The Volume and Characteristics of Insanity Defense Pleas: An Eight-State Study
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
The Volume and Characteristics
of Insanity Defense Pleas: An
Eight-State Study
Lisa A. Callahan, PhD; Henry J. Steadman, PhD; Margaret A. McGreevy, MA;
and Pamela Clark Robbins, BA
The authors document the very complex process involved in identifying insanity
defense pleas in eight states. Each state and each study county in each state
required an individual approach. Most often, county court dockets were hand
searched to identify those pleading insanity, although numerous other methodol-
ogies were used. The frequency and rate of insanity pleas and acquittals are
presented for the study states as well as descriptive data on the characteristics of
persons pleading and acquitted NGRI. Overall, the insanity defense was raised in
one percent of all felony cases. Further, only 26 percent of those raising the insanity
defense were actually acquitted NGRI. The necessity of obtaining data on insanity
pleas to adequately understand and ultimately inform future directions of insanity
defense research is discussed.
Although a substantial body of knowl- pleas is certainly not due to a lack of
edge about persons acquitted by reason interest or i m p ~ r t a n c eRather,
.~ this gap
of insanity has been developed in the appears to result primarily from one ma-
last 15 years, there is surprisingly little jor practical problem-data on insanity
information about persons who plead pleas are not centrally or systematically
"not guilty by reason of insanity" maintained.
(NGRI). Very few researchers have stud- The use of acquittal data has been
ied insanity pleas, and the studies on facilitated by its reasonable accessibility.
pleas that are available offer limited in- Fairly comprehensive information on
formation.'-' There has not been one insanity acquittals is available because
cross-jurisdictional study on insanity most persons found NGRI are commit-
pleas. Even some articles entitled "insan- ted to state mental health facilities for
ity pleas" are based almost exclusively evaluation and/or treatment. As a result.
on acquittal data, rather than plea there is often a state-level information
data.8.' The omission of data on insanity system with data on all persons acquit-
ted NGRI. Such is not the case for NGRI
Dr. Callahan is affiliated with Russell Sage College,
Troy, NY 12180. Dr. Steadman, Ms. McGreevy, and pleas. To obtain information on defend-
Ms. Clark Robbins are affiliated with Policy Research ants raising the insanity defense, county
Associates, Inc., 262 Delaware Avenue, Delmar, NY
12054. court records must be accessed, a process
Bull Am Acad Psychiatry Law, Vol. 19, No. 4, 1991 33 1Callahan et a/.
that is expensive and extremely time- the affirmative defense (Montana-1 979)
consuming. and enactment of a GBMI verdict
Certainly the biggest gap in our un- (Georgia-1982). In addition to the five
derstanding of the full process of plead- states that made one or more of the
ing and/or being acquitted by reason of above reforms, three states that made no
insanity is with unsuccessful insanity pleas. alterations in their insanity defense stat-
Recently, McGinley and Pasewark8 utes from 1979 to 1984 (New Jersey,
called attention to this deficit in their Washington, and Wisconsin) were se-
national survey of the 5 1 United States lected as comparison states.
forensic directors. Because no statewide data existed on
The data reported here provide a the frequency of insanity pleas in any of
cross-jurisdictional look at the volume the eight jurisdictions, we selected sam-
and composition of insanity pleas and ple counties based on their number of
acquittals. The data are drawn from an insanity acquittals. We selected suffi-
eight-state, NIMH funded study on in- cient counties to obtain 66 percent of all
sanity defense reform we have been con- insanity acquittals in each state. This
ducting since 1984. While the main fo- figure was chosen because it provided
cus of the study is the impact of specific information on the majority of those
reforms of insanity defense statutes, lo. ' ' utilizing the insanity plea without be-
the goal of this article is to provide a coming prohibitively expensive. We
descriptive overview of the volume, achieved our goal in all but one state. In
rates, and composition of insanity pleas Georgia, we selected counties producing
and acquittals across states. Since these 60 percent of their acquittals, since ob-
multijurisdictional data are not available taining 66 percent would have required
in the research literature, we believe this us to collect data in 15 counties, a use
descriptive report fills a crucial gap in of resources we did not have. Altogether,
the understanding of the insanity de- we selected 49 counties in the eight
fense. states.
Research Design and The initial stage of our research was
Methodology to identify all criminal defendants who
The overall objective of the research entered the insanity plea, at any time,
was to assess the impact of various types during their defense in all study coun-
of insanity defense reform. Specific re- ties. Once the insanity pleas were iden-
forms studied include changes in the tified, we abstracted information from
insanity test (California- 1982). the bur- the criminal case records. Defendants
den and standard of proof (Georgia- found NGRI were followed through the
1978 and New York- l984), the court of state mental health departments, and
jurisdiction (Ohio- 1980), and commit- those found guilty were followed
ment and release procedures (New through the departments of corrections.
York-1980), as well as the abolition of We did not attempt to follow any de-
332 Bull Am Acad Psychiatry Law, Vol. 19, No. 4, 1991Insanity Defense Pleas
fendants who were released into the sulted in pulling and reviewing 89,554
community after trial. individual case files. Another procedure
Most of the data collection was com- for identifying insanity pleas was used
pleted by local researchers who were su- in five New York counties where the
pervised by state research coordinators. dockets contained none of the necessary
These researchers and supervisors were information. This task required the re-
trained on-site by the Project Director view of all cases where a fitness to pro-
and Assistant Project Director from Al- ceed (i.e., incompetent to stand trial or
bany, New York, with continual contact IST) exam was ordered. Since as an IST
throughout the data collection process. exam is often the first step to an NGRI
Procedures plea in New York, we anticipated that
Insanity Pleas The process of gath- we would capture most cases where an
ering information about insanity pleas insanity plea was used. We were able to
was extremely complicated. We exam- rely on a computerized search in three
ined nearly 1 million indictments to find counties and in one county we were
8,979 insanity pleas. given the indictment number of all de-
The most common method of identi- fendants who were evaluated for crimi-
fying cases was to hand search the indi- nal responsibility.
vidual criminal dockets that were main- Once a case was selected by the above
tained in the county clerks' offices. This procedures, the file was pulled. The ini-
entailed reviewing every docket page of tial review of these files was to determine
every indictment for the inclusive years, if an insanity plea was ever entered. If
searching for any reference to an insan- documentation of a plea existed, the case
ity defense. This procedure led to over- then became a study case. Documenta-
sampling as we initially included any tion of an insanity plea ranged from a
cases with a reference to the mental formal notice or motion to rely on the
health of the defendant (e.g., "NGRI," defense to a notation of the plea in the
"expert," "mental illness," "psychia- case minutes. A data abstract form was
trist," "state hospital," "exam," "IST"). then completed by the field researchers.
This procedure required the field re- The form included sociodemographics,
searchers to review dockets for 580,720 target crimes, criminal justice process-
indictments. In those counties where the ing, diagnoses, known prior criminal
dockets were not available, we relied justice and mental health histories, tar-
upon other techniques for identifying get confinement, and release informa-
insanity cases. tion. As much data as possible were
In five counties, the researchers pulled collected at the county level. The data
every case file during the study years were then completed at the facility or
from the shelves or file drawers and re- facilities where the defendant was con-
viewed each to determine whether an fined and/or at the centralized infor-
insanity plea was ever raised. This re- mation center.
Bull Am Acad Psychiatry Law, Vol. 19, No. 4, 1991 333Callahan et a/.
Insanity Acquittals Findings
Although most states have a central- The frequency and rate (per 100 fel-
ized information system for persons ac- ony indictments) of insanity pleas and
quitted NGRI, case records on NGRIs acquittals are presented in Table 1.
committed to state mental health sys- Across the 49 study counties in the eight
tems are not necessarily maintained in states, the insanity defense was raised in
one location. In California we were able approximately one percent of all felony
to obtain most of our information cases (0.93%). There was wide variation
through a number of customized com- in the proportion of defendants who
puter reports from the State Depart- used an insanity defense, from a high of
ments of Mental Health. In New York 5.74% in Montana, despite its abolition
we used centralized computer records of the affirmative insanity defense in
and paper files. The most common pro- 1979," to a low in New York of 0.30%.
New York's low rate is probably due to
cedure for collecting follow-up mental
the fact that in New York City less in-
health data was to go to the actual facil-
formation is recorded early in the de-
ities. In Montana and Georgia this in-
fense process, making it nearly impossi-
volved going to only one facility in each
ble to identify those cases where an
state and two each in Washington and
NGRI plea was entered and later with-
Wisconsin. In New Jersey we collected
drawn. Our multijurisdictional findings
data in five state facilities, often return-
on the plea rate are consistent with Jan-
ing many times to locate the complete
ofsky and colleagues2recent data where
case file. In Ohio we went to 12 state they found a one-year plea rate of 1.2
psychiatric centers and obtained infor- percent of felony indictments in Balti-
mation over the telephone from two more City.
other facilities. Overall, the acquittal rate (acquittals/
Unsuccessful Insanity Pleas The pleas) across the eight study states was
data collection procedures for unsuc- 26 percent, ranging from 87 percent in
cessful insanity pleas were simplified by Washington to 7.3 percent in Montana.
the fact that most state departments of Montana's low acquittal rate undoubt-
corrections maintain centralized paper edly reflects the fact that in 1979 they
or computerized records. Generally, we abolished the affirmative insanity de-
were able to request information on all fense. The high acquittal rate in Wash-
of our cases and receive the data directly ington is probably an indication that
from state officials. In some states we insanity pleas are somehow negotiated
relied on centralized paper records as before they are entered. Across the study
well as case records in the prisons. In states there tended to be an inverse re-
Montana we collected data at their one lationship between plea and acquittal
prison, while in Ohio data collection was rates. That is, states with high plea rates
completed at the central ofice as well as had lower acquittal rates, while those
at seven prisons. with low plea rates had higher acquittal
334 Bull Am Acad Psychiatry Law, Vol. 19, No. 4, 1991Insanity Defense Pleas
Table 1
Volume of lnsanity Cases by Study State
No. of Study Felony lnsanity NGRI'
Plea2 Acquittav
Counties Study Years Indictments Pleas Acquittals
Rate Rate
(n) (n) (n) (n)
California 7 7/78-6187 225,152 1,300 665 .58 45.52
Georgia 1176-1 2/85 151,669 2,630 426 1.73 13.11
Montana 1176-1 2/85 14,227 816 58 5.74 7.31
New Jersey 1176-1 2/85 125,951 670 295 53 43.34
New York 10177-9/87 195,015 556 226 .30 39.78
Ohio 1177-1 2/83 147,477 2,005 342 1.36 15.30
Washington 7179-1 2/87 74,105 442 387 .60 87.36
Wisconsin 7/79-6185 33,613 534 156 1.59 28.24
Total 967,209 8,953 2.555 .93 26.27
'NGRI acquittals includes all acquittals identified through county records as well as acquittals (in study counties)
who were identified via state level records (state hospitals).
Plea rate is per 100 felony indictments.
Acquittal rate is the percentage of NGRI pleas that result in acquittal. It is based only on data obtained through
county level records (not used in table), and does not include acquittals identified through state records.
rates. This relationship may suggest that NGRI but found guilty were released
there is some "acceptable" range of ac- following conviction.
quittals such that a high volume of pleas Table 2 presents a descriptive profile
is offset by a relatively low acquittal rate of all persons pleading insanity and
and vice versa. those acquitted NGRI across the eight
An unexpected finding, across juris- states. Although there was variation
dictions, was that some defendants were among the states, these data provide a
found NGRI without ever entering an basic sense of the differences existing
NGRI plea. In fact, approximately 15 between those raising the insanity de-
percent of all insanity acquittees never fense and those acquitted NGRI. A com-
actually plead NGRI. Of these, most had parative look at the sociodemographic
plead not guilty. So, it is misleading to profiles of the two groups support much
assume that an NGRI acquittal is always of the earlier work that found persons
preceded by an NGRI plea. successful in an NGRI plea tend to be
Likewise, it is important to emphasize older, female, better educated, and single
that not all persons who were unsuccess- than those raising the d e f e n ~ e . ~The
.~
ful in their insanity plea were convicted. mean age (not shown in Table 2) was
Approximately 10% of those pleading 30.3 years for insanity pleas and 32.1
insanity were discharged, withdrawn, or years for those acquitted.
found not guilty, while 64% were found Because of the large volume of cases,
guilty and 26% were acquitted NGRI. comparisons between the two popula-
Furthermore, even those who were tions were statistically significant for all
found guilty did not necessarily go to major variables. Differences in the so-
prison. We found that approximately 28 ciodemographic characteristics of the
percent of persons initially pleading two groups were relatively minor. The
Bull Am Acad Psychiatry Law, Vol. 19, No. 4, 1991 335Callahan et a/.
Table 2
Characteristics of Defendants Pleading Insanity and Those Acquitted NGRI in 49 Study Counties
in Eight States'
Insanity Pleas Insanity Acquittals
Descriptor Frequency Frequency percent
(N = 8,979) Percent (N = 2,565)
Gender
Male
Female
Ethnicity
White
Black
Other minority
Diagnosis
Schizophrenia
Other major MI
Mental retardation
Personality disorder
Substance abuse
Other mental illness
Not mentally ill
Target offense
Murder
Physical assault
Other violent
Robbery
Property
Other minor
Prior history
No hospital
One or more hosps.
No arrests
One or more arrests
Type of trial
Judge
Jury
Plea bargain
' Due to the large volume of cases, differences between those pleading insanity and those acquitted were statistically
significant on all the above variables.
differences seen between the two groups pleading insanity were charged with
on diagnoses, target offense, prior his- murder, physical assault, or other vio-
tory, and type of trial were more sub- lent offenses, 65 percent of those acquit-
stantial. Just over one-half (55.2%) of ted were indicted for these violent or
those pleading NGRI were diagnosed potentially violent offenses. Insanity ac-
with schizophrenia or another major quittals were also more likely to have
mental illness (other psychosis or affec- had a prior hospitalization and to have
tive disorder) while 84 percent of those been adjudicated by bench trial (judge)
acquitted carried such a diagnoses. Sim- rather than by a jury or plea bargain
ilarly, while one-half (50%) of those than other defendants raising the plea.
336 Bull Am Acad Psychiatry Law, Vol. 19, No. 4, 1991Insanity Defense Pleas
Discussion mental illness or who claim temporary
Our research underscores the impor- insanity is clearly untrue.
tance of gathering data on all persons The range of offenses for those plead-
pleading insanity for a comprehensive ing insanity was broader than for those
understanding of the insanity defense. actually NGRI. Approximately half of
At the same time it clearly illustrates just those pleading insanity were indicted for
how difficult and time-consuming it is violent or potentially violent offenses
while the other half were charged with
to obtain such data. The data collection
robbery, property, or minor felonies. In
phase of the project preceeded, uninter-
comparison, those acquitted NGRI
rupted, from October 1985 to June
tended to have committed more serious
1990. Much of the information was dif-
offenses. Although a distinct minority
ficult to obtain. It was especially difficult
were charged with murder (l4.8%),most
to get clinical information on those un-
had committed a violent or potentially
successfully raising the plea, and data on violent offense.
prior criminal involvement were often Finally, these data indicate that the
not available for the NGRI population. decision to acquit someone was seldom
Our data underscore the rarity with made by a jury; only seven percent of
which the insanity defense is used, as it 2,500 acquittals were disposed of by a
occurs in approximately one percent of jury. Rather. the decision to acquit was
all felony '* Seven of the eight made by other key players in the crimi-
states we studied had just such a plea nal justice process including the prose-
rate. Although there was considerable cutor, defense attorney, and judge. It's
variation among the eight states in the clear that negotiation plays a central role
acquittal rate (percentage of successful in the insanity plea, just as it does in
pleas), overall, just one-quarter of those other areas of the criminal justice proc-
who raised the defense were successful. ess.
These data highlight that the insanity Acknowledgment
defense is raised very infrequently and is This research was funded in part by PHS grant MH
38329 from the Antisocial and Violent Behavior Branch
not often successful when it is raised. of the National Institute of Mental Health. We would
Our data clearly illustrate that the vast like to thank the numerous county and state officials
who cooperated with us in this research. We would also
majority of people who used the insanity like to express appreciation to Carmen Cirincione and
Eric Silver for management of a very complex data set.
defense were seriously mentally ill. Only
10 percent of the population raising the
References
defense did not receive a DSM-111 diag-
1. Boehnert C: Characteristics of successful and
nosis, and the large majority had a prior unsuccessful insanity pleas. Law Hum Behav
hospitalization. Furthermore, only the 13:3 1-9, 1989
2. Janofsky JS. Vandewalle MB, Rappeport JR:
most disturbed defendants were success- Defendants pleading insanity: An analysis of
ful in their plea. The popular concept outcome. Bull A m Acad Psychiatry Law
l7:2O3- 1 1, 1989
that the insanity defense is an "easy out" 3. Jeffrey RW, Pasewark RA, Beiber S: Insanity
for defendants who are either feigning plea: Predicting not guilty by reason of insan-
Bull Am Acad Psychiatry Law, Vol. 19, No. 4, 1991 337Callahan et a/.
ity adjudications. Bull Am Acad Psychiatry of the frequency and success of the insanity
Law 16:35-9. 1988 plea and alternate pleas. J Psychiatry Law
4. Phillips MR. Wolf AS. Coors DJ: Psychiatry l7:2O5-2 1, 1989
and the criminal justice system: Testing the 9. Hawkins MR. Pasewark RA: Characteristics
myths. Am J Psychiatry 145:605-10, 1988 of persons utilizing the insanity plea. Psychol
5. Packer IK. Homicide and the insanity de- Rep 53:191-5, 1983
fense: A comparison of sane and insane mur- 10. Steadman HJ. Callahan LA, Robbins PC.
derers. Behav Sci Law 5:25-36, 1987 Morrissey JP: The maintenance of an insan-
ity defense under Montana's abolition. Am J
6. Pasewark RA, Jeffrey RW, Beiber S: Differ- Psychiatry 146:357-60, 1989
entiating successful and unsuccessful insanity 1I. Callahan LA, Steadman HJ. Robbins PC,
plea defendants in Colorado. J Psychiatry Momssey JP: The impact of Montana's in-
Law 1 5 5 - 7 1 , 1987 sanity defense abolition. Psychiatr Q. in
7. Steadman HJ, Keitner L, Braff J, Arvanites press. 1991
TM: Factors associated with a successful in- 12. Pasewark RA. Seidenzahl D: Opinions con-
sanity plea. Am J Psychiatry 140:401-5, cerning the insanity plea and criminality
1983 among mental patients. Bull Am Acad Psy-
8. McGinley H. Pasewark RA: National survey chiatry Law 7: 199-202. 1979
Bull Am Acad Psychiatry Law, Vol. 19, No. 4, 1991You can also read
