Several years ago, I drove more than two hundred miles from Green Bay, Wisconsin, to the isolated town of Boscobel. Boscobel is hard to get to; its airport has little commercial traffic, and the nearest Greyhound station is more than thirty miles away. If Wisconsinites have heard of the place at all, it’s probably because Boscobel is home to the Wisconsin Secure Program Facility, a heavily fortified prison on the edge of town. I was there to visit one of the many inmates being held in solitary confinement.

Our visit was not exactly personal. It took place in a small room with a video screen near the ceiling, on which the image of the inmate suddenly appeared. I strained both my neck and voice to communicate with him. In the hour of conversation permitted to us, the inmate shared a story of remarkable resilience. Years of isolation in a small cell had left him with numerous physical ailments. Harsh and racist treatment from the prison staff had complicated his struggle to maintain sanity in a bleakly monotonous environment. He told me that only his practice of Islam had enabled him to survive. I left deeply impressed with his fortitude and personal dignity—and horrified at the conditions of his incarceration.

That visit to Boscobel brought home to me the degrading character of contemporary solitary confinement. In the past three decades, the United States penal system has embraced a brutal policy of isolating inmates. Despite a growing consensus that the tactic is a gross violation of human dignity, an estimated fifty- to eighty-thousand prisoners are held in isolation in the nation’s jails and prisons. The policy represents a terrible tradeoff. Widespread use of solitary confinement may bring short-term order to some penal institutions, but it does serious damage to inmates and exacts a steep moral cost.

It would be hard to exaggerate the grim reality of life in U.S. penal institutions. They are overcrowded and dangerous, and prison officials struggle to maintain order. Gangs brutalize inmates and threaten staff members. Sexual predators assault vulnerable inmates. Violent men and women with little to lose wantonly attack corrections officers. And, in recent decades, thousands of mentally ill people have entered jails and prisons. Unable to adapt to incarceration, they often create serious disturbances. Corrections officials, responsible for protecting staff and inmates, must separate violent prisoners from the rest of the population, discourage breaches of discipline, and create an atmosphere where inmates can safely pursue recreational and other activities.

To achieve these goals, prison officials employ punitive isolation, placing inmates who violate disciplinary rules in solitary confinement. There they must live in small cells for at least twenty-three hours a day, released only for exercise (usually in a small cage), showering, or the occasional visit from relatives. Their lives are carefully controlled: they cannot read, watch television, listen to the radio, or engage in other activities without official permission. Often, they can read items only from an approved list of books or periodicals. Many cannot go to the bathroom, shower, or exercise without cameras monitoring them. When leaving their cells, they undergo humiliating body-cavity searches. They may have access to a chaplain, but cannot attend religious services; Catholic inmates may go years without receiving Communion if prison officials refuse to allow it. Depending on their behavior, inmates may also be barred from seeing relatives. Visits that do take place occur with a barrier of protective glass or, like my visit, via a video screen. Inmates can remain in isolation for months or years, and many have no idea when they will be released back into the general prison population.

Scholars and activists trace the use of this kind of solitary confinement—routine, long-term, and punitive—to the federal prison in Marion, Illinois. In the 1960s Marion replaced Alcatraz as the nation’s toughest federal detention facility. It housed particularly troublesome inmates, and when officials used a variety of behavioral-modification techniques to break their wills, the prison descended into chaos. Things came to a head in 1983, when inmates brutally murdered two corrections officers. In response, Marion officials instituted a controversial lockdown, confining inmates to their cells for twenty-three hours a day. The prison’s rate of violence dropped, and what began as a temporary policy lasted for twenty-three years, despite protests from inmates and their supporters.

The Marion lockdown was one of several experiments in isolation in the last decades of the twentieth century; others occurred at a federal women’s prison in Lexington, Kentucky, and a state prison in Trenton, New Jersey. Because these experimental policies survived court challenges and halted disorder, they became models for local, state, and federal penal authorities. As prison populations exploded in the 1980s and ’90s, state after state constructed high-tech “supermax” facilities devoted entirely to solitary confinement. These prisons were often located in remote rural areas, which made it difficult for relatives to visit; architects worked meticulously to design buildings that maximized isolation. Other state prisons built new “segregation” units that could isolate as many as two hundred inmates at a time. In the 1990s, the federal government opened its supermax in Colorado, which houses convicted terrorists and other prominent criminals. Most jails and prisons now have sections dedicated to solitary confinement, typically labeled “Secure Housing Units,” “Administrative Segregation Units,” or even “Reflection Cottages.”

Corrections officials maintain that these isolation units reduce gang influence and contain violent inmates. Scholars find it difficult to substantiate these claims, since penal authorities are rarely forthcoming with data about inmates in solitary, and scholars cannot gain access to segregation units. A few, including Sharon Shalev and Daniel P. Mears, have tried to evaluate the efficacy of solitary confinement, and have found little empirical evidence that it achieves its stated goals. The threat of segregation does not clearly motivate inmates to change their behavior. Removing gang leaders from the general population may create a power vacuum that produces more violence. And some prisoners subjected to isolation become so damaged that they pose a renewed threat to staff and inmates when they return to the general prison population; if released from prison, they may also take out their rage on others in their communities.

Such realities provide reason to doubt that isolation enhances institutional safety in the long term. The difficulty of evaluating solitary confinement’s efficacy also reflects the arbitrariness with which penal institutions administer it, and the lack of public scrutiny or oversight of the process. Officials insist they employ due process in deciding whom to isolate. But few inmates are sentenced by a court to serve time in solitary confinement; such sentences are administrative matters handled by Corrections Departments. Wardens and disciplinary committees decide who should be isolated, and inmates who contest these decisions rarely prevail. To be sure, some selected for solitary have committed horrific crimes, or are violent and incapable of living with others. Many, however, are guilty of only minor disciplinary infractions. Thousands are thrown into solitary for alleged gang association, insubordination, possession of contraband material, protests against prison conditions, or even for writing allegedly subversive essays.

Prisoners can receive up to a year in segregation for minor infractions; those guilty of more serious offenses can serve years. In Louisiana’s Angola penitentiary, a man named Herman Wallace spent more than forty years in solitary before being granted a “compassionate release” while dying of cancer. Some inmates have had years tacked on to an initially short sentence in solitary for committing disciplinary offenses while in segregation. In the Secure Housing Unit (or “SHU”) at California’s Pelican Bay prison, those charged with gang affiliation were told they could exit solitary only by informing on others or renouncing their gang membership. Given the dangers of this choice, many inmates chose to remain in solitary. In other cases, inmates have been thrown into segregation while still waiting for their cases to be tried. Recently at Rikers Island in New York City, a sixteen-year-old awaiting trial spent a total of four hundred days in solitary confinement. He was released from jail after three years when the flimsy case against him fell apart. Last year, a man in New Mexico settled a civil-rights lawsuit for his treatment in a local jail, where he was incarcerated on suspicion of a DWI but never given a trial. Placed in solitary for nearly two years and denied access to dental care, he was forced to extract one of his own teeth.

Corrections officers face daily threats of violence. On constant alert, they nonetheless frequently suffer injuries at the hands of angry or mentally ill inmates. As a result, some feel safer working in a segregation unit, where inmates cannot easily assault them. Yet some corrections officers I spoke with consider the segregation unit the worst place in the prison to work, a site of continual noise, violence, and degradation. The work can be especially stressful and dangerous. When an inmate disobeys an order, groups of officers in an “extraction team” approach his cell, dressed in protective gear and armed with shields, Tasers, and other weapons. If the inmate refuses to comply, the officers will flood his cell with chemical agents. In some cases they have reportedly thrown stinger grenades, which spray rubber pellets into a concentrated area. Eventually the officers open the prisoner’s cell and violently subdue him. A few states have employed dogs in cell extractions (as documented by Colin Dayan and Human Rights Watch).

Such extractions often result in injuries to both staff and inmates. But merely living in isolation is enough to cause deep damage to inmates. Because of the closed character of most penal institutions, social scientists find it difficult to measure how solitary affects people psychologically. Most of the research that has been conducted points to solitary’s psychological harm. After a relatively short period in isolation, a person begins to disintegrate mentally and emotionally. Symptoms include constant agitation and the incapacity to tolerate new stimuli. The inmate often begins to ruminate excessively on minor offenses or relatively meaningless events. Agitation and rumination can lead to violent bouts of anger and paranoia. Sadly, such afflictions can remain with a person long after confinement is over. In addition, thousands of those now languishing in solitary confinement were already known to be mentally ill. The psychological pressure of isolation can lead to what mental-health professionals call “decompensation”: mental illnesses reappear or worsen, and inmates are unable to perform everyday functions. This deterioration may give rise to violent conflicts with staff, and may require emergency psychiatric intervention.

In addition to psychological harm, solitary confinement also inflicts spiritual damage. A person in solitary finds his horizon drastically circumscribed. Cells are painted in ugly, drab colors. Some inmates don’t see the outdoors for years. Cell doors block outside stimuli, creating sensory deprivation. Lights that remain continually on disrupt an inmate’s diurnal rhythms and disorient his sense of time. Some inmates experience hallucinations and strange out-of-body experiences. One person I talked with reported seeing a dead relative in his cell. Gradually, some inmates lose any sense of an outside world and cannot imagine anything beyond their segregation unit.

The social deprivation of solitary confinement exacts a more subtle toll. Humans are social beings, and it is by relating to other people and gauging how they respond to us that we develop a narrative of our lives—our sense of self. But those in solitary receive no feedback from others and have difficulty judging the significance of their actions. They often lack clocks or calendars, and without a way to mark time, the events of their life become disconnected. Some inmates report having trouble recognizing their own faces in a mirror. They may withdraw completely or fall into depression, sleeping continually and refusing to leave their cells for exercise or showers. Others go mad, mutilate themselves, or engage in pointless battles against corrections staff. The suicide rate in many segregation units far exceeds that of the prison’s general population.

The damage solitary confinement does to prisoners is no accident. The dehumanizing conditions in which inmates are held—the lack of sensory stimulation and human contact; the petty control over inmates’ daily lives; the disorientation with regard to time; and the threat of indefinite isolation—are, in the minds of prison officials, essential to solitary’s power as a disciplinary tool. Contemporary solitary confinement is a policy designed to do harm to the men and women subjected to it.

This would seem like an obvious case of cruel and unusual punishment under the Eighth Amendment of the Constitution. But courts have been reluctant to draw that conclusion, and current law makes it difficult for inmates in solitary even to bring cases to court. The 1996 Prison Litigation Reform Act erected a host of barriers to those who mount legal challenges to prison conditions. In cases where inmates and their lawyers have succeeded in doing so, courts have intervened to halt physical brutality against solitary inmates. But they generally refuse to go beyond such judicial actions by declaring solitary confinement itself to be cruel and unusual punishment. And while prisoners in California recently drew outside attention with a hunger strike to protest their living conditions—some of the strikers had been in solitary for more than a decade—in the end they extracted only minor concessions from prison authorities.

In some cases, courts have stopped penal institutions from placing mentally ill inmates in solitary. In 2000, inmates and lawyers brought a class-action lawsuit against Boscobel, alleging cruel and unusual punishment. A federal court ruled in favor of the inmates and barred Boscobel from keeping minors and mentally ill offenders in isolation. (In response, the Wisconsin Department of Corrections reclassified Boscobel, erasing the supermax designation. Nevertheless, it continues to hold inmates in solitary.) In 2013, lawyers alleged that a mentally ill inmate in the federal supermax in Colorado ate his own feces, built sculptures out it, and smeared it on his cell walls. Authorities ignored him and placed sandbags around his cell in an attempt to contain the foul odor. That case is still pending. But while its outcome might lead to improved conditions for mentally ill inmates in Colorado, the federal government shows little sign of abandoning solitary confinement; a promised review of its policies has not deterred it from proceeding with plans to open another supermax prison in Illinois.

Though progress at the federal level has been limited, some states have recognized the horror of solitary confinement and have made substantial changes to their penal systems. Mississippi and Maine have both moved away from solitary confinement in their maximum-security prisons, and in 2013, Illinois closed its Tamms supermax prison, a house of horrors that profoundly harmed its occupants. Following the national attention garnered by its inmate hunger strike, California promised to alter its procedures for sending inmates to the SHU. Other states have taken smaller steps, sometimes prompted by lawsuits, to reduce the numbers of those in isolation. Recently, New York announced new regulations to ensure that juveniles in solitary remain in their cells for only nineteen hours—and receive more exercise outside their cells. New York prisons will no longer place pregnant women in isolation, will institute greater due process for segregation decisions, and will be prohibited from putting mentally ill inmates into segregation.

Such reforms represent small changes to an enormous nationwide system. Yet they demonstrate that institutional security doesn’t require the use of long-term solitary confinement. Yes, in rare cases—during a riot or outbreak of mass violence—prisons must put inmates in temporary isolation. But instances of this extreme punishment should be few and far between. Prisons can place inmates in single cells to protect them or prevent them from harming others. They can institute short-term lockdowns to restore prison order or deal with violent inmates. Scholars and activists around the world have proposed ways to maintain prison order without resorting to long-term solitary confinement—pointing, for example, to Europe, where most nations have nothing like the American system of solitary. In 2011, Juan E. Méndez, a United Nations expert on torture, called for “safeguards” and “guiding principles” to govern the use of solitary confinement in the United States. He acknowledged that the tactic could be legitimate in exceptional circumstances—for example, to protect an especially vulnerable prisoner—but recommended that inmates never be isolated for more than fifteen days at a time.

Contemporary solitary confinement aims at degrading the personality, breaking the will, and forcing an inmate to submit to institutional regulations. The policy may help contain some violent criminals and bring about temporary order. But we know little about its long-term efficacy. More important, we must recognize moral limits to how we treat human beings. Solitary confinement amounts to a systematic and intentional assault on human dignity. The United States, as the country with the world’s highest incarceration rate, should be particularly alert to the human dignity inmates retain, despite their crimes. Our penal system disregards that dignity in consigning tens of thousands of Americans to the hell of solitary confinement.