No Constitutional Mandate

No Constitutional Mandate

The earlier sections treated the control test as a rule that a jurisdiction chooses to adopt or to leave out. This section asks whether any jurisdiction is required to adopt it, and the answer is no. That answer comes from two separate holdings that respond to two different constitutional claims. The first claim is that due process makes a control test part of fundamental fairness, and Leland v. Oregon rejected it.¹ The second is that the Eighth Amendment forbids punishing conduct that flows from a disease-driven compulsion, and Powell v. Texas declined to accept it.²

In both settings the question the Court answers is one of constitutional power: who gets to choose the test of responsibility. Which test is wisest lies outside that question, so a reader who thinks the irresistible impulse test is the better rule can still conclude that a state is free to reject it. The choice stays with the states because of what it involves. A test of responsibility depends on medical knowledge that remains contested and on a judgment of values about how far that knowledge should excuse.³ When specialists disagree about the first and legislators and voters disagree about the second, no constitutional text settles either dispute, and a federal court has no principled basis for fixing one answer for every state. Leland, the due process holding, comes first, and Powell, the Eighth Amendment holding, follows.

Leland and State Choice

In Leland v. Oregon, a murder prosecution, the defendant challenged an Oregon statute providing that a morbid propensity to commit prohibited acts forms no defense unless the person was shown to have been incapable of knowing the wrongfulness of those acts.⁴ In plain terms, the statute says that an urge to commit a crime, however strong, excuses only when the person also could not tell that the act was wrong. The Court read the statute as nothing more than a legislative choice of the right-and-wrong test over the irresistible impulse test, and it held that choice consistent with due process.⁵ The same decision upheld Oregon's rule that the accused must prove insanity beyond a reasonable doubt, a point covered earlier in this chapter, because the Court was reluctant to interfere with a state's policy on the burden of proof when that policy did not violate generally accepted concepts of basic standards of justice.⁶

The Court's reasoning has two parts, and each does separate work. The first part is empirical. Knowledge of right and wrong was the exclusive test of criminal responsibility in a majority of American jurisdictions, and although psychiatry had advanced a great deal since M'Naghten's Case, it had not reached the point where its learning would compel the states to give up the right-and-wrong test.⁷ A constitutional command would need something like a scientific consensus that a knowledge-only test misdescribes who can fairly be blamed, and the Court found none. The second part is about values. Choosing a test of legal sanity involves not only scientific knowledge but basic policy about how far that knowledge should determine criminal responsibility, and those who have studied the problem disagree widely.⁸ Even a perfect science of mental illness could not say how much impairment should excuse, because that is a question about blame. From these two parts the Court concluded that adopting the irresistible impulse test is not "implicit in the concept of ordered liberty."⁹

A structural reason reinforces the result. A test written into the Constitution would bind every state at once and could change only by a decision of the Court, so it would freeze one side of a live dispute and shut off the variation that lets states respond as medical understanding changes. The Due Process Clause also does not condemn a state's procedure merely because another method seems fairer or wiser to the reviewing judges, and the question is whether the state's policy offends generally accepted standards of basic justice.¹⁰

The Court later restated the principle in modern terms. In Clark v. Arizona the defendant attacked Arizona's decision to drop the cognitive-incapacity half of the M'Naghten formula from its insanity definition, arguing that the full formula was the minimum a state must provide.¹¹ The Court answered that history shows no deference to M'Naghten that could lift its formula to the level of fundamental principle, and it catalogued four historical strains: cognitive incapacity, moral incapacity, volitional incapacity, and the product-of-mental-illness test. The volitional strain is the irresistible impulse test, which asks whether a person was so lacking in volition because of a mental defect or illness that he could not have controlled his actions.¹² Because the states have combined those strains in many different ways, no single formulation serves as a baseline for due process, and the Court said that due process imposes no single canonical formulation of legal insanity.¹³

Kahler v. Kansas carried the principle further still. Kansas lets a defendant use mental illness to show that he lacked the required mens rea, but it does not acquit a defendant solely because illness kept him from recognizing the act as morally wrong, and the Court held that the Due Process Clause does not compel Kansas to do otherwise.¹⁴ Its survey again treated the irresistible impulse test as one historical variant among several, found no settled form of the insanity defense that is essential to due process, and read Leland, Clark, and Powell as confirming that states have considerable leeway in defining the defense.¹⁵ Leland and Kahler face in opposite directions. Leland upheld a state that kept the knowledge test and declined to add a control test, while Kahler upheld a state that did not recognize even the moral-capacity inquiry as an independent route to acquittal. What they share is the rule that the shape of the responsibility defense is a matter of state choice.

Students lose this issue when they read Leland as disapproving of the irresistible impulse test, when the case holds only that the Constitution does not require it. A state that adopts a control test does so as a matter of its own law, and the Constitution's silence permits both the test and its absence.

A neighboring line of cases should be kept separate from this one. In Kansas v. Hendricks the Court upheld the civil commitment of a man with pedophilia who conceded that he could not control the urge to molest children, and it treated that admitted lack of volitional control, together with a prediction of dangerousness, as what separates a person suited to civil commitment from other dangerous persons dealt with through criminal proceedings.¹⁶ Kansas v. Crane later rejected the claim that such an offender may be committed with no lack-of-control determination at all, while also holding that the Constitution does not demand total or complete lack of control, only proof of serious difficulty in controlling behavior.¹⁷ Those decisions constrain civil confinement, where the state must show why the person differs from the ordinary dangerous criminal. They say nothing about whether a criminal defendant must be offered a control-based excuse, which Leland, Clark, and Kahler leave to the states. The point that control comes in degrees returns later in this chapter, in the discussion of total incapacity.

Powell and Compulsion

Powell v. Texas pressed the second claim, and it reached further than any attack on a single insanity test. The appellant, a chronic alcoholic convicted of public drunkenness, urged that the Eighth Amendment bars punishing conduct that results from a compulsion symptomatic of a disease, and the argument could succeed only by discarding the original boundary of Robinson v. California, which the Court had limited to pure status crimes.¹⁹ The Court declined to take that step. Because Powell produced a plurality opinion and separate concurrences, its reasoning comes in more than one voice. The plurality cautioned against being drawn into defining an insanity test in constitutional terms, and said it was not yet time to write into the Constitution "formulas cast in terms whose meaning, let alone relevance, is not yet clear either to doctors or to lawyers."²⁰

Justice Black's concurrence supplies two distinct reasons for refusing the claim. The first is the absence of any limit. If the boundary of Robinson were abandoned, any new limit would soon fall away, and the states would be powerless to punish any conduct shown to result from a compulsion in the psychological sense, which would make irresistible impulse a complete defense to any crime, a result probably contrary to the law of most jurisdictions.²¹ The second goes to the premise of the argument, which assumes that punishing a person who is not morally blameworthy is cruel and unusual, even though legislatures have always had wide freedom to decide how far moral culpability must be a prerequisite to conviction.²² The first reason is practical, because a constitutional rule cannot be adjusted case by case the way a statute can. The second concerns who decides: how much blame the law requires is a policy question that the Constitution leaves to legislatures.

Limiting the claim to compulsions that are symptoms of a disease would not have cured either problem. A narcotics addict could not be punished for possessing or using drugs, a wide variety of sex offenders would be immune if their conduct were part of the pattern of a disease, and a form of the insanity defense would become a constitutional requirement throughout the Nation.²³ Such a holding would also appear to overrule Leland v. Oregon, where both the majority and the dissent stressed the indefensibility of imposing any particular test of criminal responsibility on the states.²⁴ That overlap is why the two holdings belong together. A constitutional compulsion defense is the control test mandated by another route, so Powell closes under the Eighth Amendment the door that Leland had already closed under due process. The effect would have been greatest in the states that keep the right-and-wrong test, apparently at least thirty, but it would have disturbed even states with broader defenses, since every state rule on evidence of disease and compulsion, and every jury instruction explaining those concepts, would have raised a constitutional question.²⁵ The outcome fits the general rule, restated in Justice Ginsburg's concurrence in Montana v. Egelhoff, that states have wide latitude to define criminal offenses and the extent to which moral culpability is required, subject only to the limit that a definition not offend a principle of justice so rooted in the traditions and conscience of our people as to be ranked as fundamental.²⁶

The compulsion claim rejected in Powell must be separated from the status-crime rule that survives. The concurrence describes Robinson as limited to pure status crimes and explains that crimes requiring the State to prove that the defendant actually committed a proscribed act do not raise the special problems that status offenses do.²⁷

Students lose this issue when they assume that any charge resting on a disease-driven act is constitutionally suspect, because Robinson reaches only pure status crimes. A compulsion argument against an act-based charge must come from the state's own insanity or mens rea doctrine, and the Eighth Amendment will not supply it.

Leland and State Choice

The due process question about control tests reached the Court in Leland v. Oregon, a murder prosecution in which the defendant attacked an Oregon statute providing that a "morbid propensity to commit prohibited acts, existing in the mind of a person, who is not shown to have been incapable of knowing the wrongfulness of such acts, forms no defense to a prosecution therefor."²⁹ Put plainly, the statute says that an urge to commit a crime, however strong, excuses only when the person also could not tell that the act was wrong. The Court read the provision as a legislative adoption of the right-and-wrong test in preference to the irresistible impulse test, and it held that due process did not require Oregon to choose the other way.³⁰ The holding therefore answers a question about who decides which test governs, and it answers in favor of the state legislature.

The Court's reasoning has two layers that rest on different ground. The first layer concerns what the states and the medical profession showed at the time. When Leland was decided in 1952, knowledge of right and wrong was the exclusive test of criminal responsibility in a majority of American jurisdictions, and although psychiatry had advanced a great deal since M'Naghten's Case, its progress had not reached the point where its learning would compel the states to give up that test.³¹ A constitutional command to adopt a rival test would need something close to a settled medical judgment that a knowledge-only rule misdescribes who may fairly be blamed, and the Court found no such judgment. The second layer concerns values. Choosing a test of legal sanity involves scientific knowledge and also basic policy about how far that knowledge should determine criminal responsibility, and those who had studied the problem disagreed widely.³² Even a perfected science of mental illness could not say how much impairment should excuse, because that question concerns blame, which medicine does not measure. From the two layers together the Court concluded that adopting the irresistible impulse test is not "implicit in the concept of ordered liberty."³³

The dissent in Leland shows how little disagreement there was on this point. It objected to Oregon's rule that the accused must prove insanity beyond a reasonable doubt, yet it conceded that, at the existing stage of scientific knowledge, imposing one test rather than another on the states through due process would be indefensible, and would displace a state's own choice of test "no matter how backward it may be in the light of the best scientific canons."³⁴ The side that wanted the conviction reversed thus located its constitutional objection in the allocation of the burden of proof, and left the choice of a knowledge-only test alone. That burden-of-proof ruling was treated earlier in this chapter, and the Court later declined to reconsider Leland on that point in Patterson v. New York.³⁵

Two features of the rule's shape explain why the Court left the matter with the states. A test written into the Constitution would bind every state at once and could change only by a later decision of the Court, so it would freeze one side of a live dispute and end the variation that lets legislatures respond as medical understanding changes. A legislature, by contrast, can revise its statute as the science and public judgment about blame move. The Court's general approach to due process points the same direction: a state's procedure does not violate the Fourteenth Amendment merely because another method seems fairer or wiser to the reviewing judges.³⁶ The modern statement of the test asks whether the state rule offends a principle of justice "so rooted in the traditions and conscience of our people as to be ranked as fundamental."³⁷ A knowledge-only test could hardly meet that description when a majority of jurisdictions used it.

Oregon's statute marks the pole where the rule applies. The state keeps the knowledge test, and a defendant who knew the act was wrong cannot demand a control instruction as a constitutional right. The other pole is the group of states that chose differently, which Leland leaves undisturbed. In Clark v. Arizona the Court counted fourteen jurisdictions, inspired by the Model Penal Code, that combine a volitional test with a variant of the moral incapacity test, and three more states that pair the full M'Naghten formula with a volitional formula.³⁸ Nothing in Leland touches those choices, because the Constitution neither commands nor forbids a control test. What separates the two poles is only the legislature's decision.

Later decisions have repeated the principle in broader terms. Clark found that no particular formulation of insanity has evolved into a baseline for due process, that the insanity rule is substantially open to state choice, and that due process imposes no single canonical formulation of legal insanity, and it cited Leland for the point that choosing the knowledge test over the irresistible impulse test is a matter of policy.³⁹ Kahler v. Kansas read Leland and Clark as confirming that states have considerable leeway in defining the insanity defense, and found no settled form of the defense that is essential to due process.⁴⁰ A California appellate court applied the same reasoning in ruling that due process does not compel a state to add irresistible impulse to M'Naghten.⁴¹

Students lose this issue when they read Leland as disapproving of the control test, when the case makes the test constitutionally optional in both directions. A state that adopts it does so under its own law, and a state that omits it violates nothing in the Constitution. On an exam stem in which a defendant knew the act was wrong but could not stop, the first question is which test the jurisdiction uses, because due process will not decide the outcome.

Powell and Compulsion

Leland settled that the Due Process Clause does not oblige a state to adopt a control test, which left a second constitutional route to the same destination through the Eighth Amendment. Powell v. Texas took that route and closed it. Leroy Powell, a chronic alcoholic, was convicted of being drunk in a public place, and he argued that punishing the public appearance was cruel and unusual because it flowed from a compulsion to drink that his disease produced.⁴³ The claim asked the Court to move beyond Robinson v. California, which had held that a person may not be punished for the mere status of being a narcotics addict.⁴⁴ Because the Court produced a plurality opinion and separate concurrences, the reasons are best read in layers, beginning with the lead opinion.

The lead opinion gave three connected reasons. The first was evidentiary: on that record and on the state of medical knowledge then available, the Court could not conclude that chronic alcoholics in general, or Powell in particular, were so compelled to drink and to get drunk in public that they were utterly unable to control those acts and could not be deterred at all.⁴⁵ The standard the Court used was total inability, the same strong form of control that the traditional irresistible impulse test demands, and that feature returns when this chapter reaches the total-incapacity problem. The second reason was doctrinal. The Court had never articulated a general constitutional doctrine of mens rea, the guilty-mind element of an offense, so the claim had no constitutional rule to attach to.⁴⁶ The third reason concerned who decides. Actus reus, mens rea, insanity, mistake, justification, and duress are the common-law tools that keep adjusting the criminal law to changing medical and moral views, and the lead opinion described that adjustment as having always been thought to be "the province of the States."⁴⁷

Beneath all three reasons sits a practical objection: a constitutional compulsion defense could not be kept small. The dissenting Justices proposed confining the rule to conduct that is a characteristic and involuntary part of the pattern of the disease, and the lead opinion answered that this would be "limitation by fiat."⁴⁸ Nothing in the logic of the rule would stop at chronic alcoholics. If Powell could not be convicted of public intoxication, it would be hard to see how a state could convict a person of murder who behaved normally in every other respect but suffered a compulsion to kill that was an exceedingly strong influence yet not completely overpowering.⁴⁹ The difficulty follows from how constitutional rules work. A legislature can draw a line, watch how it performs, and move it, while a rule read into the Eighth Amendment would bind every state at once and could change only by a later decision of the Court, so a badly drawn line would stay where it fell.

Justice Black's concurrence reached the same result and stated most sharply what the rule would cost. If the original boundaries of Robinson were discarded, he reasoned, the Court would be forced to hold the states powerless to punish any conduct shown to result from a compulsion, and the result would be to require recognition of irresistible impulse as a complete defense to any crime, which he described as probably contrary to the law of most American jurisdictions.⁵⁰ That last clause reports the law as the concurrence saw it in 1968, so it should be attributed that way and not read as a current count of jurisdictions. The concurrence also saw the doctrinal stake, since a holding that punishing disease-driven compulsion is unconstitutional would appear to overrule Leland by imposing a form of the insanity defense on every state.⁵¹ A constitutional compulsion defense is therefore the control test mandated by another route.

The concurrence also explains why compulsion, even when real, does not automatically excuse. Medical judgments about disease and volition answer clinical questions of diagnosis and treatment, and they have no necessary correspondence to the legal question whether punishment serves the purposes of the criminal law.⁵² Because the person who committed the act is the person punished, deterrence, isolation of the dangerous, and treatment can all be served without first deciding which part of the defendant's personality produced the act. Deterrence can still operate on a chronic alcoholic, since the threat of jail may strengthen the incentive to control the frequency and location of drinking, and civil commitment is no obvious improvement, because it may last as long as jail or longer and carry a similar stigma.⁵³

The refusal also fits a broader feature of the criminal law, which has no general principle excusing every loss of self-control. It recognizes particular defenses in areas where the condition can be verified, and it reserves the claim of individual justice for ascertainable conditions such as mental disease or defect.⁵⁴ The reason is administrative as well as moral. A system of social control that must reach the great mass of the population cannot vary its norms with each person's capacity except where the disability is gross and verifiable, and lesser impairments are better accounted for at sentencing.⁵⁵ That reasoning shows why the control test discussed earlier in this chapter requires a mental illness as its source.

The Eighth Amendment does limit punishment in some settings involving impaired self-control, and the line between those settings and Powell needs care. In Atkins v. Virginia the Court held that executing an offender with intellectual disability, a condition the opinion called mental retardation, is cruel and unusual punishment, reasoning that impairments in reasoning, judgment, and impulse control reduce moral culpability below the level that justifies the most extreme penalty.⁵⁶ That holding limits how severely a category of offenders may be punished, and it leaves the underlying crime unexcused. The opinion states that persons with such disabilities who meet the law's requirements for criminal responsibility should be tried and punished when they commit crimes, and it leaves to the states the task of developing ways to enforce the restriction.⁵⁷ Powell sought an acquittal, while Atkins removes only the death penalty. Kahler v. Kansas later grouped Powell with Leland as showing that states have considerable leeway in defining the insanity defense, and it summarized Powell as rejecting the argument that the Eighth Amendment requires an insanity defense.⁵⁸

Students lose this issue when they read Powell as saying that compulsion never matters. The holding is narrower, because compulsion is not a constitutional defense, so a defendant who knew the act was wrong but could not stop must look to the jurisdiction's own insanity test, which Leland leaves to the state. The one constitutional limit that survives is the status-crime boundary of Robinson. Powell's statute required proof of an act, appearing in public while drunk, and the concurrence found it entirely proper to confine Robinson to pure status crimes where no conduct is involved.⁵⁹