Picture a person in a hospital hallway, told that a hearing will decide whether they may leave. A doctor gives an opinion. A lawyer, if one is present, speaks briefly. A judge asks a few questions and rules. The event may last minutes, yet it can decide months of a life.
That scene sits at the center of a long argument about civil commitment, the legal power of a state to hold people in psychiatric facilities against their will. When medicine and law meet in that hallway, is anyone giving informed consent, and when consent is set aside, is the process fair? This article follows Thomas Szasz in putting liberty and consent first, and it gives the strongest opposing views too.
How American Law Mixes Medicine and Liberty
American law treats civil commitment as a deprivation of liberty that needs justification and procedure. The landmark cases form a running negotiation between medical judgment and constitutional limits. In Lessard v. Schmidt (E.D. Wis. 1972), a federal court held that Wisconsin had to prove beyond a reasonable doubt that a person was both mentally ill and dangerous. It also required notice, counsel, and a search for less restrictive alternatives. Three years later, in O’Connor v. Donaldson (1975), the Supreme Court held that a state cannot confine, without more, a nondangerous person who can survive safely in freedom alone or with willing family or friends.
Then Addington v. Texas (1979) set the burden at clear and convincing evidence, below the criminal standard. The Court reasoned that, given the uncertainties of psychiatric diagnosis, proof beyond a reasonable doubt might impose a burden the state could not meet. That compromise still shapes most commitment hearings: the state must clear more than a coin flip, yet need not meet the criminal bar.
Nearby holdings tighten or loosen the same rope. Jackson v. Indiana (1972) held that a state may not hold a defendant indefinitely solely because the defendant is incompetent to stand trial. Foucha v. Louisiana (1992) held that an insanity acquittee who is no longer mentally ill cannot be kept in a psychiatric facility on dangerousness alone. Kansas v. Hendricks (1997) offers a contrast: by a 5 to 4 vote, the Court upheld civil commitment of sexually violent predators after prison, finding the law civil rather than punitive. Critics argue that a civil label can carry the effect of punishment. Supporters answer that the law ties confinement to mental abnormality and impaired control.
On the ground, commitment standards generally rest on danger to self or others, and many states add grave disability, meaning an inability to provide for basic needs. Definitions, detention periods, and review rules differ by state. A Duke Law Journal student note argues that some people wait weeks for a hearing, and that a few states provide none. That is one author making an argument, and current statutes should be checked. The variation itself is the point.
Szasz, Consent, and the Measure of Care
Szasz was a psychiatrist who spent decades criticizing his own field. In The Myth of Mental Illness (1961), he argued that mental illness is a metaphor for conduct judged unacceptable, not a disease in the way a broken bone is. In Law, Liberty, and Psychiatry (1963), he extended the argument to law. He held that involuntary hospitalization is a form of imprisonment, that physicians who take part become agents of the state, and that courts defer to medical opinion as they would not defer to other claims against liberty.
Critics reject parts of this. Many psychiatrists argue that severe conditions are real and that Szasz understated how impaired judgment can become. E. Fuller Torrey, for one, disputes the claim that severe illness is not brain disease. The disagreement is the hinge on which commitment turns: whether the problem is best framed as disease, as social conflict, or as what Szasz called problems in living.
The thesis of this essay draws on that tradition. Informed consent is real only when a person hears the actual risks, the limits of the evidence, the alternatives, and the right to say no. If any element is missing, the consent is partial. Under commitment, the right to say no is removed by design. Szasz held that coercion is not care. The opposing view holds that a narrow, reviewed exception can protect people who cannot protect themselves.
Where Hearings and Disclosure Fall Short
Legal standards set a ceiling. Practice often sits lower. Eric Turkheimer and Charles Parry, in a 1992 American Psychologist paper titled Why the Gap?, reported that hearings are rarely adversarial and that a median of 78 percent of respondents across seven studies were committed on grave disability grounds, with or without a finding of danger. They cited Carol Warren, who found grave disability alone grew from 11 percent of emergency commitments to 52 percent at later California hearings. In their Virginia study of 190 initial and 184 recommitment hearings, 70 percent of initial and 98 percent of recommitment hearings ended in commitment.
Michael Perlin has argued that courts defer to experts and lean on ordinary common sense, and that predictions of dangerousness are unreliable. In 2018 he reported that fewer than one third of judges told patients at initial hearings about the right to counsel, and fewer than one quarter mentioned voluntary status. He also wrote that the Supreme Court once noted average hearing lengths of 3.8 to 9.2 minutes. These studies are older, and practice varies, yet the pattern is hard to ignore: a constitutional form that often behaves like a medical rubber stamp.
Disclosure of risks and alternatives shows a similar gap. A 2005 Psychiatric Bulletin survey of 30 long-term inpatients found that half reported receiving no information about their antipsychotic medication, 90 percent said they had no choice of drug, and no chart documented a discussion of benefits and side effects. A 2022 Journal of Mental Health online survey of 757 people who had taken antipsychotics reported that about 70 percent were not told about adverse effects and 28 percent recalled being offered alternatives. Self-selected samples cannot estimate national rates, and recall can differ from what was said, but the surveys still ask a plain question: if consent is the ethical measure, how often does that conversation actually happen?
Capacity Debates and International Dissent
Consent depends on capacity, which researchers have studied directly. The MacArthur Treatment Competence Study, led by Paul Appelbaum and Thomas Grisso, compared hospitalized patients with schizophrenia, depression, and heart disease. A British Journal of Psychiatry systematic review reports that 52 percent of patients with schizophrenia were impaired on the capacity measures, against 24 percent of those with depression. Many people kept capacity despite a diagnosis, which supports the view that a diagnosis alone should decide nothing. Impairment was real for others, which supports careful assessment. A 1995 American Journal of Psychiatry paper found that the legal standard chosen changes who is classed as impaired. Capacity, in other words, is partly a measurement choice, not only a clinical fact waiting to be discovered.
International bodies have taken positions that diverge from American commitment law. The UN Committee on the Rights of Persons with Disabilities, in General Comment No. 1 (2014) on Article 12 of the Convention, argued that states must abolish substitute decision-making and develop supported decision-making. In report A/HRC/35/21 (2017), Special Rapporteur Dainius Pūras argued that mental health care leans too heavily on a biomedical model and on non-consensual measures. Other commentators reply that supported decision-making may be hard to apply in acute episodes.
Alternatives that keep decision-making closer to the person have been studied as well. A 2006 Finnish Open Dialogue study by Jaakko Seikkula and colleagues, a nonrandomized comparison of 42 and 33 people, reported that 82 percent had no residual psychotic symptoms at five years. The ODDESSI trial, led by Steve Pilling and published in The Lancet Psychiatry in 2026, randomized 494 adults in acute episodes. It found no significant difference in time to relapse, though Open Dialogue was linked to lower probabilities of psychiatric inpatient admission. Those findings are mixed and limited, yet they keep alive a question Szasz would have recognized: what happens when systems try harder to preserve consent rather than override it?
The Strongest Case for Civil Commitment
The case for civil commitment rests on old and serious principles. Parens patriae holds that the state may protect people who cannot protect themselves, and the police power covers public safety. In Addington, the Court worried that a standard too strict could block needed treatment. That worry is not a straw man. Families, clinicians, and judges who support commitment often describe people who are starving, freezing, or about to harm someone, and ask what a liberty-first rule would demand of bystanders then.
Supporters also cite empirical work that softens the picture of pure coercion. In the EUNOMIA study of 11 European countries, reported by Stefan Priebe and colleagues, 55 percent of involuntarily admitted patients thought admission was right after one month and 63 percent after three months. Torrey and the Treatment Advocacy Center argue that anosognosia, a lack of awareness of illness, can remove capacity, and that court-ordered outpatient treatment prevents relapse and harm. Critics reply that anosognosia is hard to separate from disagreement with a diagnosis.
Defenders point to reforms already in place: counsel, notice, time limits, standards of proof, and judicial review. Research on coercive outpatient orders is mixed even for them. A Cochrane review found a pooled 12-month readmission risk ratio of 0.98 (95 percent confidence interval 0.82 to 1.16), and the OCTET trial reported 36 percent readmission in both groups. A fair reading of the opposing case grants the urgency of danger and incapacity, grants the procedural machinery built to cabin abuse, and still faces mixed evidence that coercion delivers what its defenders promise.
Consent as the Measure
Whatever one thinks of Szasz, his question endures. A system that sets consent aside owes people an exceptional justification, and the record above suggests that hearings and disclosure sometimes fall short. His opponents answer that some people cannot protect themselves and that the state may not look away. Both claims can be true in different cases. The harder problem is telling those cases apart without collapsing into medical deference dressed as law.
Consent is only as real as the information behind it. People should hear the risks, the limits of the evidence, the alternatives, and the right to say no, and when a court overrides that right, the process should be careful, adversarial, and slow enough to be fair. Reasonable people can disagree about where the line falls and still agree on that standard. Measuring the hallway hearing by consent, rather than by diagnosis alone, keeps medicine from swallowing liberty.
This article offers general information only. It is not medical or legal advice.
