrible things happen to defendants. If we are being fair to Aquilina, though, we can say that while she may have found it pleasurable to think about doing physical violence to Nassar, she did not necessarily mean her remarks literally. It’s understandable that a judge who had just sat through dozens upon dozens of disturbing victim testimonies about a man’s abuse would temporarily let her careful professionalism slip and convey her real human feelings.

Frankly, questions about retribution are fraught with difficulty. We can declare that vengeance is an ignoble instinct, that while it may be natural it needs to be overcome in favor of a calm, dispassionate response to crime that focuses solely on how to reduce harms. But telling somebody who has been victimized, or who has seen a loved one victimized, or who has vicariously experienced some part of the trauma of a victim, that they “shouldn’t” feel a desire to see evil punished also feels wrong. An eye for an eye may make the whole world blind, but try giving a forgiveness lecture to a parent whose child has been killed.

The truth is, criminal justice is very difficult to do well. Restorative justice initiatives, in which the victim and perpetrator try to achieve reconciliation over the perpetrator’s wrongdoing, sound ideal in theory, and have produced some impressive outcomes in individual cases. But there are instances where it’s simply inapplicable: For many domestic abuse victims, the last thing you want the justice system to do is further entangle the victim with their abuser. And you will rarely have ideal defendants or ideal victims, who are easily able to put aside all bad feelings and work together to make amends.

The American criminal justice system is, in many ways, horrifying. By now it’s well known that we jail people at staggering rates, higher than any other country in the world. This shouldn’t be the case: If every other country has figured out how to have a society with fewer imprisoned people, we are doing something wrong, and it should be fixable.

But it’s difficult to know what reforms might actually work, which would actually get us toward a society with both low crime and a low (or ideally, nonexistent) prison population. And many of the talking points among criminal justice reformers obscure this complexity. We are capable of correctly identifying a series of problems (e.g., the criminal justice system is bloated, racist, and excessively punitive; prisons do not give inmates useful skills; the stigma of criminal convictions makes re-offending more likely; the police harass and abuse communities they are supposed to protect; public defenders are underfunded and the plea bargaining process is stacked against defendants; there is no real accountability for police who needlessly injure and kill people, etc.) Yet the word “reform” is, in and of itself, nothing but a feeling that something ought to happen so that these problems do not exist. What the “something” is, well, that’s tough.

James Forman, in his Pulitzer Prize-winning Locking Up Our Own, significantly complicates the most popular narratives about mass incarceration. Forman, a former public defender in Washington, D.C. (and, full disclosure, a former teacher of mine), was horrified at seeing his black teenage clients sent off to prison for years, sometimes for minor offenses. He believes that the prison system is an unconscionably harsh moral stain on the country. But in thinking about how D.C. courts became so manifestly unfair, Forman notes something counterintuitive: the black community, which has been most affected by the growth of prisons, also produced many political initiatives to increase sentences back in the ’70s and ’80s. Forman notes that when he was in D.C. courtrooms, sometimes the judge, prosecutor, court officer, and defendant would all be black. Yet the results were still clearly unjust. This seeming paradox has caused Forman to question the idea that the criminal justice system can easily be called the “new Jim Crow.” After all, it ensnares poor people of all races, and sometimes black communities are “locking up their own.”

The “paradox” is not really a paradox at all, actually: There is no single “black community,” and black-majority cities and neighborhoods have huge class differences and political hierarchies. What Forman reveals is that many black judges and politicians at the local level actually saw crackdowns on crime in their communities as advancing the cause of racial justice; he recalls being exasperated every time a judge prepared to give a black teenage defendant a speech about how disappointed Dr. King would be if he were alive. Forman shows that as black neighborhoods were torn apart by the PCP and crack epidemics, some local leaders adopted the kind of “law and order” rhetoric more commonly associated with Richard Nixon. Black newspapers decried the “growing crime menace,” with one even calling for drug dealers to be “tarred and feathered, burned at the stake, castrated, and any other horrendous thing which can be imagined.” “We’re going to fight drugs and crime until the drug dealer’s teeth rattle,” said Atlanta mayor Maynard Jackson.

Forman suggests that local policies had a greater role in increasing mass incarceration than is usually assumed. While we tend to focus on “national campaigns, federal legislation, [and] executive orders” we should also look at “small choices” like how a public official responded to citizen complaints about used syringes in alleyways. Forman examines the mechanics of specific programs, like the anti-gun initiative Eric Holder launched when serving as U.S. Attorney for Washington, D.C. Forman shows how officials like Holder increased the use of indiscriminate searches, and that while they ostensibly targeted weapons they ended up ensnaring low-level offenders and increasing punitiveness more broadly. (Conservatives sometimes like to bring up the effects of tough gun control laws on black people, as a cynical argument against new firearm restrictions. But it’s correct that any new law, if enforced in a system that disproportionately targets minorities, will disproportionately punish minorities.) Forman doesn’t downplay the role that white policy-makers played, but he believes that until we recognize the multiplicity of causes, we won’t actually know where to start in fixing the problem.

Law professor John Pfaff, in Locked In, is similarly concerned with challenging conventional wisdom about criminal justice. Pfaff is scathing about what he calls the “standard story” pushed by critics of mass incarceration. The story they tell is that the prison boom was a consequence of the War on Drugs, which sent countless low-level drug offenders to prison, extended sentences for drug offenses to unreasonable lengths, and is driven by a profit-driven private prison system constantly hungry for more bodies. Pfaff says, while the fact that our prison population is outrageously large is not in dispute, this analysis of the causes is faulty in almost every way.

In fact, low-level drug offenses do not account for most of the growth in prison populations. Most people are imprisoned for violent offenses, and it will be necessary for reformers to accept that we can’t significantly change the numbers if we focus solely on nonviolent offenders (the ones that are easiest to persuade the public to sympathize with). In fact, prosecutors have gotten tougher on all criminals, and we will ultimately need to make the argument that violent criminals, too, deserve to be punished less. It’s going to be difficult, Pfaff says, because it requires acknowledging that “safety should not automatically be our paramount concern in every instance.” It’s easy to eliminate all risk of re-offending by never letting anyone out of prison, but that’s why the elimination of risk can’t be the sole factor guiding policy. Pfaff says there has to be a “cultural change,” in which the public becomes less aggressively punitive toward people who have harmed others. That will not be easy.

Importantly, there needs to be more focus on decisions made by prosecutors, and less on parts of the system like private prisons, which are a symptom rather than a cause. Pfaff shows that it’s not just laws or police that matter, but how prosecutors use their discretion to charge people with crimes. This makes mass incarceration largely a “state and local problem” rather than a federal one. Pfaff’s book cautions us against pushing too hard for solutions that sound sensible (say, abolishing private prisons or releasing nonviolent offenders) but that will not have much of an impact on raw numbers of incarcerated people.

Pfaff has a number of useful suggestions for what actually will work, and his book isn’t entirely discouraging and critical. The largest obstacles are actually political: When you try to close prisons, there are strong objections, because prisons are job-creators. The correctional officers’ unions are powerful, but the lobby on behalf of violent criminal offenders does not have much clout. The reason cultural change is important is that it is a necessary prerequisite to political action: No state legislature is going to seriously risk the public backlash that occurs when an offender released under some new program goes out and commits some horrible crime. In order for the legislators’ calculus to change, there mustn’t be a public backlash, because we need to have built a widespread understanding that a more merciful criminal justice system is worth a small amount of additional risk.

When you begin to assess what it would actually take to repair the institutions of justice, you can become discouraged very quickly. It’s not just that once you start getting “practical,” diving into the nuances and complications, there are all sorts of messy competing values and political obstacles. It’s also that the scale of the task is so vast. Alex Vitale, in The End of Policing, portrays a system that is quite literally beyond repair. The central thesis of his book is that efforts to reform or fix the police are doomed, because the core of the problem is the very structure of the policing system itself. Until you’ve altered that, any improvements will remain marginal and superficial. Furthermore, many reforms touted as improvements may actually make policing worse. Vitale says that “community policing, body cameras, and increased money for training reinforce a false sense of police legitimacy and expand the reach of the police into private lives.”

Vitale’s book goes through an exhaustive list of policing’s worst aspects. There are, of course, the shootings of unarmed people. (As well as armed people, who in many cases have broken no laws, as in the case of Philando Castile.) There is the increased policing of schools, and the arrest of students whose infractions would previously have been handled by the school administration. There is the criminalization of homelessness, and the police’s harassment of the poor and vulnerable. There is the policing of sex work, which puts women in jail for making a living. There is the militarization of police, who begin to resemble branches of the armed forces rather than friendly neighborhood cops-on-the-beat. And there is the political surveillance and arrest of Muslims as part of the War on Terror. The relationship between police and the policed is not one of protection but one of control. Vitale views it as a “liberal fantasy” that “police exist to protect us from the bad guys.” Because police forces are dysfunctional entities at their core, it’s hopeless to think you can change them. For example, if you require body cameras, you may end up catching some misconduct on film, but you’re also introducing a new form of electronic surveillance. If you require diversity training, officers will sit through it obligingly and then return to their regular routine. “There is no technocratic fix,” Vitale says.

This framing may or may not be helpful. It certainly does create a sense that the injustices are extreme and the solutions must match the scale of the crisis. But it also makes a fair criminal justice system seem so distant as to be impossible, since it requires nothing short of ending the police. And Vitale himself doesn’t know how to do that. Alongside his criticism of conventional reforms, he does offer ideas for improvements: decriminalize sex work and drugs, redirect policing resources toward community centers and jobs for young people, establish separate unarmed entities to deal with mentally ill people to reduce the likelihood that deadly force will be used on them. But these are sketched out only briefly, and they all sound somewhat “technocratic” themselves. We can all agree that the police should be “ended,” but what does that mean and how would it occur? I am worried that the statement “there can be no justice until there are no more police,” when combined with the obvious fact that “it is inconceivable that there will be no more police,” necessarily produces a despair that inhibits action. Vitale is actually somewhat more subtle whenever he talks about what steps can actually be taken, but he is prone to grand and unqualified statements like “the police are not here to protect you.” (Often they’re not, sometimes they are. Everything is complicated.)

So it’s important to be wary of simple conclusions, because criminal justice isn’t simple at all. But equally important is the need to avoid getting wrapped up in the details, and losing sight of matters that are simple. We can debate how best to create accountability mechanisms for prosecutors or what policies are best suited to recidivism reduction. But when you become too pragmatic and reformist in your thinking, when everything turns into  “on the one hand, on the other” equivocation, you forget that it is necessary to be outraged by the outrageous, and to state clearly that which is obvious. In Keramet Reiter’s 23/7: Pelican Bay Prison and the Rise of Long-Term Solitary Confinement, we see something that is indisputably an egregious abuse of basic human rights: the practice of keeping people in solitary confinement for years and decades at at time. Reiter tracks those who have been shuffled off to Security Housing Units and kept in their cells for, as the title suggests, 23 hours a day, 7 days a week.

For many people, it may be hard to appreciate why solitary confinement is so objectionable. After all, solitude is peaceful. But Reiter suggests that skeptics should try locking themselves in their bathroom for an hour. Then lock yourself in it for a day. Now try a year, or ten. As could be expected, one’s mental health often deteriorates rapidly. Putting someone in an empty room, with nobody to talk to, no hope of leaving, and no work to do, is a form of torture.

However much we may appreciate perspectives like Forman and Pfaff’s, then, which add nuance to criminal justice conversations, many of the issues do not require nuance. They require direct condemnation, using moral language rather than the language of policy and problem-solving. Some issues present dilemmas: How should courts treat victims’ desire for retribution? How can the New Jim Crow thesis be squared with the fact that most of the officers involved in the killing of Freddie Gray were black? If we reject the New Jim Crow thesis, what about the fact that the effects of today’s justice system on black people in many ways resemble the effects of Jim Crow? Are public prisons actually any better than private prisons? Which of today’s policing functions should be taken away from the police and delegated to other institutions? Which institutions should they be?

But while we must necessarily spend a lot of our time lost in the complications and caveats, we can never forget that we are dealing with a urgent human moral problem, and that every moment we discuss it, a million and a half people sit behind bars, some of them going mad alone in rooms that they are kept in 23 hours a day, 7 days a week.

This article originally appeared in our Jan/Feb 2018 print issue. You can purchase a copy in our online store

If you appreciate our work, please consider making a donation or purchasing a subscription. Current Affairs is not for profit and carries no outside advertising. We are an independent media institution funded entirely by subscribers and small donors, and we depend on you in order to continue to produce high-quality work.