United States of America, Plaintiff-appellee, v. Robert Lyons, Defendant-appellant, 739 F.2d 994 (5th Cir. 1984)

Despite the prodigious volume of writing devoted to the plea,⁷  the empirical data that are available provide little or no support for these fearsome perceptions and in many respects directly refute them. Both the frequency and the success rate of insanity pleas are grossly overestimated by professionals and lay persons alike; the plea is rarely made, and even more rarely successful.⁸  The number of insanity pleas based on control defects, as compared to those based on lack of cognition, must have been almost negligible.

The perception that the defendant who successfully pleads insanity is quickly released from custody is also based only on assumption. Although an acquittal by reason of insanity ends the criminal jurisdiction of federal courts (except in the District of Columbia), and of the courts of a few states, the acquittee is not simply set free. "The truth is that in almost every case, the acquittee is immediately hospitalized and evaluated for dangerousness. Usually, the acquittee remains hospitalized for an extended time."⁹ 

In sum, the available evidence belies many of the assumptions upon which much popular criticism of the insanity defense are based. The plea is rarely invoked, usually fails, and, when it is successful, the acquittees rarely go free.

Another set of objections to the plea is based on the thesis that factfinders--especially juries--are confused and manipulated by the vagueness of the legal standards of insanity and the notorious "battle of the experts" who present conclusory, superficial, and misleading testimony. These conditions, the argument runs, conspire to produce inconsistent and "inaccurate" verdicts.

Let us first put these objections in perspective. Most cases involving an insanity plea do not go to trial; instead, like most other criminal cases, they are settled by a plea bargain. In many of the cases that do go to trial, psychiatric testimony is presented by deposition, without disagreement among experts, and without opposition by the prosecution.¹⁰  And in the few cases in which a contest does develop, the defendant is usually convicted.¹¹  Hence the stereotypic "circus tent" may be raised in only a handful of cases.

The manipulated-jury argument is supported largely by declamation, not data. Although there is some evidence to support the assertion that the wording of the insanity defense has little impact on trial outcomes,¹²  one major study of jury reactions to criminal cases involving the insanity defense reached conclusions contrary to the assumption that the jury function is usurped by expert testimony. That study found that jurors responsibly and carefully consider the evidence presented, do recognize that the final responsibility for the defendant's fate rests with them, do appreciate the limits and proper use of expert testimony, and do grasp the instructions given them.¹³  Although the evidence does not warrant the conclusion that juries function better in insanity trials than in other criminal cases, it certainly does not appear that they function less effectively. And no source has been cited to the court to support the conclusion that, as an empirical matter, pleas based upon the volitional prong present an especially problematic task for the jury.

Indeed, the majority opinion does not assert that the insanity defense, particularly the control test, doesn't work; it contends that the defense can't work. The principal basis for this contention is the belief, held by "a majority of psychiatrists," that they lack "sufficient accurate scientific bases for measuring a person's capacity for self-control or for calibrating the impairment of that capacity." This argument raises practical and important questions regarding the usefulness of expert testimony in determining whether a person has the ability to conform his conduct to law; but the absence of useful expert evidence, if indeed there is none, does not obviate the need for resolving the question whether the defendant ought to be held accountable for his criminal behavior.¹⁴ 

Accountability for criminal conduct involves more than mere actus reus; it also involves the mental state with which the act is performed.¹⁵  Factfinders are required by law to determine the existence or non-existence of many mental states, for example, whether the defendant was mistaken; acted from necessity, duress, or coercion; acted without willfulness or knowledge or intention;¹⁶  or was unable to appreciate the wrongfulness of his conduct.

Our concept of responsibility in this sense is not limited to observable behavior: it embraces meaningful choice, and necessarily requires inferences and assumptions regarding the defendant's unobservable mental state. The accompanying external circumstances in cases such as coercion are of course observable; but their effect on the defendant's mental state is inferred, and no more tangible than the mental processes of a defendant pleading the control test. The difference between the concepts of excusing circumstances such as coercion and the insanity defense is that the former is based on objective assumptions about human behavior and is tested against hypothetical-objective standards such as "the reasonable person." "The insanity defense [on the other hand] marks the transition from the adequate man the law demands to the inadequate man he may be."¹⁷ 

The relevant inquiry under either branch of the insanity test is a subjective one that focuses on the defendant's actual state of mind. Our duty to undertake that inquiry is not based on confidence in the testimony of expert witnesses, but on the ethical precept that the defendant's mental state is a crucial aspect of his blameworthiness.¹⁸  The competence of psychiatrists or other experts thus is not the issue. The issue is whether a defendant should be held accountable for his conduct. The availability of expert testimony and the probative value of such testimony are basically evidentiary problems that can be accommodated within the existing test.

A recent Second Circuit case demonstrates that the volitional test can be reasonably cabined to prevent abuse. In United States v. Torniero,¹⁹  the court affirmed the trial court's refusal to allow the defendant to present evidence that his compulsive gambling disorder rendered him unable to refrain from interstate transportation of stolen jewelery. The Second Circuit established two guidelines: (1) a defendant seeking to rely on a newly recognized disorder to meet the volitional test must show "that respected authorities in the field share the view that the disorder is a disease or defect that could have impaired the defendant's ability to desist from the offense charged ..."; and (2) the alleged disease or defect must be relevant to the crime charged. The first issue was sharply disputed at a pretrial hearing, and the trial court found in favor of the government. The circuit court upheld that finding and went on to conclude that "a compulsion to gamble, even if a mental disease or defect, is not, ipso facto, relevant to the issue whether the defendant was unable to restrain himself from non-gambling offenses such as transporting stolen property."²⁰ 

Even the few cases in which the trial develops into a battle of experts provide no basis for the majority's conclusion that the prosecution faces an "all but impossible task." The prosecution appears to be able to locate experts as readily as the defense. Indeed, a defendant pleading insanity typically faces both a judge and a jury who are skeptical about psychiatry in general and the insanity plea in particular.²¹  Sharply adversarial presentation of conflicting psychiatric testimony may increase this skepticism, and thus make acquittal more unlikely. Usually the defendant will have been adjudicated sane enough to understand the proceedings against him and to assist in his defense; otherwise he would be incompetent to stand trial. The formal allocation of the persuasion burden notwithstanding, the defendant to prevail must convince the doubting factfinder that, despite present outward appearances, he was insane at the time he committed the crime.

The majority's fear that the present test invites "moral mistakes" is difficult to understand. The majority opinion concedes that some individuals cannot conform their conduct to the law's requirements. Other writers have concluded that a strictly cognitive insanity test will overlook some individuals who would be covered by a control test.²²  Without citing any data that verdicts in insanity cases decided under a control test are frequently inaccurate, the majority embraces a rule certain to result in the conviction of at least some who are not morally responsible and the punishment of those for whom retributive, deterrent, and rehabilitative penal goals are inappropriate. A decision that virtually ensures undeserved, and therefore unjust, punishment in the name of avoiding moral mistakes rests on a peculiar notion of morality.

I do not speak in this dissent to the merits of a new system for determining criminal responsibility that would not merely redefine the insanity plea but would reform the entire trial process and the treatment of those who are not absolved. Such a system would eliminate the concept of guilty mind or criminal intent as an element of most offenses; and it would consider such matters, including insanity, at a dispositional proceeding separate from a determination whether the defendant committed the act.²³  Unlike the majority, those who propose such reform thus make provision to take account for the mentally ill offender's condition at some point in the process. This view requires a revision of our criminal justice system from a punitive to a preventive basis, not merely a change in the definition of insanity; for it rejects the concept of such a defense. I do not address the question whether such a system would be better or worse. Adoption of the view would require legislative changes in our trial system and changes in our institutional treatment of at least some convicted persons.

This dissent is directed only at the amputation of one part of a comprehensive definition. That definition is itself an integral part of a social mechanism that begins with a prosecutorial decision about whether a particular defendant should even be charged and ends with three possible results: not guilty and free, guilty and sentenced, not guilty by reason of insanity and committed.²⁴ 

Judges are not, and should not be, immune to popular outrage over this nation's crime rate. Like everyone else, judges watch television, read newspapers and magazines, listen to gossip, and are sometimes themselves victims. They receive the message trenchantly described in a recent book criticizing the insanity defense: "Perhaps the bottom line of all these complaints is that guilty people go free--guilty people who do not have to accept judgment or responsibility for what they have done and are not held accountable for their actions.... These are not cases in which the defendant is alleged to have committed a crime. Everyone knows he did it."²⁵  Although understandable as an expression of uninformed popular opinion, such a viewpoint ought not to serve as the basis for judicial decisionmaking; for it misapprehends the very meaning of guilt.

Guilt is the legal embodiment of moral and philosophical concepts. As discussed above, those concepts presuppose a morally responsible agent to whom guilt can be attributed.²⁶  By definition, guilt cannot be attributed to an individual unable to refrain from violating the law. When a defendant is properly acquitted by reason of insanity under the control test, the guilty does not go free.

The majority opinion is a radical departure from the established jurisprudence of every federal circuit that has spoken on the issue. It is based only on intuitive reactions and the published recommendations, to which no one has testified, of a few professional groups. We would permit no jury to decide even an unimportant issue on such hearsay. The purposes to be served by this innovation are unclear, for it is doubtful that the decision will make the slightest impact on the enormous problem of crime. Its effect will be felt by only two small groups: a few who otherwise might have made a case for the jury but who will be deprived of a plea that in any event would likely have been bootless, and those few unfortunate persons so afflicted by mental disease that they knew what the law forbade but couldn't control their actions sufficiently to avoid violating it. The nature of their illness makes punishment useless as a means to compel them to behave lawfully; only remission of the disease will permit change. By convicting them we will not deter others who suffer the same mental defect; they too, again by definition, have no ability to obey.

In sum, I cannot join in a decision that, without supporting data, overturns a widely used rule that has not been shown to be working badly in order to adopt a change that will likely produce little or no practical benefit to society as a whole, conflicts with the fundamental moral predicates of our criminal justice system, and may inflict undeserved punishment on a few hapless individuals.