Quando ‘inocente’ significa uma prisão perpétua

O que acontece depois que um réu é inocentado por insanidade em um processo? Frequentemente a resposta é confinamento involuntário em um hospital psiquiátrico estadual – sem previsão de fim.

Nesta reportagem, o New York Times mostra como a estratégia de defesa de alegar insanidade pode ter consequências funestas que duram o resto da vida.

Na maioria dos estados norte-americanos, a não-condenação por insanidade significa a internação automática em um estabelecimento psiquiátrico. Na maioria dos casos, como em Nova York, não há limite de prazo para a internação. Naqueles em que há um prazo, como a Califórnia, os limites são baseados no tempo máximo de pena de prisão para o crime.

Leia mais no New York Times:

‘When ‘Not Guilty is a life Sentence’

In almost all states, N.G.R.I. means automatic commitment to a psychiatric facility. In most states, like New York, there is no limit to the duration of that commitment. In the states that do have limits, like California, the limits are based on the maximum prison sentence for the offense, a model that belies the idea of hospitalization as treatment rather than punishment.

Despite Ann’s determination to betray no emotion, a drop of sweat rolled down her temple as a guard painstakingly examined her lunch items. That Sunday morning, she had taken two buses, two trains and a shuttle to get from her home to the New York state psychiatric facility where her son is confined. Frustrated, she pushed back a little, but just a little, when the guard took away two sealed bottles of fruit-flavored water, a special treat that Ann had made an extra stop to buy. She watched as he held them up to examine them and concluded that they must contain caffeine — which is not allowed — because they did not read, “Does not contain caffeine.”
“They’re testing you,” she said to her son, James, after she was finally cleared, metal-detected and led upstairs to the visiting room, a spare, linoleum-floored space inside the hospital’s high-security building. Ann, who asked that her nickname and her son’s middle name be used to protect their privacy, usually comes to see James three times a week. Obstacles like these are routine. James, a middle-aged white man with thinning hair and a thickening waistline, listened to her complaints in a routine way, too, glancing up from the newspaper his mother had brought, the two of them sitting at a table, the same arch in their brown eyebrows, eating homemade coleslaw and sandwiches. They’ve been doing this a long time.
At some point in the next five hours — while the three of us ate lunch, and dessert, and later snacks between rounds of Bananagrams and Kings in the Corner — James said to me, “I shouldn’t have taken the plea.”
By the time of the arrest that would lead to James’s confinement here, he had already been hospitalized multiple times for threatening to kill himself. His problems stemmed, he said, from being sexually abused by his stepfather. But he had held down jobs — at a pizza shop, banquet setup in hotels — and after an intellectual disability was diagnosed, he and his mother say, he ran track and field in the Special Olympics, competing in Minnesota, Colorado, Germany and London. When I asked what charge, exactly, led to his arrest, he lowered his voice and said: “rape.” His mother added that a kidnapping count was “tacked on.”
In 1996, when James was 20, the police responded to a frantic 911 call near the house where he lived with his mother. At the scene, the officers found a woman bloodied and in distress. She said that James had lured her inside for a housekeeping interview — and that he’d been screaming when he started ripping her clothes off and beating her. The cops later picked him up at his grandmother’s house, a few miles away. At the police station, James signed a statement saying he understood his rights. He waived the right to representation. He signed a confession. (He and his mother now claim that the confession was coerced and that he is innocent.) When doctors subsequently evaluated him, they found him so unstable that they ruled him incompetent to stand trial. He was remanded to a hospital for several months, then sent back to jail, where he regressed again, then sent back to the hospital for several more months, stabilized once again, then sent back to jail, where in preparation for his trial, he was returned to the hospital to be evaluated for mental illness. Doctors diagnosed borderline-personality disorder, his mother says — which enabled him to plead “not responsible by reason of insanity.”
‘You can’t just punish someone for having mental illness. But that’s happening.’
James says that he understood the plea he took. In the abstract sense, he did. But the specifics of it were as mysterious to him and his family as they are to most people. Before he was arrested, James and his mother were set to move to Georgia, where they had relatives, and where Ann had friends and a job lined up. After his plea deal, Ann says, she “put everything on hold,” for what she thought would be a few years.

Instead, James, now in his 40s, has been in the hospital for almost two decades. This isn’t because he was sentenced to 20 years, or to 25. He was not sentenced at all; he is technically, legally, not responsible. The court believes beyond a reasonable doubt that he committed the act he was accused of, a prerequisite for the state to accept an insanity plea. The plea does not, however, prescribe or limit the duration of his stay. The laws that govern the practice of committing people who are acquitted because of mental illness dictate that they be hospitalized until they’re deemed safe to release to the public, no matter how long that takes.

James’s insanity acquittal placed him in an obscure, multibillion-dollar segment of domestic detention. According to a 2017 study conducted by the National Association of State Mental Health Program Directors, more than 10,000 mentally ill Americans who haven’t been convicted of a crime — people who have been found not guilty by reason of insanity or who have been arrested but found incompetent to stand trial — are involuntarily confined to psychiatric hospitals. Even a contributor to the study concedes that no one knows the exact number. While seemingly every conceivable data point in America’s prison system is meticulously compiled, not much is known about the confinement of “forensic” patients, people committed to psychiatric hospitals by the criminal-justice system. No federal agency is charged with monitoring them. No national registry or organization tracks how long they have been incarcerated or why.

In 1992, the Supreme Court ruled, in Foucha v. Louisiana, that a forensic patient must be both mentally ill and dangerous in order to be hospitalized against his will. But in practice, “states have ignored Foucha to a pretty substantial degree,” says W. Lawrence Fitch, a consultant to the National Association of State Mental Health Program Directors and former director of forensic services for Maryland’s Mental Hygiene Administration. “People are kept not because their dangerousness is because of mental illness. People stay in too long, and for the wrong reasons.”

Michael Bien, a lawyer who helped bring a successful lawsuit against the California prison system on behalf of prisoners with psychiatric illnesses, concurs. “Under constitutional law, they’re supposed to be incarcerated only if they’re getting treatment, and only if the treatment is likely to restore sanity,” he says. “You can’t just punish someone for having mental illness. But that’s happening.”

In the visiting room in New York that Sunday, as the hours went by, families came and went. Ann settled in. On James’s birthday, she brings a party: relatives, presents, a cake. And almost every week, on every visiting day, she and James try to make a life here together at the hospital — because it now seems possible that he could die there.

The insanity defense has been part of the American judicial system from its founding, carried over from our English forebears. British law has long reflected the moral sense that society has a duty not to punish people who can’t comprehend or control their crimes. But the insanity defense has always sat uneasily with the public, which tends to regard it as a means to escape justice. In the United States, such sentiments reached fever pitch in 1981, when a 25-year-old named John Hinckley Jr., hoping to win Jodie Foster’s heart, tried to assassinate President Reagan, wounding the president and three others, including James Brady, the White House press secretary, who was left permanently disabled. Hinckley was found not guilty by reason of insanity (N.G.R.I., as it is frequently abbreviated) and sent to St. Elizabeths Hospital in Washington. The country was outraged. Dan Quayle, then a senator from Indiana, called the verdict “decadent” and said the insanity defense “pampered criminals.” His Senate colleague Strom Thurmond equated it to a free ride.

In fact, despite its reputation as a “get out of jail free” card, the insanity defense has never been an easy way out — or easy to get. After a defendant is charged, the defendant, her lawyer or a judge can request evaluation by a psychiatrist. A defendant may be found incompetent to stand trial and committed for rehabilitation if she isn’t stable enough or intellectually capable of participating in the proceedings. If she is rehabilitated, she may be tried; if she cannot be, she may languish in a psychiatric hospital for years or decades. But mental illness is not exculpatory in itself: A defendant may be found mentally ill and still competent enough to stand trial. At that point, the district attorney may offer an insanity plea — some 90 percent of N.G.R.I. verdicts are plea deals. If the district attorney doesn’t offer a plea, or the defendant doesn’t take it, the case goes to trial. The defendant may still choose insanity as a defense, but then her case will be decided by a jury.

If N.G.R.I. was always difficult to get, it became even harder after Hinckley. With the Insanity Defense Reform Act of 1984, Congress restricted the judicial definition of “insanity” to only the most severe cases. Some states — Idaho, Utah, Kansas and Montana — have eliminated the defense altogether. In trials in which it is attempted, doctors may disagree, and jurors are often influenced by emotional considerations. Today, only an estimated one-120th of 1 percent of contested felony cases end in a successful N.G.R.I. defense — that is, the prosecutor disputes the insanity defense, the case goes to trial and the jury finds the defendant not guilty by reason of insanity. In addition, the legal standards for “insanity” vary among states; some define it as a defendant’s inability to know the crime was wrong or the inability to act in accordance with the law, but most define it, post-Hinckley, as only the first of these. At the trial of James Holmes, who killed 12 people and injured 70 in a movie theater in Aurora, Colo., one psychiatrist testified that he was mentally ill but that he knew right from wrong and should be considered “sane.” Another testified that he was mentally ill and incapable of reason (and, by extension, guilt). All four who examined him agreed that he had some form of schizophrenia. Jurors rejected his insanity plea.

And when an N.G.R.I. defense does succeed, it tends to resemble a conviction more than an acquittal. N.G.R.I. patients can wind up with longer, not shorter, periods of incarceration, as they are pulled into a mental-health system that can be harder to leave than prison. In 1983, the Supreme Court ruled, in Jones v. the United States, that it wasn’t a violation of due process to commit N.G.R.I. defendants automatically and indefinitely, for the safety of the public. (Michael Jones, who was a paranoid schizophrenic, had been hospitalized since 1975, after pleading N.G.R.I. to petty larceny for trying to steal a jacket.) In almost all states, N.G.R.I. means automatic commitment to a psychiatric facility. In most states, like New York, there is no limit to the duration of that commitment. In the states that do have limits, like California, the limits are based on the maximum prison sentence for the offense, a model that belies the idea of hospitalization as treatment rather than punishment. As Suzanna Gee, an attorney with Disability Rights California (a protection and advocacy agency with counterparts in every state), points out, the law allows two-year extensions as patients approach a “top date,” the limit set on their confinement. And so, she says, “it can be extended in perpetuity.”
James’s mother, Ann, now knows the predicament of forensic confinement well. At some point during James’s stay at the state hospital, she became an advocate for mentally ill offenders. “It’s like a roach motel,” she says. “At least in prison, inmates know they’re leaving. Once you check into the hospital, it’s hard to check out.”
Though forensic detentions get little attention, they can range from ethically questionable to flagrantly unconstitutional and illegal. In 1983, a national study found that N.G.R.I. patients often lost their freedom for twice as long as those actually convicted of the same offense. A study of N.G.R.I. patients in seven states between 1976 and 1985 found that in four of those states, they were confined for less time than people who were found guilty, and that in three, they were confined for longer. Scant research, conducted decades ago, seems to constitute the most recent survey of the fate of the country’s forensic
“There’s not been a lot done,” Fitch says. The federal government doesn’t collect data on forensic patients’ lengths of stay, crimes or treatment. In some cases, neither do the state or local departments in charge of their custody. In 2015, I began collecting, via request or the Freedom of Information Act, all individual length-of-stay data by legal status that existed in each state and Washington. Colorado, Wyoming, Arkansas, Missouri, California, Maine, New Hampshire, Kentucky, Wisconsin, Delaware, New Jersey, Ohio and South Carolina said they simply didn’t have that information. Alabama may or may not: In response to repeated queries, it “decided not to release forensic data,” and hospital reports are excluded from its public-records law.
Many of the above states have reported legal status and average lengths of stay. In 2014, Fitch, on behalf of the National Association of State Mental Health Program Directors, estimated, based on states’ self-reported average lengths of stay, that the national average for all N.G.R.I.s was around five to seven years. He says he finds that “horrendous,” given that civil commitments with the same diagnoses as forensic commitments can get out in under 30 days. There is no accepted body of research to suggest that lengthy institutionalization leads to better treatment outcomes. On the contrary, says Marthagem Whitlock, an assistant commissioner in Tennessee’s Department of Mental Health and Substance Abuse Services, “The deeper penetration into the system usually means more complications for the individual.” There are, as Fitch acknowledges, “people who don’t respond to treatment or who refuse treatment.” But, he argues, “it should almost never be necessary to hospitalize people that long.”
According to the state records collected for this article, in 2015, Florida had 24 N.G.R.I. patients who had been hospitalized for longer than 15 years. Texas had 27. Connecticut had 40, and Georgia had 43. New York and Washington had around 60 apiece. That’s six of the 28 states from which such data can be extrapolated, along with the District of Columbia. In those states — which exclude thousands of N.G.R.I. patients — a significant portion of N.G.R.I. patients had been hospitalized more than two years. Nearly 1,000 had been hospitalized for five to 15 years. More than 400 had been in for longer than 15. Of these, more than 100 had been in longer than 25 years and at least 60 for more than 30. And those numbers don’t present the whole picture, either: Many factors, like hospital transfers or conversion to civil commitment, can start the clock over or obscure patients’ histories. In many cases, Gee says, when patients reach out to Disability Rights California to advocate on their behalf, “if they’d not pled N.G.R.I. and just gone to prison, they might have gotten out earlier.”

Which is not to say there aren’t protocols for release. Most states do have a formal review process to judge whether N.G.R.I.s no longer fit commitment criteria: They are no longer mentally ill or are no longer dangerous as a result of their mental illness. Some states review cases on a schedule — every year, say, or in the case of New Hampshire, every five years. In others, patients (or their lawyers) have to request the review. Doctors can recommend patients for discharge at any time. Patients’ lawyers can also file writs of habeas corpus or petitions for restoration of sanity to have their cases heard in court.

At the psychiatric facility where James is a patient, as at every New York state hospital, cases are reviewed every two years or so, in a joint process by the hospital and the Office of Mental Health (O.M.H.) in Albany. In 2002, when the hospital and O.M.H. review declared James unfit for release, James requested that the court appoint him an independent evaluator and grant him a hearing.

That doctor found that he wasn’t dangerous and was ready to be transferred. The judge agreed and ordered it.

But James didn’t leave.

“Even the mechanisms for getting out,” says Pat McConahay, communications director for Disability Rights California, “are not really mechanisms for getting out.”

James’s almost-transfer 15 years ago fell under the jurisdiction of Guy Arcidiacono, now 61, the district attorney in charge of the Suffolk County Forensic Psychiatric Litigation Unit, who has handled 130 N.G.R.I. cases over 25 years. After the judge approved James for transfer, Arcidiacono and the Suffolk County district attorney’s office, joined by the New York State attorney general’s office, which represents the hospital, appealed to have the transfer order overturned. Under another judge, it was.

James’s police files are terrible to read. There was never a trial, but the allegations are disturbing. They include a statement by the woman whom the police responded to near his house saying that James sexually assaulted her, beat her unconscious, then threw her down the basement stairs. She told the police that when she came to, naked, she realized she was locked in. This, you might presume, was the basis for the rape and kidnapping charges that James said put him in the state hospital.

“Well,” Arcidiacono clarified when I contacted him, “there were two cases, actually.”

Two months earlier, James had been arrested based on another woman’s statement, taken at the emergency room, that she was pinned by her throat and raped in an empty field. (In his statement, James said she had agreed to have sex with him in exchange for crack, and that he didn’t actually intend to make good on his promise.) Charges were filed, and James was released to his mother’s custody — Arcidiacono says there’s no record of why, but it could have been because James hadn’t been indicted yet. When James pleaded N.G.R.I. to the subsequent attack, the charges from the first incident were lumped together with the new ones: in total, second-degree kidnapping, second-degree assault, second-degree aggravated sexual abuse, first-degree sexual abuse, first-degree rape and third-degree robbery.

“You can see in his case,” Arcidiacono said, “that they were serious charges.”

Like most district attorneys, Arcidiacono has substantial discretion when an N.G.R.I. patient comes up for review. Any time a hospital wants to release or transfer an N.G.R.I. patient, his office can demand a hearing. In contesting transfer or release, Arcidiacono can compel the patient to be examined by a doctor he selects. Even if that doctor agrees that the patient is ready to leave, the district attorney can still contest the release. Arcidiacono has done this and won. If a patient does win release or transfer (in New York, a forensic patient is almost always transferred to a civil, less secure facility for another unspecified amount of time before release), the district attorney can appeal, as in James’s case. When you’re dealing not with facts but with opinions, Arcidiacono told me, “reasonable people can differ.”

He went on to explain: “There are two considerations here. What is good for the defendant, and also, the safety of the public. And that’s the difficult part of these cases, balancing those two interests.”

‘Everybody except for people who take the Constitution seriously and people who are in the hospital are happy the patients are there.’
All this adds up to a difficult path to freedom for some N.G.R.I.s. In most states, as in New York, the courts have final review over forensic releases and transfers, and judges have the prerogative to side with the prosecution regardless of what doctors advise. Cas Shearin, director of investigations and monitoring at Disability Rights North Carolina, recalls a 1988 case in which, during an alcohol-related psychotic break, a man shot four strangers he thought were demons. “Year after year after year, his treating doctors went to the judge and said, ‘He’s ready to be released.’ ” The hospital where he was committed gave him increasing leave privileges, including reporting to a full-time job and visiting a girlfriend with whom he had children. After 21 years of incarceration and seven psychologists and psychiatrists testifying that he was no longer mentally ill and posed no threat to the public, he was released — though the judge still ordered that he submit to random drug tests for a year.

Politics, says Joel Dvoskin, a former New York State Office of Mental Health forensic director, can determine if “you’re going to stay locked up for a really long time, regardless of whether it’s safe to let you go.” Elected judges, fearing bad publicity, may be loath to release an offender into the community. As Ira Burnim, legal director of the Bazelon Center for Mental Health Law, a national advocacy organization based in Washington, explains the situation: “You have a mechanism to confine, for the protection of the public, these individuals when they’re mentally ill and dangerous, and the further you stray from that, the less it’s legally justified.”

John Hinckley Jr. became a famous example. Last September, Hinckley was released from St. Elizabeths Hospital in Washington, 35 years after being found not guilty. It had been two decades since his doctors declared his mental illnesses in full remission, which should have provided the basis for his release. “That anyone can justify keeping him in the hospital” for so long, Fitch says, “is just nuts.”

The question, according to Dvoskin, “becomes one of risk tolerance. America has become — to an extreme level that’s almost impossible to exaggerate — a risk-intolerant society.” Fears of people with mental illness persist, even though, according to the best estimates, only 4 percent of violent acts in the United States are uniquely attributable to serious mental illness. One study has found that those with mental illness are actually less likely to be seriously violent than the general population. (In addition, some N.G.R.I.s have been acquitted of nonviolent crimes, like public-order offenses, traffic offenses and prostitution.) Even if a mentally ill person has committed a crime, says Chris Slobogin, director of the criminal-justice program at Vanderbilt University Law School, “it doesn’t mean they’re going to do it again,” especially because their encounter with the forensic psychiatric system means they’ve received treatment. “This is a group of people that are incredibly stigmatized and misunderstood in terms of how dangerous they are.”

Recidivism for N.G.R.I.s is relatively low. Whereas, nationally, recidivism for released prisoners is above 60 percent, “people who are found N.G.R.I. tend to go back out into the community, and they tend to do really, really well,” Fitch says. The arrest rate for people in Maryland on conditional release, a kind of mental-health parole from the hospital, is less than half the arrest rate of the general population in the state. “If you provide treatment of illness and provide the supports they need, then they don’t reoffend,” Fitch says. As a 2016 study of N.G.R.I. recidivism in Connecticut — which has a post-release supervision program, too — also concluded: “The vast majority of individuals are not rearrested.”

It is not sober data analysis, however, that sticks in the public’s mind — or influences judges’ rulings. Hearing that a man in Nebraska whose most recent diagnosis was “cannabis abuse, unspecified” had been in the hospital for 37 years may evoke less sympathy or outrage when you learn that he killed six people, three of them children. Two of his victims he raped. One of them was dead when he did it. The other one, who was alive for the assault, was 10.

“It’s not an easy population to represent,” Bien says. “No one likes these people.”

On Nov. 21, 2011, after months of having delusions about aliens, conspiracies and poison, Houston Herczog, then 21, partly decapitated his father, Mark. Six days earlier, Mark Herczog had written a letter to God (“God — Help!”) asking him to save his newly unrecognizable son. After Houston was arrested in the bloodied family kitchen, he told the police: “The look in his eyes! I had to!”

I heard about the murder from my mother. Houston is my distant cousin, though I had never met him or his immediate family. During his trial, as psychiatrists testified about a psychotic break related to an onset of schizophrenia, his family — my family — prayed that he would be found not guilty by reason of insanity and sent somewhere for treatment. Charged with first-degree murder, Houston was facing 25 years to life. When he was found N.G.R.I. and sent to nearby Napa State Hospital, it was a relief.

It wasn’t until he was inside that any of us realized, from the mounting anxiety in his phone calls, the rules and restrictions that governed the possibility of his getting out.

Houston has now been in the hospital for four years. In 2015, I visited him there for the first time. After passing through a prison fence and four locked doors, I sat with him in an all-beige room. Schizoaffective disorder is the current diagnosis, and he is stable on medication, though he has periods of deep hopelessness. He struggles with what he has done; at the beginning of his incarceration, when, pharmacologically stabilized for the first time, he was suddenly lucid, he couldn’t do anything but lie in bed and cry. But he also struggles with the uncertainty of how long his confinement will last. Had he been convicted and received the minimum sentence, he would be out before his 50th birthday. His “top date,” according to his hospital paperwork, is Dec. 31, 2600 — 587 years after he was admitted.

Houston doesn’t dispute that he did what he did, but he does dispute the basis for his continued detention. “What are they rehabilitating?” he asked me that day, shaking his legs up and down when the caffeine from a Mountain Dew Code Red kicked in, his blue eyes wide behind wire glasses. He held his hand up, ticking the points off on his fingers. “Not a pattern of violent behavior, since I have no record of violent behavior before my crime. Is it my insanity? Because the treatment for my disease is medication, and I’m medicated and stable now.” And he is — high-functioning as you please. Before we began talking about his incarceration, he kept trying to engage me in a debate about feminist theory.

When he calls, Houston sometimes apologizes for not having much to say about his life — what would he have to say? — in limbo as it is. He would love to have a girlfriend, his first, but it seems unlikely to happen at the hospital. (“You know, I came here for the women,” he joked once.) He spends a portion of his unnumbered, many-numbered days arguing with staff about how many packets of sweetener he’s allowed to use at a time. (Patients are restricted to two at lunch; he likes his tea quite sweet.) His schedule consists of shuffling from group activity to group activity in a beige uniform, between feedings of institutional food three times a day, mind-numbing TV in the background, no internet or cellphones.


Houston Herczog with his father, Mark.
Credit From the family of Houston Herczog
Houston’s plan is to wait until he has been hospitalized for at least five years before he bothers to file any writs or petitions for release. That number has nothing to do with what he thinks of his mental state but with comments his first social worker at Napa made: that because of what he did, he can’t possibly be let go for a long time. (When reached for comment about whether this was a plausible conversation, a spokesman for the hospital said, “A patient with a determinate sentence length of life for a high-profile crime might end up staying in the hospital for many years. It is reasonable and therapeutic for a treatment team to discuss these realities with patients.”)

Houston’s sister Savannah says that she gets it — the urge to confine forever someone who did something horrific. Savannah was home the night her brother killed their father. Only 17 at the time, she “had to walk past them to get the phone” to call 911. When she ran to her room to dial, Houston followed her, still holding a knife, but stopped when she shut the door on him. “Part of me thinks he should still have to pay in some way, whether he was in his right mind or not,” she says. “It’s hard though, too, because it is ‘not guilty.’ I have to remind myself of that a lot. I think that’s because I was there. I saw it happen.” But, she went on, “it’s my brother, and obviously I don’t want him to spend his whole life there, because even seeing what he has to eat makes me want to cry. Being there just deteriorates him more and breaks his spirit. He realizes what he did and where he is, and he’s depressed.”

Not everyone in the family is so sympathetic, says Houston’s half sister, Cameron McDowell, 43. “We have some family members who just hate him and will never forgive him. It was just such an awful thing. And I wish Dad were here every day. I can’t even imagine what he went through that night — oh, God, it was so awful.” But, she says, “he’s gone. And we have Houston now. We have to support him. This is going to sound strange, but I’ve not once been mad at him. I really, truly, passionately believe that it’s not the person that commits the crime. It’s the illness.”

At the same time, she understands why people are afraid. Cameron’s own husband isn’t yet comfortable with the idea of Houston’s hanging around their two young kids; though he loves Houston, he’s “a little bit freaked.” “I hope my husband will change his mind when Houston gets out,” she says. She, too, worries about what will happen after his release, though for entirely different reasons. “Is he going to be so institutionalized that he won’t know how to live? That’s what breaks my heart for him.”

Like Houston’s sisters, the judicial and medical systems struggle to find a balance between the blamelessness of N.G.R.I. patients and the gravity of many of their crimes. The rights of the patient are always weighed against the public good, a standard that may include a more or less explicit desire for retribution. Those who provide treatment for forensic patients, says Michael Norko, professor of psychiatry at Yale and director of forensic services for the Connecticut Department of Mental Health and Addiction Services, “still have to answer to a court, or to the board, or to the court of public opinion. Every facility has people in it who have so violated a community, a community that is so angry, so hurt, that they’re basically pariahs.” He recalls one patient who shot a police officer; at hearings for his release, a crowd of uniformed cops repeatedly showed up and stood silently, facing the review panel. The question of a patient’s hospitalization can be “reduced to people’s grief, people’s anger, people’s fear — and the complexity of the patient’s rights and their recovery, all of those complexities get overshadowed very quickly.” When I asked Norko what happened to the patient who shot the officer, he said, “Well — well, eventually, he died.”

Because diagnoses and treatment assessments cannot predict future behavior, the standards for involuntary confinement — degree of mental illness and dangerousness — are necessarily subjective. Emotions and prejudice easily come into play, even from experts. A 2003 study in The American Journal of Forensic Psychology, for example, showed that doctors are more likely to find minorities incompetent to stand trial and more likely to diagnose psychotic disorders in African-Americans. At Napa, I spoke with a patient, a friend of Houston’s, who pleaded N.G.R.I. to a murder charge and had been hospitalized for nearly 20 years. When he was granted a hearing in which, according to court transcripts, multiple clinicians recommended his release, one doctor dissented — a doctor with whom he had had an ugly dispute and who, another doctor testified, wasn’t objective. The patient’s release was denied.

Stephen Seager, a 67-year-old psychiatrist who was at Napa for five years, writes openly about his own reactions to his patients in his 2014 memoir, “Behind the Gates of Gomorrah: A Year With the Criminally Insane.” He describes being in the hallway with a group of them heading to lunch as being “engulfed in a wave of hungry psychopaths.” When one of his patients tells him a story about his childhood, he writes, “I didn’t like thinking that some of the men even had childhoods.” Asked in court about a patient’s diagnosis, he gave the admitting diagnosis — bipolar disorder, manic with psychotic features — though he writes that he knew he wasn’t currently mentally ill. But he did think he was dangerous. When I asked him about this, he said: “The point is they’re supposed to be dangerous because they’re mentally ill, but if they get better and they’re still dangerous, what do you do?” He is well aware of the import of expert testimony in retention decisions. “Most of the time,” he said, “judges take our opinion on it.

“I look for the safety of the community,” he went on. “I live here. Sometimes you just have to say something for everybody’s best interest, regardless of whether they’re mentally ill or not.”

Of patients who “just never quite get better,” in doctors’ estimations, Seager said: “Oh, they’ll be here till they die.”

Napa State hospital’s vast, drought-dry campus is roamed by a pack of screeching peacocks. At 1,255 beds, it is one of the largest state psychiatric hospitals in the country. For most of his time there, Houston has been on what’s called a discharge unit, with one to three roommates, where it’s possible to get privileges like walking to the visitors’ center alone. After a bout of suicidal thoughts, he was moved, indefinitely, to a locked unit where all activities and access to the courtyard are supervised.

Houston’s treatment consists of up to 20 hours of group classes a week. His schedule at one point included Emotional Management, Substance Recovery, Current Issues in Mental Health, Self-Esteem, Fitness/Easy Exercise, Leisure Skills/Computer, Fitness/Weight Lifting, Discharge Planning, Wellness Recovery Action Plan, Coping Skills/Fitness, Leisure Skills/Journal and Conditional Release Prep. He says they watch a lot of videos. I once talked to a fellow patient of Houston’s on the phone who said he was heading to a class where the day’s lesson was learning to make cheesecake.

“Cheesecake!” he said. “Imagine that.”

State forensic hospitals vary widely in size — anywhere from just one ward to 1,500 beds — and they also vary in the activities and treatments they offer. Public psychiatric hospitals across the country may be accredited by an independent nonprofit called the Joint Commission. But it doesn’t require, for a start, the use of evidence-based therapies. One-on-one psychotherapy can be hard to get. (Houston began receiving such therapy only after his mother spent two years asking for it, his family says. Napa declined to comment, citing privacy laws.) State mental-health agencies may not be required to report what kind of care hospitals are giving, unless something goes so wrong that it attracts the attention of federal authorities. (Many do, however, voluntarily participate in national data-collection systems.)

Newsletter Sign UpContinue reading the main story
Get the Morning Briefing by Email
What you need to know to start your day, delivered to your inbox Monday through Friday.


Enter your email address
Sign Up

You agree to receive occasional updates and special offers for The New York Times’s products and services.

The American archetype of bad mental-health practices is the movie “One Flew Over the Cuckoo’s Nest”: doped-up patients locked away from society to suffer and, ultimately, to perish. Some advocates argue that the conditions at psychiatric institutions aren’t always much better. Until recently, Oregon State Hospital, where “Cuckoo’s Nest” was filmed, may in ways have been worse.

In 2006, The Oregonian won a Pulitzer Prize for its exposé of the institution. It reported that there was a male staff member who worked on one ward for only a year in the late ’80s but still managed to sexually abuse six female patients; there was a wrongful-death suit in 2003; a rape lawsuit in 1995 that accused the hospital of what The Oregonian described as a “longstanding pattern of sex abuse,” some of it by a staff member who, the paper reported, had nearly been fired the year before, “after he and several other staff asphyxiated a patient while restraining him for refusing to take off his shoe.”

The problems weren’t unique to Oregon: Nebraska state psychiatric hospitals settled two major lawsuits, one in 1996 and another in 2006, which claimed that staff members and patients raped and sexually assaulted female patients; two lawsuits have been brought against Massachusetts in the last 10 years by families of forensic patients who were subjected to illegal restraining methods, one whose condition declined and another who died. Napa had more than 4,000 reported patient-on-patient or patient-on-staff assaults in 2014; in 2010, a staff member was killed.

Today Oregon State Hospital still has the towering 1883 brick face of the “Cuckoo” era, but inside, it’s all sparkling new facilities, yoga props and a physical-therapy pool — thanks, in part, to a state senator, Peter Courtney. On a 2004 tour of the hospital, Courtney was astonished to learn of 3,600 badly corroded cans of cremains in an outbuilding, still unclaimed. Around the same time, after the death of yet another patient, the Justice Department became involved. Oregon State is now the first state psychiatric hospital in the nation to use “collaborative problem solving,” instead of, say, automatic “seclusion and restraint” of patients who start to become difficult. Courtney pushed funding for the reforms through Oregon’s State Legislature and brought in a new superintendent, Greg Roberts.

“Very often when you find serious problems, the problem is one of three things: leadership, leadership or leadership,” Roberts says. (A notable problem of leadership was found at Napa: The man who was its executive director between 2007 and 2010, Claude Edward Foulk, is now serving 248 years in prison for sexually molesting children.) Before Roberts came to Oregon, he was the person New Jersey sent to take over “problematic” hospitals that had “a serious negative incident or series of negative incidents.” He revamped three there, one of them twice.

More than 10 years into its transformation, Oregon State has come a long way. But, Roberts acknowledges, it still has a long way to go. Among the hospital’s 937 forensic patients, there were 1,908 incidents of seclusion and restraint in 2014. And while Roberts’s reforms included securing timely releases for patients — between 2012 and 2015, the hospital discharged 90 percent of its “guilty except insanity” (Oregon’s N.G.R.I.) patients with “top dates” early — it still has several patients who’ve been in for more than 15 years, who it acknowledges long ago ceased to be dangerous. Staff recently discharged one patient who had been in for more than 30.

Unfortunately, there aren’t a lot of other places for forensic patients to go. When the push to deinstitutionalize psychiatric patients began in the 1970s, many inpatient facilities were shut down or downsized in favor of community integration and outpatient services, which never adequately materialized because of a lack of funding. But institutions have remained the preferred repository for forensic patients, who fill an increasing share of the remaining 42,000 state psychiatric hospital beds. In Pennsylvania, the proportion of forensic patients increased by 379 percent between 1988 and 2008. In California, some 90 percent of approximately 7,000 state-hospital patients are now forensic. Nationwide, nearly one-third of “consumers” in state hospitals in 2007 were forensic, and that number is “rapidly expanding,” according to the federal Substance Abuse and Mental Health Services Administration (Samhsa). In the last 10 years, forensic patients’ costs jumped to $4.25 billion from $2.5 billion; forensic patients now account for 43.9 percent of total state psychiatric hospital expenditures. This means less room for others who might be seeking psychiatric care, while at the same time, states have cut billions from mental health, leaving even scantier outpatient systems. In Hawaii, as a 2015 Samhsa report notes, “There is no voluntary admission to the [state psychiatric] facility because of the volume of admissions from the court system.” The same lack of services that contributes to people’s reaching a mental-health crisis that ends in arrest also keeps people institutionalized longer afterward.

Whitlock, of Tennessee’s mental-health department, says that there are N.G.R.I. patients whom her state’s hospitals are trying to discharge but for whom they “just can’t find a provider.” Under her lead, Tennessee, more than any other state, has been trying not to commit forensic patients to begin with. Since 1974, it has done its pretrial competency evaluations on an outpatient basis, no insignificant matter given that elsewhere these can take six months or longer. In 2009, the state also started doing post-N.G.R.I. evaluations to see if N.G.R.I.s even needed hospitalization, also on an outpatient basis. “We did not think there was a need to spend 60 or 90 days in the hospital to determine if N.G.R.I.s fit the involuntary-commitment standard,” Whitlock says. “Once you get someone into the hospital, it’s hard to get the court to take them back out. We could probably determine it in a day.” Now only 55 percent of Tennessee’s N.G.R.I.s are committed. Their typical length of stay ranges between seven months and 4.5 years, and they get out, on average, in about two. The state’s new initiative adds to the body of evidence that less hospitalization doesn’t lead to higher crime rates. Since Tennessee stopped automatically committing N.G.R.I.s, says Jeff Feix, the state’s director of forensic services, “the recidivism rate we have is no different than it was before.”

Despite the clinical benefits and cost savings — in 2015, according to a Samhsa report, the average annual cost of one forensic patient, nationwide, was $341,614 — Tennessee’s model is still unusual. There is no outcry, from the public or politicians, for alternatives to indefinitely institutionalizing N.G.R.I.s. After 45 years of studying the issue and filing lawsuits on behalf of patients, Michael Perlin, an emeritus professor at New York Law School and an expert on mental-disability law, thinks he knows why: “Everybody except for people who take the Constitution seriously and people who are in the hospital are happy the patients are there. Prosecutors, police, they’re glad they’re not going anywhere. I believe that the disability rights community has never gotten substantially involved in the issue because some of the people have been charged with very horrific crimes.”

As he put it: “This is an area that everybody kind of wishes would go away.”

Nearly two decades into James’s confinement, it remains unclear if he will ever be considered fit for release. He says that currently his entire medication regimen consists of “fish oil twice a day, calcium, vitamin D and two Kool-Aids and prune juice and Metamucil.” In the 2004 appellate court overturning of his transfer, the hospital doctor testifying in favor of his continued retention pointed out that there was no medication for his diagnosis of antisocial personality disorder — just therapy, to which he was resistant. The hospital forensic committee said that they believed James was insufficiently willing to take responsibility for his actions. One of his doctors called him “sexually preoccupied”; the appellate court concluded that “the disorder will continue to cause him to be dangerous at least until such time as he decides he wants to change and begins working seriously with his treatment providers.”

The court-appointed psychiatrist at the hearing two years earlier, when James was ordered for transfer, said quite the opposite. In his opinion, James “acknowledged the wrongfulness of his actions, and although he does not show much remorse or regret for his actions as a result of limited insight, he certainly realizes that what he did was wrong and against the law.” He said that James “has benefited from [the state psychiatric facility] as much as he will benefit, and at this point he could be transferred to a civil hospital, where he will continue to benefit from the structured environment and continue to receive individual treatment, hoping that he will eventually gain further insight.” But in its appeal at that time, as in every review since, the hospital successfully petitioned to retain him.

How much James’s perceived dangerousness is due to his illness and how much to his extended hospitalization can be difficult to untangle. During a phone call last year, he confessed that lately he’d been “going through a lot of crap.” He was referring to recent fights on the ward. The 2004 appeal notes that he was assaulted twice in 2002 and “got into an altercation with another patient” that year; all of the incidents read as demerits against him. Such disputes are not uncommon for patients with personality disorders, says Norko, the Yale professor and clinician. A hospital is generally “not a good place for them,” he says. “Clinically you try very hard not to hospitalize people who have personality disorders. They can’t quite seem to ever get it right. To them it looks like: ‘The staff want me to do this, they want me to do that, they keep changing the rules,’ and they don’t understand the rules, and they get into arguments, they get into fights. It’s all part of their personality disorder, which in itself isn’t necessarily all that dangerous, but it’s kind of hard to move somebody along when their record is dotted with all of these altercations.”

Perhaps the most cleareyed view of the compromises inherent in N.G.R.I. commitments comes from Paul Appelbaum, professor and director of the division of law, ethics and psychiatry at Columbia University. Appelbaum acknowledges that some N.G.R.I.s are “unnecessarily detained for a longer period than what seems to be warranted by their mental disorder and its impact on their likelihood of being violent in the future.” But, he says, such exaggerated concerns about public safety may be necessary to the survival of the insanity defense. “There are injustices that are imposed on individuals,” Appelbaum says. “But I also see at a 30,000-foot level why the system works that way, and recognize perhaps the paradox that if it didn’t work that way, we might lose the insanity defense altogether, or at the very least have an even more restrictive system that we have to deal with.”

For many, Appelbaum says, an N.G.R.I. verdict is still superior to a conviction. “It exempts them from a formal finding of guilt, which can be important later in their lives. It enables them to serve their time of confinement in what is generally a much better and safer environment than an overcrowded state prison.” While one way to look at indefinite confinement is as an unbearable uncertainty, another is that it offers hope: “Compared to a life without the possibility of parole, you know, maybe that’s better.”

Yet for the insanity defense to live up to the moral imperative it was designed to embody — exculpating those with diminished responsibility for their acts — better mechanisms for evaluating release will need to be adopted, Slobogin says. In an attempt, in part, to protect N.G.R.I.s’ constitutional rights to be released when they don’t fit commitment criteria, the American Bar Association recently revised its criminal-justice mental-health standards, which were adopted in 1986. The association recommended that forensic patients be detained only if there is clear and convincing evidence — a standard that, if it had to be quantified, means about 75 percent certainty — that they’re mentally ill and dangerous; the current legal burden of proof in most jurisdictions is a “preponderance of the evidence,” or 51 percent certainty.

“It’s immoral to deprive someone of liberty because you’re mad at them for being found N.G.R.I.,” says Slobogin, who was on the task force. “Under our new standards, after a year you have to have very strong proof of dangerousness, or you can’t detain them.” But the standards are just recommendations, and some states didn’t even follow the ones from the ’80s. Slobogin concedes that in many jurisdictions, they may have no impact at all.

Mac McClelland is a three-time National Magazine Award finalist and the author of “Irritable Hearts: A PTSD Love Story.” She last wrote for the magazine about the pop star Azis.

Sign up for our newsletter to get the best of The New York Times Magazine delivered to your inbox every week.

A version of this article appears in print on October 1, 2017, on Page MM34 of the Sunday Magazine with the headline: ‘They’ll Be Here Till They Die’. Today’s Paper|Subscribe