This research piece is close to my heart, as I spent several years directing the Missouri Department of Mental Health’s (DMH) Forensic program, in which I was a liaison between the mental health system and the criminal justice system when questions of competency to stand trial, and responsibility at the time of the criminal offense (or the insanity defense) were raised in criminal proceedings. I quickly learned that the public’s perception of the use of the insanity defense, and even competency to stand trial, did not match up with the reality of what it meant to be found not guilty by reason of insanity. How to help bridge that education gap became one of my priorities, and when I joined our Department of Sociology and Criminal Justice Studies, my research interests continued around these individuals at the intersection of the criminal justice and mental health systems.
The current project centers around those individuals found Not Guilty by Reason of Insanity (known as NGRI – and I’ll use those initials throughout the rest of this discussion) for a criminal charge of murder in the State of Missouri. Let’s start out with some information about the insanity defense in general.
Most of the public has strong feelings about the insanity defense, as it is typically associated with a criminal defendant “getting off”, or not receiving the appropriate amount of punishment. Usually the insanity defense receives a great deal of public attention for high profile cases, such as John Hinckley (attempted assassin of Pres. Reagan); Jeffrey Dahmer (serial killer); Lee Boyd Malvo (the Washington DC sniper); or David Berkowitz (Son of Sam serial killer) – however, of these cases, ONLY John Hinckley successfully utilized the insanity defense. We may see the insanity defense utilized in the upcoming criminal proceedings for the Colorado theater killings matter.
Pleading insanity means that the defendant is incapable of forming the necessary criminal intent to commit a crime. Among other elements, in order to be found guilty of a crime, the defendant must have mens rea (or guilty mind – in other words, be capable of intending to do what was done). If the defendant, because of a severe mental illness, cannot form the requisite criminal intent and is incapable of knowing right from wrong, then that person can be found NGRI, and is usually committed to a locked mental health facility for varying lengths of time. In Missouri, for example, such a commitment is indefinite, and the person can only be released upon a judge’s order.
The insanity defense is not utilized very often – most research supports that it is ATTEMPTED in approximately 1% of all felony charges, and is SUCCESSFUL in only 1/4th of that 1% of cases (see research by Henry J. Steadman and others). That translates to a very small number of cases successfully pleading insanity! Yet that number is not what the public expects (see Michael Perlin’s work on the myths of the insanity defense).
When the insanity defense is attempted for a defendant accused of murder, it adds to the public outrage. The public wants to be assured that such NGRI acquittees are not released too quickly! This risk assessment issue is mirrored in the correctional system when there are questions of parole, or returning inmates to the community.
This public concern, or fear, really, in large part, resulted in the criminal justice system’s “get tough on crime” approach starting about 20 years ago. We see that reflected in “three strikes and you’re out” legislation; or mandatory minimum sentencing as examples. The “get tough on crime” approach was reflected in the mental health system when several states adopted a Guilty but Mentally Ill plea to be used in place of an insanity defense. This plea meant the person’s mental illness did make some contribution to the offense, but the individual should still serve their sentence in prison, not a mental health facility like NGRI acquittees. Other states, like Missouri, attempted to make it more difficult for NGRI acquittees to be released by changing the venue, or location, to file for such a court ordered release, and adding increased testimony requirements to state statutes governing such cases as of 1996.
Whether these “get tough on crime” initiatives affected NGRI acquittees prompted us to attempt to answer three research questions. First, were NGRI acquittees for murder different than those found NGRI for other crimes? Next, did crime severity affect the likelihood of being released from a mental health facility to the community – do those murder acquittees reside in the mental health hospital for longer periods of time? Finally, did the length of hospitalization change for murder and other offenses AFTER those legislative changes of 1996?
In order to answer those questions, we had access to 27 years of insanity acquittal data for the state of Missouri, giving us 1130 NGRI acquittees overall (those in the system between July 1, 1979 and June 30, 2007). Missouri is an interesting state to review, as research indicates that it ranks 6th among reporting states for the frequency of successful insanity pleas. Missouri has seen a decrease in insanity pleas after the 1996 legislation toughening up the release requirements. Using an existing dataset, we completed a secondary analysis of the data. Approximately thirteen percent of NGRI acquittees had been acquitted of murder, 50% for other violent crimes, and almost 37% for nonviolent offenses. To deal with some missing data, my co-author utilized a multiple imputation technique.
To answer research question #1, we compared the NGRI murder acquittees to other acquittees, and there were differences. Using multiple logistic regression, being female, and being acquitted in an urban county (compared to a rural county) INCREASED the odds of a NGRI murder acquittal. Those defendants never married, diagnosed with an OTHER AXIS I or psychiatric disorder, and having a greater number of previous psychiatric hospitalizations DECREASED the odds of a NGRI murder acquittal.
The answer to research question #2 was that NGRI murder acquittees were LESS likely to obtain conditional releases to the community than those found NGRI for other crimes. When they did obtain such releases, it was after they had been hospitalized substantially longer than those acquitted of other offenses. HOWEVER, for acquittees who had NEVER been released, there was no statistical difference in current length of hospitalization.
As to research question #3, or did the length of hospitalization change for murder and other offenses AFTER those legislative changes in Missouri, the answer was different than anticipated. Those NGRI murder acquittees were hospitalized, on average, MUCH longer after the legislative changes compared to other acquittees, BUT lengths of hospitalization INCREASED SIMILARLY for ALL crime categories – in otherwords, all NGRI acquittees, regardless of committing crime, were hospitalized substantially longer AFTER the legislative changes.
So, what does this all mean? Since gender (female) increased the odds of a NGRI murder acquittal, we looked to criminal justice system trends with female offenders. While the number of female offenders in the criminal justice (CJ) system incarcerated for murder has DECREASED, that is not the case for female NGRI murder acquittees – female NGRI murder acquittees account for 22% of ALL female NGRI acquittees. So, are women more favorably treated in sentencing, if the idea is that an insanity acquittal is a lesser disposition? This may be an example of potential system bias towards gender (and should be the focus of additional research).
Since there were really no clinical differences between NGRI murder acquittees and all other acquittees, that may mean that the seriousness of the mental illness is considered in assessing whether ALL acquittees met the statutory test for this defense. Fewer previous psychiatric hospitalizations may mean that successful NGRI murder acquittees don’t seek mental health treatment PRIOR to the murder, and without treatment those symptoms may have exacerbated, or increased, and led to the murder itself. The NGRI murder acquittees were less likely to have a prior felony conviction, and this may mean that those who have been convicted before may face a HIGHER burden to show that they did not know right from wrong since they had past interactions with the CJ system. As to release, NGRI murder acquittees may have to demonstrate longer periods of psychiatric stability and nondangerous behavior prior to release, and that crime seriousness is an important factor for potential release decisions.
What about the get tough on crime initiatives? Since lengths of hospitalization increased post legislation for ALL NGRI acquittees, it may mean that the increasingly conservative get tough on crime attitudes have, in fact, affected ALL NGRI acquittees. So while the get tough on crime initiatives may not have had as punitive or deterrent an effect on correctional sentencing practices, it has, perhaps unintended, affected the mental health system. Why does this matter? Because in an era of decreasing budgets, such increasing lengths of stay also increase mental health system operating costs (for example, in Missouri, 44% of long-term inpatient beds are occupied by such forensic clients) AND may not truly solve the public’s dilemma with defendants who also have mental illness. This unintended consequence may lead to additional criminalization of the mentally ill. It is hoped that fear should not drive policies and practices of either the criminal justice or mental health system, but instead that policies be grounded in research. Research in this area is fascinating, and working with these questions really never seem like work at all.
Read Related Articles
What Do We Mean when We Talk About Punishment?
The Hidden Costs of the Prison Boom for the Mental Health of Women
New Study Supports Link Between Inequality and Crime
Social Science is Changing how to view Policing