A Paradox of Bias: Racial Differences in Forensic Psychiatric Diagnosis and Determinations of...
-
Upload
timothy-allen -
Category
Documents
-
view
213 -
download
1
Transcript of A Paradox of Bias: Racial Differences in Forensic Psychiatric Diagnosis and Determinations of...
A Paradox of Bias: Racial Differences in Forensic PsychiatricDiagnosis and Determinations of Criminal Responsibility
Brea L. Perry • Matthew Neltner • Timothy Allen
� Springer Science+Business Media New York 2013
Abstract Although there is substantial evidence that
African Americans receive unequal treatment in both the
healthcare and criminal justice systems, less research has
investigated the role of race when these two systems con-
verge. Here, we examine the influence of race on patterns
of forensic psychiatric diagnosis and determinations of
criminal responsibility in pre-trial correctional facilities
(e.g., forensic psychiatric hospitals). Data are from a
medical chart review of 129 randomly selected competency
evaluations that occurred in a pre-trial correctional psy-
chiatric facility. Consistent with previous research, findings
indicate that African Americans are disproportionately
diagnosed with highly stigmatized psychotic spectrum
disorders relative to whites. In addition, they unexpectedly
indicate that African Americans are significantly more
likely than whites to be found not criminally responsible by
the court-appointed evaluating mental health professional,
controlling for sociodemographic characteristics, number
of violent and non-violent charges, and other potential
confounding variables. Mediation analysis reveals the
important and previously undocumented finding that the
effect of race on criminal responsibility determinations is
fully mediated by differential diagnosis. This suggests that
patterns of racial inequality and potential bias in the
diagnostic process may confer medical resources and other
benefits for African Americans in the context of the
criminal justice system.
Keywords Race � African American � Psychiatric
diagnosis � Criminal justice � Forensic psychiatry � Not
guilty by reason of insanity (NGRI)
Introduction
Racial and ethnic inequality is evident in both the Ameri-
can healthcare system and criminal justice system. With
respect to health care, research suggests that African
Americans have less access to health services and tend to
receive delayed treatment and lower quality acute and
long-term care than whites (Wright and Perry 2010;
Smedley et al. 2002; Williams and Rucker 2000). Dispar-
ities are particularly pronounced in the area of psychiatric
treatment, with documented differences in treatment-
seeking, barriers to receiving care, higher likelihood of
involuntary hospitalization, and provider bias that affects
clinician–patient interactions and treatment outcomes
(Segal et al. 1996; Snowden 1999; Snowden and Pingitore
2002; van Ryn and Burke 2000). Of central concern for this
analysis are racial disparities in diagnosis, wherein African
Americans are disproportionately likely to be diagnosed
with a psychotic disorder (Blow et al. 2004; Neighbors
et al. 1999; Strakowski et al. 2003). In general, undiag-
nosed or misdiagnosed mental illness among racial and
ethnic minorities is a major public health concern as it
results in worse acute and long-term outcomes for those
affected (Wang et al. 2005). In addition, because psychotic
disorders are more highly stigmatized by the American
B. L. Perry (&)
Department of Sociology, University of Kentucky, Lexington,
KY, USA
e-mail: [email protected]
M. Neltner
University Health Service, University of Kentucky, Lexington,
KY, USA
T. Allen
Department of Psychiatry, University of Kentucky, Lexington,
KY, USA
123
Race Soc Probl
DOI 10.1007/s12552-013-9100-3
public, this pattern of misdiagnosis has critical implications
for the social status and life chances of labeled individuals
(Phelan et al. 2000; van Dorn et al. 2005).
In recent decades, the mental health and criminal justice
systems have become increasingly intertwined, with
greater numbers of people with serious mental illness now
being detained and treated in correctional facilities than in
psychiatric hospitals (Lamb and Weinberger 2001). Since
the deinstitutionalization movement began in the 1960s,
there has been a sharp reduction in in-patient psychiatric
hospital capacity that poses barriers to providing appro-
priate long-term care for the most severely impaired and
socioeconomically disadvantaged patients (Ehrenkranz
2001; Lamb and Bachrach 2001). Consequently, there has
been an increase in untreated mental illness in urban areas,
leading to homelessness, crime, and arrests (Markowitz
2006; Mechanic and Rochefort 1990). This trend has been
termed the criminalization of mental illness (Abramson
1972).
The mental health and criminal justice systems converge
most overtly in psychiatric pre-trial correctional facilities
and mental health courts. Criminal defendants are evalu-
ated by psychiatrists and psychologists for competency to
stand trial and criminal responsibility in cases where
mental illness is suspected. In a very small minority of
cases, individuals with mental illness may be determined
not guilty by reason of insanity (NGRI), meaning (in many
jurisdictions) that ‘‘…the defendant, as a result of severe
mental disease or defect, was unable to appreciate the
nature and quality or the wrongfulness of his acts’’ (Federal
Insanity Defense Reform Act, 1984). In such cases,
defendants are either released from custody or committed
to a psychiatric facility rather than a prison and released
when they pose no further threat rather than serving out a
mandated sentence in prison. An NGRI finding is a pref-
erable legal outcome in most cases since social conditions
and access to mental health treatment are substantially
better in psychiatric hospitals relative to prisons (Baillar-
geon et al. 2010; Birmingham 2003; Forrester et al. 2010).
On the whole, African Americans are disadvantaged
across nearly all sectors of the criminal justice system—
from higher arrest rates to harsher sentencing (Higginbo-
tham 2002; Pettit and Western 2004; Roberts 2004). Par-
adoxically, there is modest evidence that African
Americans are more likely than whites to be found NGRI
(Poulson 1990). However, mechanisms underlying this
counter-trend are not well understood. More research is
needed to determine how race affects criminal responsi-
bility determinations in criminal courts and diagnostic
decisions in forensic psychiatric facilities, and whether
these are related.
Here, we use data from 129 pre-trial competency eval-
uations in a forensic psychiatric facility obtained through
retrospective chart review. We examine whether African
American patients were more likely to be diagnosed with a
psychotic disorder and to be recommended not criminally
responsible relative to whites by the court-appointed
evaluator. Then, we assess whether the impact of race on
criminal responsibility determinations works through racial
disparities in diagnosis, conferring additional medical care
and other resources for African Americans in this small
sector of the criminal justice system.
Racial Disparities in Psychiatric Diagnosis
and Treatment
African Americans are about three to four times as likely as
whites to be diagnosed with psychotic disorders such as
schizophrenia (Blow et al. 2004), and only about a third of
the effect of race can be explained by socioeconomic status
(SES) differences across racial groups (Bresnahan et al.
2007). A substantial proportion of racial disparities in
psychiatric diagnosis reflects real differences in the inci-
dence of disorder and is likely attributable to social epi-
demiological factors such as racism and segregation into
impoverished neighborhoods (Williams 1999; Williams
and Jackson 2005). However, another source is provider
behavior and clinical decision-making (van Ryn 2002; van
Ryn and Fu 2003), which may artificially inflate rates of
psychotic disorders among African Americans.
Substantial evidence indicates that patient race signifi-
cantly affects psychiatric diagnosis such that African
Americans are more likely to receive a diagnosis of
schizophrenia and less likely to be diagnosed with
depression than whites similar on relevant characteristics
(DelBello et al. 2001; Kales et al. 2000; Lawson et al.
1994; Loring and Powell 1988; Neighbors et al. 1999;
Raybur and Stonecypher 1996; Takei et al. 1998; Trier-
weiler et al. 2000). For instance, Strakowski et al. (2003)
found that African American men diagnosed with affective
disorder by expert consensus were significantly more likely
than other patients to be diagnosed with a schizophrenia
spectrum disorder by clinical assessment and structured
interview. Research across a variety of settings and sam-
ples indicates that African Americans are 10–40 % more
likely to be diagnosed with psychotic spectrum disorders
than whites and other comparison groups (for reviews, see
Adebimpe 1981 and Neighbors et al. 1999).
Several potential explanations for over-diagnosis of
psychotic spectrum illness among African Americans have
been offered. It may be that diagnostic criteria are biased
and ethnocentric, making the DSM a less valid and reliable
diagnostic tool for some racial or ethnic groups relative to
others (Funtowicz and Widiger 1995; Widiger and Spitzer
1991). Alternatively, the application of diagnostic criteria
to different racial or ethnic groups by clinicians may be
Race Soc Probl
123
unintentionally biased. For example, there is evidence that
clinicians differentially attribute and weigh various symp-
toms (e.g., hallucinations, paranoia, and elevated mood) in
making diagnostic decisions about African Americans
versus other groups (Trierweiler et al. 2000). These pat-
terns can be the result of patients’ cultural mistrust of the
medical system, ineffective communication and weak
therapeutic alliance between clinicians and patients, or
clinicians’ cultural misunderstandings and racial prejudices
(Neighbors et al. 1999; Snowden 2003; Vasquez 2007).
Although clinicians are expected to objectively weigh
biomedical and behavioral evidence in making diagnostic
and treatment decisions, disregarding race, gender, socio-
economic status, and other sociodemographic factors, bias
may be unavoidable. Stereotyping and social categorization
are components of an adaptive cognitive strategy that helps
humans make sense of vast amounts of complex informa-
tion, increasing the speed and efficiency of cognitive pro-
cessing (Kunda 1999; Stangor 2000). Stereotypes about a
group may be applied to individuals during the clinical
encounter, unconsciously affecting beliefs and expectations
about patients (Lewis et al. 1990). For instance, van Ryn
and Burke (2000) found that race negatively influenced
clinicians’ ratings of patients’ intelligence, personality, and
adherence to treatment recommendations after controlling
for socioeconomic status, mental health status, gender, and
other patient and clinician characteristics. Studies find that
through this lens of racial or ethnic stereotypes, similar
patient behavior can be interpreted in very different ways,
particularly if the behavior is ambiguous (Dunning and
Sherman 1997; Lepore and Brown 1997; Sagar and
Schofield 1980; Trierweiler et al. 2000).
Other racial differences in psychiatric treatment have
been identified, some of which may contribute to differ-
ential diagnosis. Controlling for patient behavior and other
clinical factors, clinicians spend less time evaluating
African American patients relative to whites (Cooper et al.
2003; Segal et al. 1996). Consistent with disparities in
diagnosis, African Americans are more likely to be pre-
scribed antipsychotic medications (Dixon et al. 2001; Segal
et al. 1996). Also, African Americans are more likely to be
brought to psychiatric treatment by legal means, emer-
gency room use, and involuntary hospitalization (Akutsu
et al. 1996; Rosenfield 1984; Takeuchi and Cheung 1998;
Snowden and Cheung 1990; Snowden 1999), and less apt
to voluntarily seek or receive psychiatric treatment than
whites (Department of Health and Human Services 1999;
Snowden and Pingitore 2002; Wang et al. 2005).
Perceived dangerousness is the strongest factor in pre-
dicting support of forced treatment for mental illness
(Corrigan et al. 2003) and is closely linked to the presence
of psychotic symptoms by the American public, mental
health treatment providers, and family members of
individuals with psychotic disorders (Phelan et al. 2000;
van Dorn et al. 2005). Consequently, perceptions of dan-
gerousness and psychotic symptoms are associated both
with support for forced or coerced treatment and with more
stigmatizing attitudes and greater desire for social distance
from individuals with mental illness (Link et al. 1999;
Pescosolido et al. 2007; van Dorn et al. 2005; Watson et al.
2005). This research suggests that there may be a link
between misdiagnosis of psychotic disorders among Afri-
can Americans, perceived dangerousness of racial and
ethnic minorities with mental health problems, and invol-
untary entry into treatment through the criminal justice
system.
Racial Inequality in the Criminal Justice System
Racial and ethnic minorities experience inequality in the
criminal justice system (Maurer and King 2007), which has
become increasingly involved in detaining individuals with
serious mental illness (Lamb and Weinberger 2001). Pettit
and Western (2004) estimate that about twenty percent of
African American men are imprisoned by age 30 compared
to only three percent of white men in the same birth cohort.
African Americans, especially young men, are dispropor-
tionately likely to be both arrested and convicted of crimes
(Higginbotham 2002). Factors that contribute to this pat-
tern are family structure, living in areas of concentrated
poverty, low SES, and police and juror racial bias (Higg-
inbotham 2002; Kirk 2008). Once convicted, African
Americans receive harsher punishments than whites, with
disproportionate numbers of the minority group being
imprisoned for 1 year or more (Roberts 2004; Sweeney and
Haney 1992).
An area of the criminal justice system in which African
Americans may have an advantage is determinations of
criminal responsibility. In a very small minority of cases,
juries may determine that a defendant is not responsible for
his or her crimes due to mental disorder or defect, also
known as NGRI. One study suggests that jurors are more
likely to find a defendant NGRI if he or she is African
American (Poulson 1990). In addition, race has been shown
to affect jury determinations in other instances, for exam-
ple in cases where the jury is concerned that pre-trial
publicity is racist (Fein et al. 1997); the jury in such cases
tends to give the minority defendant more leeway to offset
discrimination. However, whether and how race influences
determinations of responsibility for criminal activity
among patients with mental illness is not well understood.
In cases where mental illness is suspected, defendants
are evaluated for psychiatric disorders that might affect
competency to stand trial and criminal responsibility
(Knoll and Resnick 2008). Attornies assigned to a case may
also request a psychiatric evaluation. Psychiatrists or
Race Soc Probl
123
psychologists provide observation evidence (descriptions
of behaviors and cognitions that provide evidence of
mental illness), mental disease evidence (psychiatric dis-
orders for which a defendant meets diagnostic criteria, if
any), and capacity evidence (whether the mental illness
reduced the defendant’s capability to perform mental pro-
cesses; Wortzel and Metzner 2006). The NGRI defense is
employed in \1 % of cases in the US court system and is
only successful in about one quarter of those cases (Silver
et al. 1994).
However, when a jury makes an NGRI determination,
defendants are often committed to a psychiatric facility for
an indeterminate period until they pose no further threat
rather than to a correctional facility for a pre-determined
sentence. Psychiatric facilities are preferred restrictive
environments since prisons provide comparatively fewer
and lower-quality mental health services, and are charac-
terized by conditions likely to exacerbate mental illness
(e.g., overcrowding, physical and emotional abuse by fel-
low prisoners and prison staff, social isolation; Baillargeon
et al. 2010; Birmingham 2003; Forrester et al. 2010).
Additionally, because inmates with mental illness detained
in prisons do not receive appropriate long-term mental
health treatment, their likelihood of recidivism and sub-
sequent re-arrest is high (Ditton 1999; NAMI 2004).
Despite the increasing role of mental health courts in the
criminal justice system and the sharp increase in mental ill-
ness among inmates in recent decades (Lamb and Wein-
berger 2001), little is known about how race or ethnicity
affect diagnostic determinations by clinicians in forensic
psychiatric facilities. Furthermore, whether and how race
influences determinations of responsibility for criminal
activity among pre-trial evaluees with mental illness is not
well understood. There is much at stake both in terms of
taxpayer burden and outcomes among prisoners with mental
illness. On average, in 2001, the annual cost per state inmate
was $22,650 (Stephan 2004), and costs are estimated to be
nearly fifty percent higher among inmates with serious
mental illness (Lovell et al. 2001). In addition, inmates with
mental health problems have higher recidivism rates, leading
to multiple imprisonments and progressively lengthy sen-
tences (Baillargeon et al. 2009). Since African Americans
are disadvantaged in both the criminal justice and mental
health treatment systems, understanding how race affects
diagnostic and legal outcomes when these systems converge
is critical.
The present study is of 129 pre-trial evaluees referred to
a forensic psychiatric facility for competency, criminal
responsibility evaluation, and/or psychiatric treatment.
During an initial examination of these data, we unexpect-
edly found that white pre-trial evaluees were significantly
more likely to be determined to be responsible for their
crimes by mental health evaluators than were African
Americans. The purpose of this study is to determine
whether differential diagnosis with psychotic mental illness
(e.g., schizophrenia, schizoaffective disorder, psychosis not
otherwise specified) in the forensic psychiatric facility
partially or fully accounts for the racial discrepancy in
being found responsible for crime.
Methods
Patients were sent to the Kentucky Correctional Psychiatric
Center (KCPC) by court order. Transfer to KCPC is typically
requested by Defense Counsel for competency to stand trial
and criminal responsibility evaluations. By definition, these
patients were thought to be cognitively or emotionally
impaired by a non-clinician prior to their referral. Medical
charts were selected randomly for review from all KCPC
discharges in 20061. A database was created recording
information from the medical chart, including sociodemo-
graphic information, criminal and psychiatric history, Axis
I–III diagnosis, and evaluator determinations of criminal
responsibility. Because the charts represent historical data,
and no identifying information was recorded, the Institu-
tional Review Board at the University of Kentucky waived
the informed consent requirement.
While 194 charts were selected for review, only 131 of
these contained information about criminal responsibility
determinations. Though efforts were made to contact
KCPC to obtain this information, it was not recorded in
either patient charts or the state’s electronic database in 63
cases. In addition, two cases were listed as ‘‘Hispanic’’
with no indication of race. These 65 cases (34 % of the
sample) were dropped from all analyses. Concerns about
the bias this potentially introduces are minimal since a
comparison of cases with and without missing data reveal
no significant differences by gender, race, socioeconomic
status, psychiatric diagnosis, or other study variables.
Measures
Sociodemographic variables are included in multivariate
models as controls. These include gender (1 = female;
0 = male) and race (1 = white; 0 = African American).
Age and educational attainment are measured in years. Two
1 According to statistics from the Kentucky Justice and Public Safety
Cabinet (2007), 23 % of all individuals arrested in Kentucky between
2003 and 2007 were African American. This figure is very similar to
the distribution of African Americans in our randomly selected
sample (22 %). However, this figure is substantially higher than the
overall percentage of the Kentucky population that is African
American (8 %). This suggests that African Americans are dispro-
portionately likely to be arrested in Kentucky, but once arrested, they
are probably not disproportionately likely to be sent to a psychiatric
correctional facility for evaluation prior to standing trial.
Race Soc Probl
123
additional variables measuring diagnosis with a cognitive
impairment (e.g., mental retardation) or learning disability
(1 = yes; 0 = no) and history of illegal drug abuse or
dependence (1 = yes; 0 = no) are included in models as
independent variables. Also, two independent variables
measure treatment and behavior while at KCPC are included
in models. Length of stay is measured in days and a dichot-
omous variable indicates whether the patient was physically
restrained while at KCPC (1 = yes; 0 = no). Independent
variables measuring criminal history are separated into total
number of violent and non-violent prior convictions and
current charges. Alternative coding strategies (e.g., sepa-
rating previous convictions from current charges) did not
alter regression results and resulted in poorer model fit.
With respect to dependent variables, Axis I diagnosis is
measured as a series of dichotomous indicators representing
affective disorders, substance use disorders (SUDs), and
psychotic disorders. Affective disorders include major
depression, bipolar, anxiety, and adjustment disorders.
Psychotic disorders include schizophrenia, schizoaffective
disorder, dementia, and psychosis NOS. All of the evaluated
patients were diagnosed with a psychiatric disorder. Diag-
nostic categories are used as dependent variables in the first
set of models and independent variables in the second set.
Finally, a dichotomous dependent variable represents mental
health evaluator determinations of criminal responsibility
(1 = responsible; 0 = not responsible).
Analysis
Binary logistic regression is employed to model the effects
of race and other independent variables on patients’ odds of
being diagnosed with a psychotic disorder and their odds of
being found responsible for their crimes. To facilitate the
use of these results in meta-analyses, odds ratios for the
effects of race are converted to effects sizes and presented
in text (Chinn 2000). For each outcome, related groups of
variables are added in a stepwise fashion, resulting in four
restricted models and one full model with all covariates.
This strategy permits a preliminary assessment of media-
tion, which is then fully tested using the sgmediation
command in Stata with a bootstrapped estimation of the
indirect effect (MacKinnon and Dwyer 1993)—a method
that has been shown to produce less biased estimates than
the Baron and Kenny (1986) and Sobel (1986) methods in
simulation studies (MacKinnon et al. 1995).
Initially, multinomial logistic regression was employed
to predict a nominal diagnosis outcome (where
1 = affective disorder; 2 = psychotic disorder; 3 = sub-
stance use disorder) rather than a binary one (1 = psy-
chotic disorder; 0 = affective disorder or SUD). However,
a binary model is presented for the following reasons: (1)
the number of patients with an Axis I SUD is relatively
small, introducing estimation bias associated with small
cell size; (2) Wald tests did not identify statistically sig-
nificant differences in the effects of independent variables
on the odds of being diagnosed with an affective disorder
versus a SUD; (3) results regarding race and psychotic
disorders are the same whether patients with SUDs are
omitted or combined with affective disorders; and (4)
interpretation of binary models is more straightforward and
comprehensible by a broader audience of readers. Full
results are available upon request. Finally, multicollinearity
was assessed using variance inflation factors (VIFs). None
of the VIFs exceed 1.5, suggesting that the level of mul-
ticollinearity is unproblematic.
Results
Descriptive Findings
Sample descriptive statistics are presented in Table 1.
About 12 % of the sample is female, 78 % is white, and
22 % is African American. Mean age is 33.20, and mean
year of schooling is 10.19. About 22 % of patients in the
sample were diagnosed with cognitive impairment or a
learning disability, and 78 % had a history of abusing
illegal drugs. With respect to primary Axis I diagnosis,
32 % were diagnosed with a psychotic spectrum disorder,
58 % with an affective disorder, and 10 % with a substance
use disorder. The average length of stay in the forensic
psychiatric facility is 46.41 days, and 19 % of patients
were secluded or restrained at least once while at the
facility. The mean number of past convictions and current
violent charges against patients in the sample is 1.92, and
the mean number of non-violent convictions and charges is
12.00. Finally, 81 % of patients evaluated at the forensic
psychiatric facility were determined responsible for their
crimes by the court-appointed evaluator.
Multivariate Findings on Predictors of Diagnosis
The effects of sociodemographic and other independent
variables on diagnosis with a psychotic disorder are
depicted in Table 2. According to Model 1, white patients
are estimated to be 78 % less likely than African Ameri-
cans to be diagnosed with a psychotic disorder (p \ 0.01)
versus an affective disorder or SUD. This constitutes a
large effect size (d = 0.89; CI 0.32–1.46). In addition,
higher levels of education are associated with a reduction
in the odds of being diagnosed with a psychotic disorder
(OR = 0.73; p \ 0.01). Findings in Model 2 indicate that
neither cognitive impairment nor history of drug abuse has
a significant effect on diagnosis. However, as shown in
Model 3, length of stay in the forensic psychiatric facility is
Race Soc Probl
123
positively associated with the odds of being diagnosed with
a psychotic disorder (OR = 1.03; p \ 0.01). Neither
number of violent nor non-violent convictions and charges
significantly predict psychotic diagnosis (See Model 4).
The full model with all covariates is presented in Model
5 of Table 2. The effects of race, educational attainment,
and length of stay remain significant in the full model, and
coefficients are slightly larger or unchanged. This suggests
that the effects of these variables are not confounded or
mediated by criminal charges, treatment factors, cognitive
impairment, or drug abuse history. Holding covariates at
their means, the predicted probability that an African
American at the forensic psychiatric facility is diagnosed
with a psychotic disorder is 56 %, compared to only 21 %
for white patients. Likewise, the predicted probability that
a patient with an 8th grade education is diagnosed with a
psychotic disorder is 48 %, compared to 15 % for a patient
with a high school degree and only 4 % for a patient with a
college degree. In all, findings are consistent with previous
research, suggesting that there are substantial racial and
socioeconomic status disparities in diagnosis of psychotic
disorders.
Multivariate Findings on Predictors of Court
Determination
Results from the regression of criminal responsibility
determination by mental health evaluators on race and
other independent variables are presented in Table 3.
According to Model 1, white patients are estimated to be
Table 1 Sample descriptive statistics (n = 129)
Mean SD Range
Female 0.12
Race/ethnicity
White 0.78
African American 0.22
Age in years 33.20 10.53 18–75
Educational attainment in years 10.23 2.37 2–20
Cognitive or learning disability 0.29
History of drug abuse 0.78
Axis I diagnosis
Psychotic disorder 0.32
Affective disorder 0.58
Substance abuse/dependence 0.10
Length of stay in days 46.41 32.65 4–330
Was physically restrained 0.19
Number of violent charges 1.92 2.68 0–18
Number of non-violent charges 12.00 48.31 0–506
Found criminally responsible 0.81
Table 2 Binary logistic
regression of psychotic
diagnosis on independent
variables (n = 129)
Table presents odds ratios
(standard errors in parentheses);
two-tailed tests; *** p \ 0.001,
** p \ 0.01, * p \ 0.05
(1) (2) (3) (4) (5)
Female 1.13 1.05 0.84 1.00 0.66
(0.72) (0.70) (0.61) (0.66) (0.51)
White 0.22** 0.24** 0.20** 0.19*** 0.20**
(0.10) (0.12) (0.10) (0.09) (0.10)
Age in years 1.02 1.02 1.01 1.03 1.01
(0.02) (0.02) (0.02) (0.02) (0.03)
Educational attainment in years 0.73** 0.70** 0.74** 0.70** 0.67**
(0.08) (0.08) (0.08) (0.08) (0.09)
Cognitive or learning disability 0.59 0.45
(0.30) (0.26)
History of drug abuse 0.39 0.41
(0.20) (0.23)
Length of stay in days 1.03** 1.03**
(0.01) (0.01)
Was physically restrained 0.56 0.51
(0.33) (0.31)
Number of violent charges 0.93 0.94
(0.08) (0.09)
Number of non-violent charges 0.99 0.99
(0.00) (0.00)
Pseudo-R2 0.14 0.17 0.22 0.15 0.26
Likelihood ratio X2 22.64*** 26.90*** 35.46*** 24.82*** 41.56***
Race Soc Probl
123
nearly three times more likely than African Americans to
be determined responsible for their crimes (OR = 2.85;
p \ 0.05). This constitutes a moderate effect size
(d = 0.58; CI 0.01–1.14). In addition, being older is
associated with a reduced likelihood of being determined
criminally responsible (OR = 0.94; p \ 0.05). As shown
in Model 2, neither cognitive impairment nor drug abuse
history is significantly related to criminal responsibility
determinations. However, being diagnosed with a psy-
chotic disorder rather than an affective disorder or SUD is
strongly predictive of criminal responsibility such that
those with a psychotic diagnosis are 94 % less likely to be
determined criminally responsible than those with another
diagnosis (p \ 0.001). This constitutes a very large effect
size (d = 1.53; CI 0.79–2.23). In the model that includes
diagnosis, the effects of race and age are substantially
reduced and become non-significant, indicating a possible
mediating relationship. Results in Models 3 and 4 dem-
onstrate that length of stay, physical restraint, and charges
are not significantly related to criminal responsibility
evaluations.
Findings from the full model are presented in Model 5 of
Table 3. Here, age and psychotic disorder remain signifi-
cant, suggesting that these effects are not confounded by
any of the independent variables included in this analysis.
Number of violent past convictions and current charges
reaches statistical significance in the full model, as well,
such that each additional conviction/charge is associated
with a 70 % increase in the odds of being determined
criminally responsible (p \ 0.05). The predicted probabil-
ity of being determined responsible if diagnosed with an
affective disorder or SUD is 99 %, compared to only 73 %
for those diagnosed with a psychotic disorder. Also, the
predicted probability of being determined criminally
responsible for patients with no violent convictions or
charges is 91 %, compared to 97 % for patients with two
violent convictions or charges and 100 % for those with six
or more.
Race becomes non-significant in the full model and the
coefficient is substantially reduced compared to Model 1.
Results from bootstrapped estimation support the presence
of mediation, indicating that the indirect effect of race
through psychotic diagnosis is 0.23 (OR = 1.26; p \ 0.05)
and the direct effect is only 0.01 (non-significant). This
constitutes a small effect size (d = 0.13; CI 0.05–0.28). In
all, 95 % of the total effect of race on criminal responsi-
bility determination is mediated through diagnosis with a
psychotic disorder. African Americans are significantly
more likely to be diagnosed with a psychotic disorder,
which in turn substantially decreases their likelihood of
Table 3 Binary logistic
regression of criminal
responsibility on independent
variables (n = 129)
Table presents odds ratios
(standard errors in parentheses);
two-tailed tests; *** p \ 0.001,
** p \ 0.01, * p \ 0.05
(1) (2) (3) (4) (5)
Female 0.59 0.49 0.64 0.81 0.75
(0.49) (0.39) (0.44) (0.58) (0.68)
White 2.85* 1.04 2.91* 3.87* 1.07
(1.75) (0.67) (1.53) (2.21) (0.83)
Age in years 0.94* 0.96 0.94* 0.93** 0.92*
(0.02) (0.03) (0.02) (0.02) (0.03)
Educational attainment in years 1.03 0.90 1.01 1.06 0.87
(0.10) (0.11) (0.10) (0.11) (0.12)
Cognitive or learning disability 3.48 3.11
(3.08) (2.94)
History of drug abuse 1.17 1.02
(0.77) (0.72)
Psychotic diagnosis 0.06*** 0.03***
(0.04) (0.03)
Length of stay in days 0.99 1.00
(0.01) (0.01)
Was physically restrained 0.81 0.50
(0.52) (0.42)
Number of violent charges 1.44 1.70*
(0.29) (0.44)
Number of non-violent charges 1.08 1.11
(0.07) (0.09)
Pseudo-R2 0.10 0.31 0.11 0.17 0.41
Likelihood ratio X2 12.35* 38.56*** 13.42* 21.14** 50.53***
Race Soc Probl
123
being determined responsible for their crimes by a court-
appointed evaluator.
Discussion
Consistent with previous research (Poulson 1990), our
findings indicate that African Americans are less likely than
whites to be found responsible for their crimes by mental
health evaluators at a forensic hospital even after controlling
for socioeconomic status, violence, number of prior offenses,
and other factors related to responsibility determinations.
However, this relationship is fully mediated by higher levels
of diagnosis with psychotic spectrum disorders among
African Americans in the forensic psychiatric facility.
These data indicate that individuals with psychotic
symptoms are nearly 25 times more likely to be found not
criminally responsible than those with affective or substance
use disorders—a finding that is unsurprising given the cri-
teria for NGRI laid out by the Federal Insanity Defense
Reform Act and most state insanity statutes. Namely,
defendants must have a severe mental illness that prohibits
them from knowing that their actions were wrong (cognitive
prong) and/or prohibits them from conforming their behavior
to the requirements of the law (volitional prong). While
affective disorders can be severe, unlike psychotic spectrum
disorders they are not often accompanied by delusions or
perceptual distortions that impair the experience of reality
(APA 2000). Additionally, defendants with psychotic spec-
trum disorders may be perceived by jurors as less responsible
for their crimes because their symptoms are often bizarre,
severe, and visibly distressing—clear signs of a medical
rather than a perceived moral condition. Along these lines,
research by Corrigan et al. (2003) suggests that people are
more likely to adopt a sympathetic orientation toward those
with mental disorders when they are perceived as having
little control over their illness.
More remarkably, since the court-appointed evaluator’s
opinion is accepted by the court in the vast majority of cases,
African Americans’ disproportionate diagnosis with psy-
chotic disorders probably confers medical resources and
other benefits in the context of this small sector of the
criminal justice system. Specifically, it leads to placement in
an environment that is more conducive to positive mental
health and legal outcomes and is less dangerous, disorga-
nized, and isolating (Baillargeon et al. 2010; Birmingham
2003; Ditton 1999; Forrester et al. 2010; NAMI 2004). This
is paradoxical to the impact of psychotic spectrum diagnoses
in the community, which have been associated with greater
stigma, social isolation, and reduced life chances (Link et al.
1999; Pescosolido et al. 2007; Phelan et al. 2000; Van Dorn
et al. 2005; Watson et al. 2005). However, diagnosis with a
psychotic disorder may have long-term consequences for
inmates following community reintegration, particularly if
this diagnosis is inappropriate and leads to ineffective
treatment and poor outcomes.
Differential diagnostic patterns observed in these data are
likely attributable to a variety of mechanisms. As previous
research has suggested, clinicians may be unintentionally
biased in their application of diagnostic criteria, or the criteria
themselves might be biased (Lewis et al. 1990; Trierweiler
et al. 2000; van Ryn and Burke 2000; Widiger and Spitzer
1991). These processes may be exacerbated in the context of
forensic psychiatric evaluations. First, clinicians’ biased per-
ceptions in this type of clinical interaction may be especially
negative and strong since stereotypes and images of African
Americans as criminal perpetrators are pervasive in American
culture (Kennedy 1997; Russell 1998). In other words, racial
biases are likely to be particularly salient and influential since
the individuals being evaluated appear to validate racial ste-
reotypes of criminality. In addition, clinician–patient inter-
actions and communication may be strained and ineffective to
an even greater degree than is typical of race-discordant
clinical encounters (Cooper et al. 2003; Johnson et al. 2004;
van Ryn 2002; Vasquez 2007), increasing the likelihood of
misdiagnosis. That is, in forensic psychiatric evaluations, the
power differential between a white doctor and minority
patient is exacerbated by the deviant label and probably also
by vast social class inequalities.
Racial disparities in diagnosis found in these data may
also reflect real differences in rates of disorder among whites
and African Americans in the criminal justice system.
African Americans are less apt to voluntarily seek or receive
psychiatric treatment than whites (Department of Health and
Human Services 1999; Snowden and Pingitore 2002; Wang
et al. 2005), often resulting in delayed treatment or no
treatment for mental illness (Snowden 2001). When symp-
toms of untreated mental illness eventually reach crisis
levels, it increases the likelihood of criminal activity and of
being brought to psychiatric treatment by legal means,
emergency room use, and involuntary hospitalization
(Akutsu et al. 1996; Takeuchi and Cheung 1998; Snowden
and Cheung 1990; Snowden 1999). Thus, for African
Americans with psychotic spectrum disorders, the forensic
psychiatric evaluation may constitute their first real contact
with the mental health treatment system. Conversely, whites
with symptoms of psychosis may be more likely to seek
treatment earlier and voluntarily, reducing the likelihood that
they will end up in the criminal justice system (Markowitz
2006; Mechanic and Rochefort 1990).
Limitations
Because data are from a chart review of pre-trial evaluees
referred for psychiatric evaluation, there are many points
where selection bias could be introduced. For example,
Race Soc Probl
123
because the detail and accuracy of information in medical
charts vary depending on both the individuals recording
and reviewing the chart, there may have been errors in the
data. However, there is no reason to believe that any sys-
tematic errors that might have biased results occurred.
Also, a relatively small convenience sample of pre-trial
evaluees in one state pre-trial forensic hospital was
employed, and there is substantial variation in forensic
psychiatric procedures and conditions across the American
criminal justice system. There is also a selection bias in
which patients sent to the forensic hospital are selected by
legal counsel, judges, and medical personnel in the jail
without systematic criteria. Consequently, these findings
may not be generalizable to other states or systems. We
hope that this research serves as a starting point for larger,
nationally representative studies on racial and ethnic dis-
parities in diagnosis and court determinations in correc-
tional psychiatric contexts.
Implications for Practice and Policy
From a clinical standpoint, it is critical to develop cultur-
ally sensitive assessment and treatment models for use in
correctional psychiatric facilities, as well as in the broader
community. There are several factors working against the
therapeutic alliance when the patient and clinician are from
differing racial and ethnic backgrounds (Vasquez 2007).
For example, due to a history of abuse by medical
researchers and practitioners, African Americans may be
skeptical of white clinicians and suspicious of the mental
health treatment system in general (King 1992). At times,
these attitudes may be mislabeled as symptoms of paranoia
during the diagnostic process, leading to higher rates of
diagnosis with psychotic spectrum disorders among Afri-
can Americans (Neighbors et al. 1999). One key compo-
nent to reducing such misunderstandings and promoting
recovery is development of a strong therapeutic alliance
(Horvath and Luborsky 1993; Martin et al. 2000). Pro-
moting mutual trust and understanding takes time. When
clinicians put more effort into engaging the patient, it
appears to reduce racial discrepancies in psychiatric treat-
ment (Segal et al. 1996; Davis et al. 2011). It is also
important to increase the diversity of the mental health
workforce so that racial and ethnic minorities can be
matched with clinicians who are more culturally sensitive
and have shared ethnic experiences.
With respect to public policy, this research provides
additional evidence that it is critical to reduce racial and
ethnic disparities in mental health services. Research sug-
gests that African Americans tend to activate alternative
coping skills and sources of support rather than seeking
formal psychiatric services, often resulting in delayed
treatment or no treatment for mental illness (Snowden
2001). Resistance to help-seeking may be part of the
African American cultural legacy of bearing up to prob-
lems and remaining strong in the face of adversity—a
coping mechanism that evolved as a result of slavery
(Poussaint and Alexander 2000). Also, because perceptions
of stigma associated with mental illness are higher among
African Americans than other racial and ethnic groups
(Anglin et al. 2006), fear of public exposure may prevent
them from seeking services for psychiatric symptoms.
Policies and programs are needed that target minority
communities to reduce the stigma associated with mental
health services utilization and to make treatment-seeking
more normative and accessible.
Acknowledgments The authors extend special thanks to Ms. Jen-
nifer Haynes for support and feedback related to this project, and to
Tyler Jones, MD, who helped initiate the chart review and data col-
lection. The first and second authors contributed equally in this
research. Address correspondence to Brea Perry, Department of
Sociology, University of Kentucky, 1515 Patterson Office Tower,
Lexington, KY 40506 (email: [email protected]).
References
Abramson, M. F. (1972). The criminalization of mentally disordered
behavior: Possible side effect of a new mental health law.
Hospital and Community Psychiatry, 23, 101–105.
Adebimpe, V. R. (1981). Overview: White norms and psychiatric
diagnoses of black patients. American Journal of Psychiatry,138, 279–285.
Akutsu, P. D., Snowden, L. R., & Organista, K. C. (1996). Referral
patterns to ethnic specific and mainstream programs for ethnic
minorities and whites. Journal of Counseling Psychology, 43,
56–64.
American Psychiatric Association. (2000). Diagnostic and StatisticalManual of Mental Disorders (Revised (4th ed.). Washington,
DC: American Psychiatric Association.
Anglin, D. M., Link, B. G., & Phelan, J. C. (2006). Racial differences
in stigmatizing attitudes toward people with mental illness.
Psychiatric Services, 57(6), 857–862.
Baillargeon, J., Binswanger, I. A., Penn, J. V., Williams, B. A., &
Murray, O. J. (2009). Psychiatric disorders and repeat incarcer-
ations: The revolving prison door. American Journal of Psychi-atry, 166, 103–109.
Baillargeon, J., Hoge, S. K., & Penn, J. V. (2010). Addressing the
challenge of community reentry among released inmates with
serious mental illness. American Journal of Community Psy-chology, 46, 361–375.
Baron, R. M., & Kenny, D. A. (1986). Moderator-mediator variables
distinction in social psychological research: Conceptual, strate-
gic, and statistical considerations. Journal of Personality andSocial Psychology, 51, 1173–1182.
Birmingham, L. (2003). The mental health of prisoners. Advances inPsychiatric Treatment, 9, 191–199.
Blow, F. C., Zeber, J. E., McCarthy, J. F., Valenstein, M., Gillon, L.,
& Bingham, C. R. (2004). Ethnicity and diagnostic patterns in
veterans with psychoses. Social Psychiatry and PsychiatricEpidemiology, 39, 841–851.
Bresnahan, M., Begg, M. D., Brown, A., Schaefer, C., Sohler, N.,
Insel, B., et al. (2007). Race and risk of schizophrenia in a US
Race Soc Probl
123
birth cohort: Another example of health disparity? InternationalJournal of Epidemiology, 36, 751–758.
Chinn, S. (2000). A simple method an odds ratio to effect size for use
in meta-analysis. Statistics in Medicine, 19, 3127–3131.
Cooper, L. A., Roter, D. L., Johnson, R. L., Ford, D. E., Steinwachs,
D. M., & Powe, N. R. (2003). Patient-centered communication,
ratings of care, and concordance of patient and physician race.
Annals of Internal Medicine, 139, 907–915.
Corrigan, P. W., Markowitz, F. E., Watson, A. C., Rowan, D., &
Kubiak, M. A. (2003). An attribution model of public discrim-
ination towards persons with mental illness. Journal of Healthand Social Behavior, 44(2), 162–179.
Davis, T. D., Deen, T., Bryant-Bedell, K., Tate, V., & Fortney, J.
(2011). Does minority racial-ethnic status moderate outcomes of
collaborative care for depression? Psychiatric Services, 62(11),
1282–1288.
DelBello, M. P., Lopez-Larson, M. P., Soutullo, C. A., & Strakowski,
S. M. (2001). Effects of race on psychiatric diagnosis of
hospitalized adolescents: A retrospective chart review. Journalof Child and Adolescent Psychopharmacology, 11, 95–103.
Ditton, P. M. (1999). Mental health and treatment of inmates and
probationers. Bureau of Justice Statistics Special Report. Avail-
able at http://bjs.ojp.usdoj.gov/index.cfm?ty=pbdetail&iid=787.
Accessed 29 Jan 2012.
Dixon, L., Green-Paden, L., Delahanty, J., Lucksted, A., Postrado, L.,
& Hall, J. (2001). Variables associated with disparities in
treatment of patients with schizophrenia and comorbid mood and
anxiety disorders. Psychiatric Services, 52, 1216–1222.
Dunning, D., & Sherman, D. A. (1997). Stereotypes and tacit
inference. Journal of Personality and Social Psychology, 73,
459–471.
Ehrenkranz, S. M. (2001). Emerging issues with mentally ill
offenders: Causes and social consequences. Administration andPolicy In Mental Health, 28, 165–180.
Fein, S., Morgan, S. J., Norton, M. I., & Sommers, S. R. (1997). Hype
and suspicion: The effects of pretrial publicity, race, and
suspicion on jurors’ verdicts. Journal of Social Issues, 53,
487–502.
Forrester, A., Chiu, K., Dove, S., & Parrott, J. (2010). Prison health-
care wings: Psychiatry’s forgotten frontier. Criminal Behaviourand Mental Health, 20, 51–61.
Funtowicz, M. N., & Widiger, T. A. (1995). Sex bias in the diagnosis
of personality disorders: A different approach. Journal ofPsychopathology and Behavioral Assessment, 17, 145–165.
Higginbotham, A. L, Jr. (2002). Unequal justice in the state criminal
justice system. In S. L. Gabbidon, H. T. Greene, & V. D. Young
(Eds.), African American classics in criminology and criminaljustice (pp. 135–158). Thousand Oaks, CA: Sage.
Horvath, A. O., & Luborsky, L. (1993). The role of the therapeutic
alliance in psychotherapy. Journal of Consulting and ClinicalPsychology, 61(4), 561–573.
Johnson, R. L., Roter, D., Powe, N. R., & Cooper, L. A. (2004).
Patient race/ethnicity and quality of patient–physician commu-
nication during medical visits. American Journal of PublicHealth, 94, 2084–2090.
Kales, H. C., Blow, F. C., Bingham, C. R., Copeland, L. A., &
Mellow, A. M. (2000). Race and inpatient psychiatric diagnoses
among elderly veterans. Psychiatric Services, 51, 795–800.
Kennedy, R. (1997). Race, Crime and the Law. New York: Vintage.
King, P. A. (1992). The dangers of difference. The Hastings CenterReport, 22(6), 35–38.
Kirk, D. S. (2008). The neighborhood context of racial and ethnic
disparities in arrest. Demography, 45(1), 55–77.
Knoll, J. L., & Resnick, P. J. (2008). Insanity defense evaluations:
Toward a model for evidence-based practice. Brief Treatmentand Crisis Intervention, 8, 92–110.
Kunda, Z. (1999). Social cognition: Making sense of people.
Cambridge, MA: MIT Press.
Lamb, H. R., & Bachrach, L. L. (2001). Some perspectives on
deinstitutionalization. Psychiatric Services, 52, 1039–1045.
Lamb, H. R., & Weinberger, L. E. (2001). Persons with severe mental
illness in jails and prisons: A review. New Directions for MentalHealth Services, 90, 29–49.
Lawson, W. B., Hepler, N., Holladay, J., & Cuffel, B. (1994). Race as
a factor in inpatient and outpatient admissions and diagnosis.
Hospital and Community Psychiatry, 45, 72–74.
Lepore, L., & Brown, R. (1997). Category and stereotype activation:
Is prejudice inevitable? Journal of Personality and SocialPsychology, 72, 275–287.
Lewis, G., Croft-Jeffreys, C., & David, A. (1990). Are British
psychiatrists racist? British Journal of Psychiatry, 157, 410–415.
Link, B., Phelan, J. C., Bresnahan, M., Stueve, A., & Pescosolido, B.
A. (1999). Public conceptions of mental illness: Labels, causes,
dangerousness, and social distance. American Journal of PublicHealth, 89, 1328–1333.
Loring, M., & Powell, B. (1988). Gender, race, and DSM-III: A study
of the objectivity of psychiatric diagnostic behavior. Journal ofHealth and Social Behavior, 29(1), 1–22.
Lovell, D., Allen, D., Johnson, C., & Jemelka, R. (2001). Evaluating
the effectiveness of residential treatment for prisoners with
mental illness. Criminal Justice and Behavior, 28, 83–104.
MacKinnon, D. P., & Dwyer, J. H. (1993). Estimating mediated
effects in prevention studies. Evaluation Review, 17, 144–158.
MacKinnon, D. P., Warsi, G., & Dwyer, J. H. (1995). A simulation
study of mediated effect measures. Multivariate BehavioralResearch, 30, 41–62.
Markowitz, F. E. (2006). Psychiatric hospital capacity, homelessness,
and crime and arrest rates. Criminology, 44, 45–72.
Martin, D. J., Garske, J. P., & Davis, M. K. (2000). Relation of the
therapeutic alliance with outcome and other variables: A meta-
analytic review. Journal of Consulting and Clinical Psychology,68(3), 438–450.
Maurer, M., & King, R. S. (2007). Uneven justice: State rates of
incarceration by race and ethnicity. Report from the sentencingproject. Available online at http://www.sentencingproject.org/doc/
publications/rd_stateratesofincbyraceandethnicity.pdf. Accessed
15 Dec 2011.
Mechanic, D., & Rochefort, D. A. (1990). Deinstitutionalization: An
appraisal of reform. Annual Review of Sociology, 16, 301–327.
National Alliance on Mental Illness. (2004). Spending Money in Allthe Wrong Places: Jails and Prisons. Available online at
http://www.nami.org/Template.cfm?Section=Fact_Sheets&
Template=/ContentManagement/ContentDisplay.cfm&Content
ID=14593. Accessed 27 Dec 2011).
Neighbors, H. W., Trierweiler, S. J., Munday, C., Thompson, E. E.,
Jackson, J. S., Binion, V. J., et al. (1999). Psychiatric diagnosis
of African Americans: Diagnostic divergence in clinician-
structured and semistructured interviewing conditions. Journalof the National Medical Association, 91, 601–612.
Pescosolido, B. A., Fettes, D. L., Martin, J. K., Monahan, J., &
McLeod, J. D. (2007). Perceived dangerousness of children with
mental health problems and support for coerced treatment.
Psychiatric Services, 58, 619–625.
Pettit, B., & Western, B. (2004). Mass imprisonment and the life
course: Race and class inequality in US incarceration. AmericanSociological Review, 69, 151–169.
Phelan, J. C., Link, B. G., Stueve, A., & Pescosolido, B. A. (2000).
Public conceptions of mental illness in 1950 and 1996: What is
mental illness and is it to be feared? Journal of Health and SocialBehavior, 41, 188–207.
Poulson, R. L. (1990). Mock juror attribution of criminal responsi-
bility: Effects of race and the guilty but mentally ill (GBMI)
Race Soc Probl
123
verdict option. Journal of Applied Social Psychology, 20(19),
1596–1611.
Poussaint, A. F., & Alexander, A. (2000). Lay my burden down:Suicide and the mental health crisis among African Americans.
Boston, MA: Beacon Press.
Raybur, T. M., & Stonecypher, J. F. (1996). Diagnostic differences
related to age and race of involuntarily committed psychiatric
patients. Psychology Report, 79, 881–882.
Roberts, D. E. (2004). The social and moral cost of mass incarceration
in African American communities. Stanford Law Review, 56(5),
1271–1305.
Russell, K. K. (1998). The color of crime. New York: New York
University Press.
Sagar, H. A., & Schofield, J. W. (1980). Racial and behavioral cues in
black and white children’s perceptions of ambiguously aggres-
sive acts. Journal of Personality and Social Psychology, 39,
590–598.
Segal, S. P., Bola, J. R., & Watson, M. A. (1996). Race, quality of
care, and prescribing practices in the psychiatric emergency
room. Psychiatric Services, 47(3), 282–286.
Silver, E., Cirincione, C., & Steadman, H. J. (1994). Demythologizing
inaccurate perceptions of the insanity defense. Law and HumanBehavior, 18, 63–70.
Smedley, B. D., Stith, A. Y., & Nelson, A. R. (2002). Institute ofmedicine (US). Unequal treatment confronting racial and ethnicdisparities in healthcare. Washington, DC: National Academies
Press.
Snowden, L. R. (1999). Psychiatric inpatient care and ethnic minority
populations. In J. M. Herrera, W. B. Lawson, & J. J. Sramek
(Eds.), Cross cultural psychiatry (pp. 261–273). New York: John
Wiley and Sons Ltd.
Snowden, L. R. (2001). Barriers to effective mental health services
for African Americans. Mental Health Services Research, 3(4),
181–187.
Snowden, L. R. (2003). Bias in mental health assessment and
intervention: Theory and evidence. American Journal of PublicHealth, 93(2), 239–243.
Snowden, L. R., & Cheung, F. K. (1990). Use of inpatient mental
health services by members of ethnic minority groups. AmericanPsychologist, 453(3), 347–355.
Snowden, L. R., & Pingitore, D. (2002). Frequency and scope of
mental health service delivery to African Americans in primary
care. Mental Health Services Research, 4(3), 123–130.
Sobel, M. E. (1986). Some new results on indirect effects and their
standard errors in covariance structure models. SociologicalMethodology, 16, 159–186.
Stangor, C. (2000). Stereotypes and prejudice. Philadelphia, PA:
Psychology Press.
Stephan, J. J. (2004). State prison expenditures, 2001. Bureau ofJustice Statistics Special Report. Available online at http://bjs.
ojp.usdoj.gov/index.cfm?ty=pbdetail&iid=1174.
Strakowski, S. M., Keck, P. E, Jr, Arnold, L. M., Collins, J., Wilson,
R. M., Fleck, D. E., et al. (2003). Ethnicity and diagnosis in
patients with affective disorders. Journal of Clinical Psychiatry,64(7), 747–754.
Sweeney, Laura T., & Haney, Craig. (1992). The influence of race on
sentencing: A meta-analytic review of experimental studies.
Behavioral Sciences and the Law, 10, 179–195.
Takei, N., Persaud, R., Woodruff, P., & Murray, R. M. (1998). First
episodes of psychosis in Afro-Caribbean and white people. An
18-year follow-up population-based study. British Journal ofPsychiatry, 172, 147–153.
Takeuchi, D. T., & Cheung, M. K. (1998). Coercive and voluntary
referrals: How ethnic minority adults get into mental healthtreatment. Ethnicity and Health, 3, 149–158.
Trierweiler, S. J., Neiighbors, H. W., Munday, C., Thompson, S.,
Binion, V. J., & Gomez, J. P. (2000). Clinician attribution
associated with diagnosis of schizophrenia in African American
and non-African American patients. Journal of Consulting andClinical Psychology, 68, 171–175.
U.S. Department of Health & Human Services. (1999). Mental health:culture, race, ethnicity—supplement, a report of the surgeongeneral. http://mentalhealth.samhsa.gov/cre/default.asp.
Van Dorn, R. A., Swanson, J. W., Elbogen, E. B., & Swartz, M. S.
(2005). A comparison of stigmatizing attitudes toward persons
with schizophrenia in four stakeholder groups: Perceived
likelihood of violence and desire for social distance. Psychiatry:Interpersonal and Biological Processes, 68, 152–163.
van Ryn, M. (2002). Research on the provider contribution to race/
ethnicity disparities in medical care. Medical Care, 40(1),
140–151.
van Ryn, M., & Burke, J. (2000). The effect of patient race and socio-
economic status on physicians’ perceptions of patients. SocialScience and Medicine, 50, 813–828.
van Ryn, M., & Fu, S. S. (2003). Paved with good intentions: Do
public health and human service providerscontribute to racial/
ethnic disparities in health? American Journal of Public Health,93, 248–255.
Vasquez, M. J. T. (2007). Cultural differences and the therapeutic
alliance: An evidence-based analysis. American Psychologist,62(8), 878–885.
Wang, P. S., Berglund, P., Olfson, M., Pincus, H. A., Wells, K. B., &
Kessler, R. C. (2005). Failure and delay in initial treatment
contact after first onset of mental disorders in the national
comorbidity survey replication. Archives of General Psychiatry,62(6), 603–613.
Watson, A. C., Corrigan, P. W., & Angell, B. (2005). What motivates
support for legally mandated mental health treatment? SocialWork Research, 29, 87–94.
Widiger, T. A., & Spitzer, R. L. (1991). Sex bias in the diagnosis of
personality disorders: Conceptual and methodological issues.
Clinical Psychology Review, 11, 1–22.
Williams, D. R. (1999). Race, SES, and health: The added effects of
racism and discrimination. Annals of New York Academy ofSciences, 896, 173–188.
Williams, D. R., & Jackson, P. B. (2005). Social sources of racial
disparities in health. Health Affairs, 24, 325–334.
Williams, D. R., & Rucker, T. D. (2000). Understanding and
addressing racial disparities in health care. Health CareFinancing Review, 21(4), 75–90.
Wortzel, H., & Metzner, J. (2006). Clark v. Arizona: Diminishing the
right of mentally ill individuals to a full and fair defense. Journalof the American Academy of Psychiatry and the Law Online, 34,
545–548.
Wright, E. R., & Perry, B. L. (2010). Medical sociology and health
services research: Past accomplishments and future policy
challenges. Journal of Health and Social Behavior, 51(S),107–119.
Race Soc Probl
123