Concept
Decline and Current Status
Product of Mental Disease or Defect
Decline and Current Status
The three objections just described left the District of Columbia court with a test whose wording it had repaired but whose structure remained under attack, and the court's final step shows where that pressure led. In United States v. Brawner the court overruled Durham in favor of a modified version of the Model Penal Code test, abandoning a formula it had announced eighteen years earlier and decrying the “trial by label” that the product test had produced.1 The replacement deserves attention first, because it shows how a test can keep a causal phrase and still avoid the difficulties that undid the product test. The second part turns to the constitutional rule that leaves each state free to choose its own formulation, which explains how the product test can still be found in one state today.
Move to Substantial Capacity
The test that replaced Durham is the one approved by the American Law Institute and set out in Model Penal Code § 4.01. Under the substantial-capacity test, a person is not responsible for criminal conduct if, at the time of the conduct and as a result of mental disease or defect, the person lacks substantial capacity either to appreciate the criminality of the conduct (the Code's bracketed alternative reads wrongfulness) or to conform the conduct to the requirements of law.2 The second paragraph adds that the terms mental disease or defect do not include an abnormality manifested only by repeated criminal or otherwise antisocial conduct.3 A reader coming from the product test will notice that a causal phrase survives, because the incapacity must arise as a result of the disease. The causal link has moved, though. It now runs from the disease to an impairment of two named capacities, which witnesses can describe from the person's history and symptoms, and it no longer runs from the disease to the act itself, which is the but-for reconstruction that critics found impossible to answer.
The Code's draftsmen rejected the product formula because of the ambiguity of the word product, and they chose instead to restate the older cognitive and volitional tests in broader terms.4 Their reason was that impairment of cognition and impairment of the capacity for self-control are the conditions that must be taken into account to exclude from penal sanctions the people whom the threat of punishment cannot deter. That reasoning answers the overbreadth objection directly, since the capacity filter that critics wanted placed on top of causation is now the rule itself. Each prong was also reworded to repair a known weakness. The Code says appreciate rather than know, which lets testimony about a person's emotional or affective makeup bear on the question, and it speaks of conforming conduct, which avoids the suggestion drawn from the irresistible impulse test that lost control appears only in sudden acts and never in brooding or planned ones.
The word substantial has its own rationale. The older tests were usually read to demand a complete impairment of cognition or of self-control, whereas the Code asks only for a lack of substantial capacity.5 Candid witnesses who must infer a mental state at a time they did not observe can rarely say more than that capacity was substantially lacking, even when they know the disorder was extreme, and many advanced disorders leave rudimentary capacities in place while still presenting inappropriate occasions for criminal punishment.6 The draftsmen conceded that substantial supplies no precise measure of degree, but they concluded that identifying the degree of impairment with precision was impossible both verbally and logically.7
The result in that scene follows a general pattern: a delusional defendant who has some idea of what the act is, or that it is wrong, fails the right-wrong test but might still lack substantial capacity under the Code.8 The relaxation is modest in practice, because even under the more flexible test few mentally ill defendants qualify for the defense.9 The other side of the line is just as important. A disorder that only colors a person's mood and leaves both capacities substantially intact excuses no one under the Code, however clearly it contributed to the act.
Paragraph (2) of the Code answers a different problem, the one the District of Columbia met in Blocker v. United States. There the court held that a defendant with a sociopathic personality was entitled to a new trial after the psychiatric profession, or a portion of it, began calling sociopathy a mental disease.10 Under Durham a change in clinical vocabulary could change who was excused, because the disease had to be nothing except a factor causing the crime.11 The Code closes that door. The draftsmen reasoned that the psychopath differs from the normal person only in degree and that opinion divides on whether psychopathy is a disease. A condition shown only by repeated criminal behavior also proves disease only circularly, since that behavior is exactly what the disease must be shown to have produced, and excluding it avoids litigating what is essentially a matter of terminology.12
The Code test did not remain the last word. Virtually all of the federal courts of appeals adopted it, and a significant minority of states did the same, usually by statute, but Congress repudiated it for federal prosecutions in 1984 in favor of a M'Naghten-style statutory formulation.13 The change followed the verdict in which John Hinckley was found not guilty by reason of insanity on the charge of attempting to murder President Reagan, which provoked a public outcry.14 Under 18 U.S.C. § 17 it is an affirmative defense that, as a result of a severe mental disease or defect, the defendant was unable to appreciate the nature and quality or the wrongfulness of the acts, and mental disease or defect does not otherwise constitute a defense.15 The statute drops the capacity-to-conform prong, raises the threshold from substantial incapacity to inability, and requires the defendant to prove insanity by clear and convincing evidence.16 The exam trap is to assume that the Model Penal Code test governs federal prosecutions because it replaced Durham in the District of Columbia. Many jurisdictions narrowed the defense after the Hinckley verdict, though portions of the substantial-capacity test continue to be used.17
New Hampshire Alone
The product test's disappearance from the District of Columbia raises the question of how any American court can still apply it, and the answer lies in the Constitution's silence about insanity formulas. The insanity defense itself is not enshrined in the Constitution, so a state may abolish it, as Idaho, Kansas, Montana, and Utah have done, although evidence of mental illness can still be offered to raise a reasonable doubt about the mens rea for the crime.19 As to the formula, the Supreme Court wrote in Clark v. Arizona that “History shows no deference to M'Naghten that could elevate its formula to the level of fundamental principle, so as to limit the traditional recognition of a State's capacity to define crimes and defenses.”20 If the dominant formula is not constitutionally fundamental, no rival formula is either, and that includes the product test.
The same hesitation to constitutionalize appears in Powell v. Texas. The Court affirmed the conviction of a chronic alcoholic for public drunkenness and, in rejecting an extension of Robinson, observed that the extension would push it toward defining an insanity test in constitutional terms, naming both Durham's product formula and M'Naghten's right-wrong test as examples of the competing formulations at stake.21 The opinion gave a reason that applies directly here: a constitutional rule would reduce or eliminate the experimentation of the States and freeze the developing dialogue between law and psychiatry into a rigid constitutional mold.22 Experimentation is the process that produced the product test, its adoption in the District of Columbia, and its abandonment there, and the same latitude lets New Hampshire keep it.
Because state choice produces variety, it helps to sort the formulations by what each one examines, since sorting shows where the product test sits. The four families are these:
- Cognitive incapacity asks whether a disease of the mind left the accused not knowing the nature and quality of the act, or not knowing that it was wrong.23
- Moral incapacity arises where wrong is read to mean morally wrong rather than legally wrong, a question on which American jurisdictions are divided.24
- Volitional incapacity asks whether the disease left the accused unable to conform conduct to law, an approach that remains the minority one because it is hard to distinguish an irresistible impulse from an impulse that was simply not resisted.25
- Product of mental illness asks only whether the unlawful act was the product of mental disease or defect, so it examines no capacity at all.26
The Model Penal Code formulation combines the cognitive and volitional families, which is why it is not listed separately.27 The product test is therefore the only one of the four that asks about a causal relationship rather than about a capacity, and that difference is what makes the same facts come out differently across jurisdictions.
On an exam problem the first step is therefore to identify the formulation the jurisdiction uses, because the outcome depends on which family of test applies. Among the states, the product test now survives only in New Hampshire, which shares it with the Virgin Islands, a territory.28 That status completes a symmetry in the rule's history. New Hampshire's court first applied the formula in the nineteenth century, and for the next eighty-three years no other American jurisdiction adopted it or gave it serious consideration.29 New Hampshire's experience during those decades offered no further explanation of what product means, so the formula carried its original ambiguity into the District of Columbia, where it was explained, narrowed, and finally abandoned.30 Students sometimes infer from Brawner that the product test no longer exists anywhere, but New Hampshire still applies it. The District of Columbia court abandoned its version for reasons of workability, and one jurisdiction's judgment about workability binds no other.
States may likewise decide who must prove insanity. Once the prosecution has proved every fact constituting the crime beyond a reasonable doubt, a state may refuse to sustain the affirmative defense of insanity unless the defendant demonstrates it by a preponderance of the evidence, as Patterson v. New York confirmed.31 The federal statute is one example of a defendant-borne burden, so burden allocation is a second point, after the test itself, on which jurisdictions are free to differ.
Move to Substantial Capacity
The test that replaced the product rule is the formulation the American Law Institute approved in Model Penal Code § 4.01, and it is easiest to learn by reading it against the rule it displaced. Paragraph (1) provides that a person is not responsible for criminal conduct if, at the time of the conduct and as a result of mental disease or defect, the person lacks substantial capacity either to appreciate the criminality of the conduct (the Code's bracketed alternative reads wrongfulness) or to conform the conduct to the requirements of law.33 Paragraph (2) adds that the terms mental disease or defect do not include an abnormality manifested only by repeated criminal or otherwise antisocial conduct.34 The result is the substantial-capacity test, and measured against the product test it differs in three ways: what the causal link connects, what the experts are asked to describe, and how much impairment is enough. Each difference has its own reason, and the first explains why the Code looks, at a glance, like a relative of the rule it replaced.
A reader coming from the product test will notice that a causal phrase survives, because the incapacity must arise as a result of the disease.35 The drafters nevertheless rejected the product formula on account of the ambiguity of the word product, and what they changed was the thing the causal phrase connects.36 Under the product test, as the courts later explained it, the jury had to decide whether the defendant would have committed the crime but for the disease, which critics described as a mental exercise that cannot be performed, since no one knows what a person would have done with a different mind.37 Under the Code, the causal link runs from the disease to an impairment of cognition or self-control, so the question becomes what effect the disease had on the defendant's cognitive or behavioral capacities.38 Witnesses can speak to that question from the defendant's history and symptoms. A court that adopted the Code test for its circuit made the same point when it observed that judges and jurors, relieved of the burden of divining precise causal relationships, can concentrate on the ultimate decisions that belong to them.39
The two capacities the Code names come from the older tests, which the drafters restated more broadly instead of discarding.40 They treated impairment of cognition and impairment of volitional capacity as the conditions that must be taken into account to exclude from penal sanctions the people whom the threat of punishment cannot deter.41 Each branch was also reworded to repair a known weakness. The Code says appreciate instead of know, which lets testimony about a person's emotional or affective makeup bear on the question, and a court adopting the test observed that bare intellectual awareness of wrongfulness, divorced from an understanding of its moral or legal import, has little significance.42 The Code says conform, which avoids the suggestion drawn from the irresistible impulse test that lost self-control shows itself only in sudden or spontaneous acts and never in brooding or planned ones.43 This design also answers the overbreadth objection to the product test, because the capacity filter that critics wanted placed on top of causation is now part of the rule itself.
The third difference is the degree of impairment required. The older cognitive and volitional tests were usually read to demand a complete impairment, while the product test could be read to demand only a slight one, and the Code takes a position between those extremes by requiring a lack of substantial capacity.44 A court adopting the test explained that the word excludes both extremes: incapacity of any degree is not enough, and total incapacity is unnecessary.45 The reason lies in the evidence the test must be applied to. Candid witnesses asked to infer a mental state at a time they did not observe can rarely say more than that capacity was substantially lacking, even when they know the disorder was extreme, and many advanced disorders leave rudimentary capacities in place while still presenting inappropriate occasions for criminal punishment.46 The draftsmen conceded that substantial supplies no precise measure of degree, but they concluded that identifying the degree of impairment with precision was impossible.47
That pairing follows a general pattern: a delusional defendant who has some idea of what the act is, or that it is wrong, fails a test demanding total cognitive incapacity but may still lack substantial capacity under the Code.48 The relaxation is modest in practice, because even under the Code's more flexible approach few mentally ill defendants qualify for the defense.49 The second defendant shows where the product test and the Code part ways. An illness that contributed to the act may satisfy a causal inquiry, yet the Code excuses only when the illness takes away a named capacity to a substantial degree, so students lose this issue when they stop at causation and treat causal contribution without substantial incapacity as enough.50
Paragraph (2) answers a problem the product test left open, which is who counts as mentally diseased at all. Neither test defines mental disease or defect, except that the Code excludes an abnormality manifested only by repeated criminal or otherwise antisocial conduct.51 The draftsmen aimed the exclusion at the so-called psychopathic personality, reasoning that the psychopath differs from a normal person only in degree and that opinion is divided on whether psychopathy is a disease.52 A circularity problem also supports the exclusion. Repeated criminal behavior is the very phenomenon that must, by hypothesis, result from disease before irresponsibility is established, so offering that behavior as the proof of the disease proves nothing independent, and excluding it avoids litigating what is essentially a matter of terminology.53 Under the product test the disease had to be nothing except a factor causing the crime, so the label attached to a condition decided the case.54 That vulnerability was visible in 1957, when a staff conference at the hospital supplying most of the District's psychiatric witnesses reclassified psychopathic personality as a mental disease, and juries were abruptly told that defendants previously treated as responsible were now to be acquitted.55
When the District of Columbia Circuit abandoned the product test in 1972, it adopted the Code formulation and laid over it the definition of mental disease or defect it had developed under the product test, an abnormal condition of the mind that substantially affects mental or emotional processes and substantially impairs behavior controls.56 The Code's own reach later narrowed. Since 1984 many jurisdictions, including the federal courts, have moved to a M'Naghten-style test in response to the Hinckley verdict, although portions of the substantial-capacity test continue to be used.57 The volitional branch in particular remains a minority position, because it is hard to distinguish an irresistible impulse from an impulse that was simply not resisted.58 A problem that invokes the Code test by name is therefore answered through capacity, while a problem set in a federal prosecution calls for the federal formulation.
New Hampshire Alone
The District of Columbia's retreat from the product test, described above, ended the rule's run in the jurisdiction that made it famous, but it did not remove the rule from American law.60 The clearest picture of where the test now stands comes from Clark v. Arizona, in which the Supreme Court surveyed American insanity law and sorted it into four traditional strains. Cognitive incapacity asks whether mental illness left the defendant unable to understand what he was doing, and moral incapacity asks whether it left him unable to understand that the act was wrong; both come from the alternatives stated in the M'Naghten rule. Volitional incapacity, the irresistible-impulse idea, asks whether illness left the defendant unable to control his actions. The fourth strain, the product-of-mental-illness test, asks simply whether the act was a product of mental disease or defect, and the Court noted that it had been used as early as 1870.61
The Court also counted how the jurisdictions had combined these strains, and the count shows how isolated the product test had become. The tally, as the Court reported it in 2006, was as follows.62
- Seventeen states and the federal government use a recognizable version of M'Naghten with both its cognitive and moral incapacity components.
- One state uses M'Naghten's cognitive incapacity test alone, and ten states, including Arizona, use the moral incapacity test alone.
- Fourteen jurisdictions, inspired by the Model Penal Code, combine volitional incapacity with some variant of moral incapacity, so that either one excuses.
- Three states combine a full M'Naghten test with a volitional incapacity formula.
- New Hampshire alone uses the product-of-mental-illness test.
Only the last group has a single member. The tally is a 2006 snapshot, and this chapter reports it as the Court gave it, but later state-law summaries still place New Hampshire alone among the states and add the Virgin Islands, a territory rather than a state, as the one other place that applies the product rule.63 New Hampshire is also where the formula began: its courts adopted it around 1870, and the court that decided Durham in 1954 patterned its rule on New Hampshire's.64
Today New Hampshire frames the inquiry as whether a mental disease or defect caused the charged conduct, and the test is broad enough to take in whether the defendant knew right from wrong, whether the defendant acted impulsively, and whether the defendant suffered delusions or hallucinations.65 That breadth follows from the structure of the rule. The test names no capacity that the defendant must have lost, so each kind of impairment matters only to the extent that it caused the conduct, and that is why the product test sits apart from the other three strains, each of which asks what the illness took away.
A rule that nearly every other state has left behind can survive because of what Clark decided about due process. The case approached the question from the narrowing side. Arizona had first codified the full M'Naghten statement, then in 1993 removed the cognitive-incapacity part and left moral incapacity alone, and Clark argued that M'Naghten's two-part formula was the minimum that due process allows.66 The Court rejected the argument. History showed no deference to M'Naghten that could raise its formula to a fundamental principle, the traditional approaches differed so widely that no formulation had become a due process baseline, and the insanity rule, like the definition of crimes, is substantially open to state choice.67 The Court summed up the point this way: "due process imposes no single canonical formulation of legal insanity."68
The reason lies in what a legal insanity test borrows from medicine. The legal definition depends on concepts of mental abnormality that shape the expert testimony offered at trial, and medical definitions devised to justify treatment, like legal definitions devised to excuse from conventional responsibility, change over time and provoke disagreement.69 A constitutional rule would therefore bind every state to one formulation at a moment when neither doctors nor lawyers could say with confidence which one is right, and it would end the experimentation by which states learn from one another's tests. A plurality in Powell v. Texas made that point when it declined to write an insanity test into the Constitution, warning that doing so would freeze the developing dialogue between law and psychiatry into a rigid constitutional mold. The plurality even used the Durham formula as its example, observing that a defendant tried under the right-wrong test could otherwise claim a constitutional entitlement to the product test.70
The same latitude protects narrow tests. As Clark describes Leland v. Oregon, the Court found no due process violation when a state chose M'Naghten over the irresistible-impulse test, because scientific knowledge did not require otherwise and the choice of test is a matter of policy.71 Even the dissent in Leland, which objected to Oregon's burden of proof, treated it as indefensible to use due process to impose one test of culpability on the states, however backward that state's choice might look by the best scientific standards.72
Two situations mark the line. A state that keeps the product test, as New Hampshire does, acts within its latitude, and nobody can attack the choice by pointing out that the rest of the country has moved on.73 A defendant who complains that a state's test is too narrow, as Clark did about Arizona's moral-incapacity-only definition, loses for the same reason.74 The challenge fails just as clearly when it runs in the broadening direction.
Texas shows how differently a state can make the choice. In Williams v. State, a Texas appellate court restated that insanity there is an affirmative defense available when, as a result of severe mental disease or defect, the actor did not know the conduct was wrong, and that the defendant must prove it by a preponderance of the evidence.75 Texas recognizes no separate diminished-capacity defense; evidence of mental illness aimed at the culpable mental state is treated as a failure of proof on mens rea.76 A defendant whose severe disease caused the offense but who knew the conduct was wrong therefore has no insanity defense in Texas, while the same facts would raise the causal question under New Hampshire's test. Neither result raises a constitutional problem, which is why two such different rules can coexist.
On an exam problem, the first step is to identify which formulation the jurisdiction uses, because the product inquiry applies today only in New Hampshire among the states, and a problem set anywhere else calls for a capacity analysis or the federal formulation. Students lose points on this issue when they treat a test's rarity as a reason to call it unconstitutional or extinct. Rarity records how few legislatures and courts have chosen the test, and the Constitution asks nothing more of that choice.