This is a summary of The Effect of the Guilty but Mentally Ill Verdict on the Outcome of a Jury Trial
Author: Michael A. Eagan, Institution: Texas State University
Click here for the original version
This paper traces the insanity defense from its earliest roots through the modern Guilty but Mentally Ill (GBMI) verdict, then asks whether GBMI actually changes what juries decide.
The defense started long before McNaughton. Roman law excused people found non compos mentis, and English courts later ran three successive tests. The “good and evil” test, used from the 14th to 16th centuries, drew on religious ideas and asked whether a defendant could tell good from evil.
The “wild beast” test followed in 1724, after Rex v. Arnold, and shifted the question from moral failing to cognitive failing: could the defendant understand what he was doing at all? Regina v. Oxford in 1840 refined this further, asking whether the defendant understood the nature and consequences of his act.
Guilty but mentally ill
The modern standard came from the 1843 trial of Daniel McNaughton, who killed the prime minister’s secretary while trying to assassinate Robert Peel. The jury found him not guilty by reason of insanity after nine medical experts testified to his condition.
Public outrage, including from Queen Victoria, pushed the House of Lords to formalize the McNaughton Rule: defendants are presumed sane, and to be excused, they must show a defect of reason from disease of the mind that left them unable to know the nature of their act or that it was wrong. Critics called this rule too vague, since “wrong” could mean legally wrong, morally wrong, or something psychiatric, and courts never agreed on which.
Dissatisfaction with McNaughton produced several alternatives. The irresistible impulse test, adopted by many states in the 1920s, asked whether a mental disease removed the defendant’s willpower to resist an urge, even if he knew it was wrong. It fell out of favor because jurors had no reliable way to judge what counted as truly irresistible.
The Durham Rule
The Durham Rule, from a 1954 D.C. Circuit case, asked only whether the unlawful act was the product of mental disease or defect. Courts abandoned it in 1972 because it gave too much weight to psychiatric testimony and had no clear legal standard. The American Law Institute’s Model Penal Code test, created in 1955, combined a cognitive prong with an irresistible impulse prong, asking whether the defendant lacked substantial capacity to appreciate the wrongfulness of his conduct or to conform his conduct to the law.
This became the most widely adopted standard, used by 26 states plus D.C. Diminished capacity, a related but separate concept, doesn’t produce a not guilty verdict; it can only reduce a charge or sentence, and several states, including California, have abolished it outright.
The Hinckley case
John Hinckley’s 1982 acquittal for shooting Ronald Reagan reshaped the field again. Under the ALI standard then used federally, prosecutors had to prove sanity beyond a reasonable doubt, and since experts on both sides disagreed, the jury had to acquit. Public backlash led over half the states to rewrite their statutes back toward McNaughton, a few states abolished the insanity defense entirely, and Congress passed the Insanity Defense Reform Act of 1984, which made insanity an affirmative defense with the burden shifted to the defendant, who now has to prove insanity by clear and convincing evidence.
The Hinckley case also drove the spread of the Guilty but Mentally Ill verdict, first passed in Michigan in 1975. GBMI gives juries four options: guilty, not guilty, not guilty by reason of insanity, or guilty but mentally ill. A GBMI verdict requires proof of every element of the crime, proof of mental illness, and a finding that the defendant doesn’t meet the stricter McNaughton insanity standard. Sentencing under GBMI works exactly like a standard guilty verdict; treatment, if any, happens during incarceration, and the defendant serves out the full sentence regardless of whether treatment succeeds.
Core question
The paper’s core question is whether adding this fourth option changes trial outcomes. State-level data from Michigan, South Carolina, Georgia, and Illinois mostly shows no statistically significant drop in not-guilty-by-reason-of-insanity verdicts after GBMI statutes passed; Michigan’s NGRI rate actually rose. Pennsylvania and Georgia showed some decline, but confounding factors, like Pennsylvania’s simultaneous shift in burden of proof, made it hard to isolate GBMI’s effect. Mock trial research tells a different story.
A study using 140 undergraduates found a significant displacement of NGRI verdicts once GBMI was offered as an option. A follow-up with 327 participants, using the same case, found the GBMI option cut guilty verdicts by about two-thirds and NGRI verdicts by about half.
Insanity pleas are rare
The paper notes that insanity pleas are rare and usually unsuccessful. They come up in under 1% of felony cases and succeed only 15 to 25% of the time when raised. As of writing, 26 states plus the federal government use McNaughton, 20 states plus D.C. use the ALI test, and only three states use GBMI. Texas, notably, uses McNaughton with the irresistible impulse test and has no GBMI verdict.
The author concludes that real-world statistics don’t show a clear GBMI effect, but mock trial data and anecdotal patterns suggest a modest one exists. The bigger failure, per the paper, is that GBMI hasn’t delivered the treatment it promised: convicted defendants serve their full sentences whether or not their mental illness resolves. The paper calls for more real-world data collection and controlled mock trial designs that isolate the GBMI variable specifically.