Page 1
Racial and ethnic differences in psychopathic personality
Richard Lynn*
University of Ulster, Coleraine, Northern Ireland, BT52 1SA, UK
Received 19 June 2000; received in revised form 15 December 2000; accepted 27 January 2001
Abstract
This paper proposes that there are racial and ethnic differences in psychopathic personality con-
ceptualised as a continuously distributed trait, such that high values of the trait are present in blacks and
Native Americans, intermediate values in Hispanics, lower values in whites and the lowest values in East
Asians. Part one of the paper sets out the evidence for this thesis. Part two applies the thesis to the unre-
solved problem in The Bell Curve that racial and ethnic differences in a number of social phenomena such
as crime, welfare dependency, rates of marriage, etc. cannot be fully explained by differences in intelligence
and proposes that some of the residual disparities are attributable to differences in psychopathic person-
ality. Part three of the paper integrates the theory with Rushton’s r-K theory of race differences. # 2001
Elsevier Science Ltd. All rights reserved.
Keywords: Psychopathic personality; Conduct disorder; Attention deficit hyperactivity disorder; Crime; Marriage;
Sexual behaviour; Race differences; r-K theory
1. Introduction
This paper offers a contribution to a problem posed by Herrnstein and Murray (1994) in The
Bell Curve. Herrnstein and Murray argue that racial and ethnic differences in a number of social
phenomena such as crime, poverty, long-term unemployment, teenage pregnancy and the like are
partly explicable in terms of differences in intelligence. They show, however, that differences in
intelligence cannot explain entirely the racial and ethnic differences in these phenomena and they
therefore conclude that some other factor or factors must also be involved. They conclude that
‘‘Some ethnic differences are not washed away by controlling for either intelligence or for any
other variables that we examined. We leave those remaining differences unexplained and look
forward to learning from our colleagues where the explanations lie’’ (p. 340).
0191-8869/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0191-8869(01)00029-0
Personality and Individual Differences 32 (2002) 273–316
www.elsevier.com/locate/paid
* Present address: The Whitfield Institute, Whitfield Court, Glewstone, Ross-on-Wye, Herts, HR9 6AS, UK. Tel.:
+44-1989-770908.
E-mail address: rlynn@uk.packardbell.org
Page 2
In this paper it is proposed that a component in the solution to this problem lies in racial and
ethnic differences in psychopathic personality considered as a continuously distributed person-
ality trait. It is argued that of the major racial and ethnic groups, East Asians score lowest on
psychopathic personality, whites score next lowest followed by Hispanics, while blacks and
Native Americans score highest. Of these five populations, East Asians, whites, blacks and Native
Americans can be considered as racial groups while Hispanics are an ethnic group from Latin
America and the Caribbean with a common Spanish heritage. The first part of the paper sets out
the evidence for this thesis. The second part applies the thesis to the solution of the problem
raised by Herrnstein and Murray. The third part of the paper discusses the relation between the
present thesis and Rushton’s r-K theory of race differences.
2. Psychopathic personality
We begin by describing the nature of psychopathic personality. The condition was identified in
the early nineteenth century by the British physician John Pritchard (1837) who proposed the
term ‘‘moral imbecility’’ for those deficient in moral sense but whose intellectual ability was
unimpaired. The term psychopathic personality was first used in 1915 by the German psychiatrist
Emile Kraepelin (1915) and has been employed as a diagnostic label throughout the twentieth
century. In 1941 the condition was described by Cleckley (1941) in what has become a classical
book The Mask of Sanity. He described the criteria for the condition as being general poverty of
affect, defective insight, absence of nervousness, lack of remorse or shame, superficial charm,
pathological lying, egocentricity, inability to love, failure to establish close or intimate relation-
ships, irresponsibility, impulsive antisocial acts, failure to learn from experience, reckless beha-
viour under the influence of alcohol, and a lack of long term goals.
In 1984 the American Psychiatric Association dropped the term psychopathic personality and
replaced it with ‘‘anti-social personality disorder’’. Some authorities such as Lykken (1995)
regard this as simply a synonym for psychopathic personality. Others, such as Hare (1994), con-
sider that there is some difference between the two concepts and that anti-social personality dis-
order is a less satisfactory term because it fails to give sufficient emphasis to the psychological
features as opposed to the behavioral characteristics of the condition. Despite these fine distinc-
tions, for practical purposes psychopathic personality and anti-social personality disorder can be
regarded as largely synonymous descriptions of the same condition.
In 1994 the American Psychiatric Association (1994) issued a revised Diagnostic Manual in
which it listed 11 features of anti-social personality disorder. These are: (1) inability to sustain
consistent work behavior; (2) failure to conform to social norms with respect to lawful behavior;
(3) irritability and aggressivity, as indicated by frequent physical fights and assaults; (4) repeated
failure to honor financial obligations; (5) failure to plan ahead or impulsivity; (6) no regard for
truth, as indicated by repeated lying, use of aliases, or ‘‘conning’’ others; (7) recklessness regard-
ing one’s own or others’ personal safety, as indicated by driving while intoxicated or recurrent
speeding; (8) inability to function as a responsible parent; (9) failure to sustain a monogamous
relationship for more than one year; (10) lacking remorse; and (11) the presence of conduct dis-
order in childhood. This is a useful list of the principal constituents of the condition, subject to
the reservation that it does not explicitly include the deficiency of moral sense although this is
274R. Lynn/Personality and Individual Differences 32 (2002) 273–316
Page 3
implicit in virtually all the listed behaviors. Despite the replacement of the concept of psycho-
pathic personality with that of anti-social personality disorder by the American Psychiatric
Association, many psychiatrists and psychologists such as Hare (1994) and Lykken (1995) regard
the term psychopathic personality as preferable and continue to use it.
In recent years the most frequently used instrument for the assessment of psychopathic per-
sonality is Hare’s Psychopathy Checklist (Hare, 1991, 1994; Harpur, Hart, & Hare, 1994). This
consists of a list of 20 symptoms of psychopathic personality which are identified as present or
absent by raters or by self-assessment. These symptoms fall into two main clusters or factors
identified by factor analysis. The first of these is a syndrome of eight ‘‘emotional-interpersonal’’
characteristics (glibness/superficial charm; grandiose sense of self-worth; pathological lying; con-
ning/manipulative; lack of remorse or guilt; shallow affect; callous, lack of empathy; failure to
accept responsibility for actions). The second factor consists of a syndrome of nine ‘‘social
deviance’’ characteristics (need for stimulation/proneness to boredom; parasitic lifestyle; poor
behavior controls; early behavior problems; lack of realistic long-term goals; impulsivity;
irresponsibility; juvenile delinquency; revocation of conditional release). The Psychopathic
Checklist contains three further characteristics which load about equally on both factors.
These are promiscuous sexual behavior; many short-term sexual relationships; and criminal
versatility. The twenty items of the Psychopathy Checkist are scored 0 (No), 1 (Maybe) or 2
(Yes). This treats psychopathic personality as a continuously distributed personality trait
with scores ranging between 0 and 40. Research using this instrument has shown a number
of informative features of psychopathic personality such as the higher scores normally
obtained by males but it has not been used to any significant scale to investigate racial and
ethnic differences.
2.1. The MMPI Psychopathic Deviate Scale
The MMPI (Minnesota Multiphasic Personality Inventory) is a questionnaire consisting of a
series of scales for the measurement of a variety of psychiatric conditions regarded as con-
tinuously distributed in the population, such as hysteria, mania and depression. The MMPI was
constructed in the late 1930s by Hathaway and McKinley (1940) and is one of the most widely
used instruments in personality research. It contains a scale for the measurement of psychopathic
personality known as the Psychopathic Deviate Scale. This was constructed by writing a number
of questions, giving them to criterion groups of those manifesting psychopathic behaviour and
‘‘normals’’, and selecting for the scale the questions best differentiating the two groups. The cri-
terion group manifesting psychopathic behaviour consisted of 17–24 year olds appearing before
the courts and referred for psychiatric examination because of their ‘‘long histories of delinquent-
type behaviours such as stealing, lying, alcohol abuse, promiscuity, forgery and truancy’’ (Archer,
1997, p. 20). The common feature of this group has been described as their failure to ‘‘learn those
anticipatory anxieties which operate to deter most people from committing anti-social beha-
viour’’ (Marks, Seeman, & Haller, 1974, p. 25). The manual describes those scoring high on the
scale as follows: irresponsible, antisocial, aggressive, having recurrent marital and work pro-
blems, and underachieving (Hathaway & McKinley, 1989). A number of subsequent studies have
shown that the Psychopathic Deviate scale differentiates delinquents and criminals from non-
delinquents and non-criminals (e.g. Elion & Megargee, 1975).
R. Lynn/Personality and Individual Differences 32 (2002) 273–316 275
Page 4
During the 60 or so years following its publication the MMPI has been administered to
numerous groups. Normative data for the means obtained in the United States by ‘‘normal’’ (i.e.
non-psychiatric) samples of blacks, whites, ethnic Japanese Americans, Hispanics and Native
Americans have been published by Dahlstrom, Lachar, and Dahlstrom (1986). I have calculated
these data as d scores (differences expressed in standard deviation units) in relation to a white
value of zero, positive signs indicating scores higher than those of whites and negative signs
indicating signs lower than those of whites. Scores of males and females have been combined
giving equal weight to both sexes. The results of these calculations are shown in the first row of
Table 1. The ethnic Japanese are entered in the table under ‘‘E. Asians’’ (East Asians) because
‘‘Asians’’ is the term most frequently used for this group in the American literature and I have
added the prefix ‘‘East’’ to differentiate them from South Asians from the Indian sub-Continent.
Most American Asians are ethnic Chinese, Japanese, Koreans and Filipinos, although they also
include Vietnamese and some others from southeast Asia. Hispanics are not a homogeneous
racial group but rather an ethnic group from Spanish speaking Latin America and the Caribbean
which includes ethnic Europeans, Native Americans, blacks and racial hybrids. The data set out
in the first row of Table 1 show that blacks and Native Americans obtain higher means than
whites on the scale (by 0.29d and 0.44d, respectively), Hispanics obtain the same mean as whites,
while Japanese Americans obtain a lower mean (?0.31d).
The MMPI was revised and standardised in the second half of the 1980s. The revised version of
the test is called the MMPI-2 and much useful information about the test is provided in the
manual (Hathaway & McKinley, 1989). The standardisation sample numbered 2500 and was
selected to match the national population of the United States in terms of geographical location,
age, educational level, socio-economic status, earnings, marital status and ethnicity. Of the 54
questions in the Psychopathic Deviate scale of the original test, four were replaced in the MMPI-
2. The manual provides means and standard deviations for five racial and ethnic groups in the
standardisation sample. These are whites, blacks, Asians, native Americans and Hispanics. No
details are given regarding the national or ethnic origins of the Asian group, but as noted the
majority of these in the United States are ethnic Chinese, Japanese, Koreans and Filipinos. I have
converted the means of the five groups to d scores in relation to a white zero and combined males
and females, in the same way as for the original MMPI. These results are shown in row 2 of
Table 1. The results are consistent with those of the original MMPI in so far as blacks and Native
Americans obtain higher mean scores than whites, while East Asians obtain lower scores. The
result for Hispanics is discrepant in so far as they obtain a substantially higher mean than whites
Table 1
Psychopathic deviate scale of the MMP1 (d)
No.Location Test BlacksE. Asians Hispanics N. Americans WhitesReference
1
2
3
4
5
USA
USA
Japan
Nigeria
USA
MMP1
MMP1-2
MMP1-2
MMP1-2
MMP1-A
0.29
0.48
?0.31
?0.18
?0.36
0.00
0.70
0.44
0.74
0.00
0.00
Dahlstrom et al., 1986
Hathaway & McKinley, 1989
Japanese MMP1, 1993
Nzewi, 1998
Archer, 1997
0.50
0.33 0.36 0.00
6 Mean 0.40
?0.280.350.59 0.00
276R. Lynn/Personality and Individual Differences 32 (2002) 273–316
Page 5
in the MMPI-2 standardization sample but the same mean as whites in the original MMPI. Row
3 gives data for the Japanese from the Japanese manual of the MMPI-2 based on a nationally
representative standardisation sample of 1022. It will be seen that the Japanese obtain a lower
mean score than American whites. Row 4 gives data from an administration of the MMPI-2 to
200 Nigerian male and female students in Nigeria and shows that they obtained an average score
half a standard deviation higher than white American students.
In 1992 the MMPI-A was published as a version of the test designed for adolescents. Norma-
tive data for blacks, whites and ‘‘others’’ have been provided by Archer (1997) and are shown in
row 5. In this row ‘‘others’’ are entered as Hispanics because these are by far the largest third
ethnic group. It will be seen that blacks and Hispanics score about a third of a standard deviation
higher than whites.
The racial and ethnic differences in psychopathic deviance scores presented in Table 1 show
reasonably consistent results for blacks and Native Americans, both of which score consistently
higher than whites in the range of 0.29–0.74d, and also for East Asians who consistently score
lower than whites in the range of ?0.18 to ?0.36d. The results for Hispanics are less consistent
but average out at 0.35d. To obtain overall estimates for the five groups the results in the table
have been averaged giving equal weight to each study. These averages are shown in the bottom
row of the table and indicate the highest level of psychopathic personality in Native Americans,
followed by blacks, Hispanics, whites and East Asians. It is proposed that these means should be
adopted provisionally as the best available estimates of the racial and ethnic differences in psy-
chopathic personality.
2.2. Conduct disorder
The concepts of psychopathic personality and anti-social personality disorder are not used for
children or young adolescents up to the age of 15 years. Children and young adolescents mani-
festing psychopathic and anti-social behaviours are instead regarded as having conduct disorders.
The principal criteria set out by the American Psychiatric Association (1994) for a diagnosis of
conduct disorder are persistent stealing, lying, truancy, running away from home, fighting, arson,
burglary, vandalism, sexual precocity and cruelty. Childhood conduct disorder is therefore an
analogue of psychopathic personality in older adolescents and adults and a diagnosis of psycho-
pathic personality is not normally made without evidence of conduct disorder in childhood. A
number of studies have shown that conduct disorder in children is a frequent precursor of psy-
chopathic behaviour in later adolescence and adulthood (e.g. Bernstein, Cohen, Skodal, Bezir-
ganian, & Brook, 1996; Loeber, 1990; Mealy, 1995). Conduct disorder is most generally measured
by rating scales of which the most widely used are the Child Behaviour Checklist (CBCL) con-
structed by Achenbach (1992) and the Teachers Rating Scale (TRS) constructed by Conners
(1989). These rating scales consist of a number of the expressions of conduct disorder and the
teacher, or sometimes the parent, identifies those manifested by the child being assessed. The
ratings are summed to give a score.
A number of studies reporting differences in mean scores obtained by children of different
racial and ethnic groups are shown in Table 2. Where the results are reported as means and
standard deviations, the racial and ethnic differences have been calculated as d scores, as in
Table 1. The first row presents data for blacks and whites from a study of 1027 children in North
R. Lynn/Personality and Individual Differences 32 (2002) 273–316 277
