Matthew Howard, Michael Dayton, Kirk Foster, Michael Vaughn, John Zelner. Handbook of Children, Culture, and Violence. Editor: Nancy E Dowd, Dorothy G Singer, Robin Fretwell Wilson. Sage Publication. 2008.
Psychopathy has been recognized as a serious psychiatric malady for more than two centuries. Formal psychopathy assessments, however, have only recently been included in mental health evaluations of juvenile offenders. A small subpopulation of youth offenders—perhaps 5 to 8 %— commits a majority of general and violent crimes. The notion that members of this group are psychiatrically disordered gained widespread currency with the publication of Dr. Terrie Moffitt’s (1993) classic paper distinguishing “life-course persistent” and “adolescence-limited” delinquent subtypes. Moffitt observed the following:
[L]ongitudinal research consistently points to a very small group of males who display high rates of antisocial behavior across time and in diverse situations. The professional nomenclature may change, but the forces remain the same as they drift through successive systems aimed at curbing their deviance: schools, juvenile-justice programs, psychiatric treatment centers, and prisons. The topography of their behavior may change with changing opportunities, but the underlying disposition persists throughout the life course, (p. 678) [italics added]
The disposition underlying persistent, life-course criminality was, Moffitt argued, rooted in early neuropsychological vulnerabilities and criminogenic environmental influences, which interact to produce the disorder.
As investigators have attempted to better account for the conspicuous heterogeneity among youth offenders (Loeber, Farrington, & Waschbusch, 1998; Tolan & Gorman-Smith, 1998; Wolfgang, Figlio, & Sellin, 1972; Wolfgang, Thornberry, & Figlio, 1987), other taxonomies of antisocial youth have emerged. Many of these taxonomies, including the distinction between psychopathic and nonpsychopathic juvenile offenders, describe two primary offender subgroups: first, a relatively small group of early-onset, criminally versatile, chronic offenders who frequently have histories of violent behavior and comorbid attention-deficit/hyperactivity disorder (ADHD) and conduct disorder (CD); and second, a substantially larger group who offend later, have lower rates of interpersonal violence and psychopathology, and whose offending terminates in adolescence (Quay, 1993).
Implicit in many juvenile offender taxonomies is the notion that neurological or other biological factors play a key role in the development of the more serious youth offender subtype. The origins of the substantially more common and comparatively benign juvenile offender subtype, on the other hand, are considered primarily social in nature. While it is evident that valid schemes for subtyping juvenile offenders might lead to increased understanding and better treatment of early antisocial behavior, it is probable that such typologies would have seriously deleterious consequences for youth diagnosed with the more pernicious subtype. Many social service and legal practitioners regard “juvenile psychopaths” as untreat-able, at least given currently available pharmacological, cognitive-behavioral, and social/criminological interventions (Young, Justice, Erdberg, & Gacano, 2000). There is some evidence (see Lyon & Ogloff, 2000) that adolescent offenders in the United States and Canada are increasingly subjected to assessments for psychopathy and that important determinations are made on the basis of these findings, including transfer decisions. Thus, it is critical that legal and social service practitioners be aware of key issues and recent findings vis-à-vis juvenile psychopathy research and practice.
Legal and social service practitioners in the adult criminal justice and juvenile justice systems likely will encounter the notion of psychopathy at some point in their professional careers. They should be aware of issues and research relevant to the construct. This chapter presents a comprehensive overview of the psychopathy research and clinical literature. The chapter reviews contemporary approaches to assessment, characteristics of youth diagnosed with psychopathy, potential applications of psychopathy evaluations, and possible effects of psychopathy diagnoses on legal proceedings.
Contemporary Perspectives on Psychopathy
One of the most significant events in psychopathy research was the publication of the Psychopathy Checklist (PCL) (1991), available currently in a revised format referred to as the PCL-R (Hare, 1996a, 1996b). The PCL-R operationalizes psychopathy in a manner generally consistent with Cleckley’s (1941) clinical description. The PCL-R consists of 20 items reflecting a variety of personality and behavioral attributes that are each scored from 0 (not present) to 2 (definitely present) by a trained rater (Bodholt, Richards, & Gacono, 2000). Total PCL-R scores range from 0 to 40; scores of 30 or more are generally used to identify adult psychopaths. Ideally, ratings are based on semi-structured interviews with the target individual and family members or friends who have significant knowledge of the individual and a review of available criminal justice/mental health file records. On occasion, psychopathy diagnoses are based only on file records.
Factor analytic studies of the PCL-R have consistently identified two independent, though moderately positively correlated (r= 0.5-0.6) factors (Bodholt et al, 2000). Factor 1—an interpersonal and affective dimension—incorporates items referring to dispositional glibness/superficial charm, a grandiose sense of self-worth, pathological lying, conning and manipulative behavior, lack of remorse or guilt, shallow affect, callousness/ lack of empathy, and a failure to accept responsibility for his or her actions (Hare, 1996a, 1996b). Factor 2 includes items reflecting an unstable and antisocial lifestyle such as those assessing need for stimulation/ proneness to boredom, parasitical lifestyle, poor behavioral controls, early-life problem behaviors, lack of realistic long-term goals, impulsivity, irresponsibility, a history of juvenile delinquency, and court-ordered revocation of conditional release. Factor 1 scores tend to correlate most highly with measures of narcissism and egocentrism, whereas Factor 2 scores correlate significantly positively with measures of substance abuse, criminal behavior, and antisocial personality disorder criteria (American Psychiatric Association, 2000). Although most research to date has employed the two-factor model of the PCL-R, several studies now support a three-factor model and a two-factor (4-facet subscale) model (Cooke & Michie, 2001; Falkenbach, Poythress, & Heide, 2003; Skeem, Mulvey, & Grisso, 2003).
Studies of the PCL-R indicate that the measure possesses good internal consistency, inter-rater, and test-retest reliability (Bodholdt et al., 2000). Validity assessments indicate that PCL-R scores predict general and violent recidivism (Salekin, Rogers, & Sewell, 1996), and institutional violence among forensic patients (Heilbrun, Hart, Hare, Gustafson, Nunez, et al., 1998). Lo date, hundreds of studies of adults have used the original PCL, PCL-R, or one of the other versions of the instrument such as the PCL:SV (Screening Version) (Sparrow & Gacono, 2000).
Only recently, however, have efforts been made to apply PCL-based and other psychopathy assessments to antisocial children, adolescent offenders, and members of the general adolescent population. Investigations of PCL-defined psychopathy in youth began with Forth, Hart, and Hare’s (1990) study of adolescent offenders. PCL-R Items 9 (parasitical lifestyle) and 17 (many short-term relationships) were deleted and the scoring criteria for Items 18 (juvenile delinquency) and 20 (revocation of conditional release) were modified to reflect adolescent offenders’ more limited opportunities for interaction with the justice system relative to adult offenders (Forth & Mailloux, 2000). Lhis modification of the PCL-R was used in a number of studies of adolescents (Edens, Skeem, Cruise, & Cauffman, 2001; Seagrave & Grisso, 2002). However, several recent investigations have used the newer 20-item Psychopathy Checklist: Youth Version (PCL: YV), which is a modified version of the original PCL-R that was explicitly designed for adolescents (Forth & Mailloux, 2000).
As with the PCL-R, factor analytic studies of the PCL:YV have identified two factors— an affective/interpersonal dimension (Factor 1) and an antisocial lifestyle/behavior factor (Factor 2)—that underlie juvenile psychopathy (Brandt, Kennedy, Patrick, & Curtin, 1997). Lhe alpha and inter-rater reliabilities of the PCL:YV appear to be acceptable for total scores, but more research evaluating the inter-rater reliability of the subscales is needed. A number of studies of adolescents have established the construct validity of the modified PCL-R and PCL:YV scales. However, no widely accepted cutpoints for the diagnosis of juvenile psychopathy have been established for PCL-based measures. Researchers have raised serious concerns regarding the developmental appropriateness of some PCL:YV items, such as those assessing parasitical lifestyle and many short-term relationships (Brandt et al., 1997). Edens étal. (2001) observed that “although the scoring criteria for several problematic items (e.g., impulsivity, irresponsibility, and need for stimulation/proneness to boredom) have been revised in an attempt to better tailor them to adolescent respondents, the stability of these items over significant time periods appears to be an open issue” (Edens et al., 2001, p. 61).
Several studies have recently examined the predictive validity of two- and three-factor versions of the PCL:YV. Corrado, Vincent, Hart, and Cohen (2004) found that factor scales from both models predicted general and violent recidivism in 182 male adolescent offenders followed for approximately 1 year. Measures of antisocial behavior rather than psychopathic personality traits accounted for most of the instrument’s explanatory power.
Other instruments designed to assess psychopathy in youth are discussed below (Fnck, 2002; Fnck, O’Brien, Wootton, & McBurnett, 1994; Lynam, 1997). Factor analyses of these scales also support a two-factor model of psychopathy similar to that assessed by the various PCL-based measures, although some support for three-factor models has also been adduced.
The relationship between the DSM conduct disorder diagnoses and juvenile psychopathy is asymmetrical. A majority of youth diagnosed as psychopathic meet the more behaviorally based conduct disorder criteria, but only a minority of youth who meet conduct disorder criteria also meet juvenile psychopathy criteria. Conduct disorder diagnoses are also far more prevalent than psychopathy diagnoses among adolescent offender populations across a variety of juvenile justice settings (Forth, 1995). According to the Diagnostic and Statistical Manual of Mental Disorders (DSM) (APA, 2000), youth meet conduct disorder criteria if they evidence “a repetitive and persistent pattern of behavior in which the basic rights of others or age-appropriate societal norms or rules are violated, as manifested by the presence of three (or more)… criteria in the past 12 months, with at least one criterion present in the last 6 months” (p. 98). A total of 15 criteria for conduct disorder are enumerated in the DSM (e.g., deliberately engaging in fire setting).
Despite all these PCL-based studies of youth, it remains unclear whether psychopathy is more appropriately regarded as a discrete disorder or a continuously distributed characteristic. One recent taxometric analysis supported the notion that persistently antisocial youth constitute a naturally occurring discrete class (i.e., a “taxon”) of youth (Skilling, Quinsey, & Craig, 2001). Most studies of youth to date have used psychopathy measures as a continuous as well as categorical variable, implicitly embracing the notion that psychopathic traits may exist on a continuum among youth.
Studies to date suggest that PCL-R-based measures may eventually hold some utility vis-à-vis prediction of offending in youth; however, the use of these instruments in clinical or forensic settings at present would be premature. The developmental appropriateness and factor structure of these measures have not been clearly established, whereas the potentially adverse consequences of false-positive diagnoses with juvenile psychopathy are obvious.
In addition to the PCL-derived instruments, a number of other measures have been developed to assess juvenile psychopathy including the Antisocial Process Screening Device, Child Psychopathy Scale, and P-Scan (Edens et al., 2001; Seagrave & Grisso, 2002). To date, none of the juvenile psychopathy measures has been validated to the point necessary to justify its routine use in clinical or forensic settings. Each of the respective measurement approaches—structured interview versus self-report—has its drawbacks. Structured interviews are time consuming and expensive to conduct and require access to and a careful review of available clinical/forensic file data, whereas the validity of self-report psychopathy measures has been called seriously into question. Substantially more research is needed that addresses the reliability, validity, and utility of various approaches to juvenile psychopathy assessment before such measures can be considered for clinical and forensic uses.
Characteristics of Psychopathic Youth
Prior investigations assessing a range of additional variables have also helped to identify characteristics of psychopathic youth. Juvenile psychopathy has been evaluated in relation to demographic features such as age, ethnicity, and gender; sociodevelop-mental experiences such as child abuse and neglect; psychiatric problems such as substance abuse and other co-occurring psychiatric disorders; and patterns of criminal offending including the age of onset, and the nature and frequency of antisocial conduct.
Gender, ethnicity, and age have not been demonstrated as significant factors in psychopathy. Relatively few studies have examined gender differences in adolescent psychopathy. Three studies found that females had lower psychopathy scores, on average, than males: Gretton (1998); Rowe (1997); and Stanford, Ebner, Patton, and Williams (1994). However, only in one investigation (Stanford et al., 1994) was the difference statistically significant, perhaps due to the comparatively small number of subjects participating in these studies (Forth & Mailloux, 2000).
Five studies comparing Caucasian adolescents to Native Canadian, African American, and Hispanic youth reported nonsignificant differences with regard to psychopathy (Brandt et al., 1997; Hume, Kennedy, Patrick, & Partyka, 1996; McBnde, 1998; Meyers, Burket, & Harris, 1995; Pan, 1998). Forth, Hart, and Hare (1990) found significantly lower psychopathy scores for native Canadian than Caucasian youth. Cross-sectional studies examining the relationship between age and psychopathy among adolescents have not generally identified significant associations (Forth & Burke, 1998).
Clinical lore strongly supports an association between substance abuse and psychopathy (Cleckley, 1941). However, research investigating the link between the two in adults and adolescents is limited. Rutherford, Alterman, and Cacciola (2000) reviewed the adult psychopathy literature and concluded that, in general, there is a moderate association between measures of substance abuse and dependence and the PCL-R total and Factor 2 (antisocial lifestyle) subscale scores, respectively. For example, Hart and Hare (1989) reported correlations of r = .31 and r= .40, respectively, for PCL-R total and Factor 2 scores and a measure of drug abuse/dependence symptoms among 80 forensic psychiatric adult patients. Smith-Stevens and Newman (1990) found significantly higher rates of lifetime alcohol abuse/dependence and drug abuse/ dependence among psychopathic Wisconsin adult inmates compared to their nonpsychopathic cohorts. PCL-R total and Factor 2 scores were significantly inversely associated with age at first intoxication and first arrest, whereas Factor 1 (core personality traits) scores were significantly inversely associated only with age at first arrest. Rutherford et al. (2000) concluded that, “Antisocial Lifestyle [i.e., Factor 2] consistently had a stronger relationship than Psychopathic Personality Traits [i.e., Factor 1] to alcohol and drug abuse/dependence among male offenders” (p. 354).
Studies of adolescent psychopathy and substance abuse report mixed findings. Mailloux, Forth, and Kroner (1997) examined the relationship between PCL:YV scores and measures of lifetime alcohol and drug problems. Total and Factor 2 (antisocial lifestyle) scores, respectively, were significantly related to Michigan Alcoholism Screening Test (MAST) (r’s = .46 and .41) and Drug Abuse Screening Test (DAST) (r’s = .42 and .48) scores, age at drug use initiation (r’s = −.50 and −.50), and number of drugs tried (r’s = .56 and .54). Factor 1 (core personality traits) scores were not significantly correlated with MAST and DAST scores and had lower, although statistically significant, associations with age of onset of drug use (r = −.39) and number of illicit drugs tried (r = .46). Brandt et al. (1997) studied 130 adolescent offenders with multiple felony convictions and did not find PCL-R total, Factor 1 (core personality traits), or Factor 2 (antisocial lifestyle) scores to be significantly associated with a measure of substance abuse based on file records, although the assessment of substance abuse was relatively crude.
Forth (1995) also examined associations between PCL:YV scores and the MAST and DAST in a community youth sample (mean age = 17.2, SD = 1.6) and a sample of young offenders (mean age = 17.5,SD= .90). Total (r = .48), Factor 1 (core personality traits) (r = .33), and Factor 2 (antisocial lifestyle) (r= .47) scores were significantly associated with MAST scores among community youth. Similar associations were observed between total (r= .56), Factor 1 (core personality traits) (r= .41), and Factor 2 (antisocial lifestyle) (r= .52) scores and the DAST measure of lifetime drug-related problems among community youth. Only total and Factor 2 scores, respectively, were significantly associated with MAST (r’s= .23 and .28) and DAST scores (r’s = .28 and .36) among serious youth offenders and the observed correlations were lower for the offender group than the community sample.
In sum, findings with adolescent offenders approximate those obtained with adult offenders with regard to substance abuse measures (Rutherford et al., 2000). Antisocial lifestyle traits are more consistently and strongly associated with substance abuse than are the core personality traits of psychopathy. The causal nature of the relationship between Factor 2 (antisocial lifestyle) traits and substance abuse is unclear. It is possible that substance abuse is just one of many manifestations of an impulsive and risk-taking lifestyle, or that it plays an important independent role in the development of various antisocial outcomes. That is, early substance abuse might reflect a predisposition to antisocial behavior and contribute to further and more aberrant conduct by disinhibiting behavior and impairing judgment.
The conflation of psychopathy with moral turpitude can be traced back two centuries to the notion of moral insanity. Early studies of moral reasoning compared psychopathic youth, variously defined, to other delinquents (Fodor, 1973) or to non-incarcerated, non-delinquent youth (Campagna & Harter, 1975), and found more rudimentary levels of moral reasoning in psychopathic youth. Jurkovic and Prentice (1977) and Lee and Prentice (1988) found lower levels of moral development in psychopathic youth compared to normal youth, but had mixed results with regard to moral reasoning. Lee and Prentice (1988) identified nonsignificant differences between psychopathic and other delinquent youth, whereas Jurkovic and Prentice (1977) found the moral reasoning of psychopathic youth significantly less developed than that of other delinquent, but nonpsychopathic, youth.
Trevethan and Walker (1989) compared the moral development and moral orientation of 14 psychopathic adolescents to 15 delinquent but nonpsychopathic youth and 15 normal youth recruited from a local high school. Psychopathic youth were nearly 1 year older, on average, than youth in the delinquent and normal groups. Participants were asked to respond to hypothetical and real-life moral dilemmas during a lengthy interview. Psychopathic youth differed significantly from normal youth, but not from other delinquents, with regard to their stage of moral reasoning. Psychopathic youth did, however, display a significantly more “egoistic utilitarian” moral orientation than did normal youth or other delinquent youth in response to real-life moral dilemmas. Trevethan and Walker concluded that “although there were no differences across groups when discussing hypothetical dilemmas, when it was a situation in which they had actually been involved, psychopaths more frequently expressed the moral legitimacy of concerns for themselves” (1989, p. 100).
Blair (1997) compared the moral judgments of 16 psychopathic and 16 nonpsychopathic residents of a school for behaviorally and emotionally disturbed youth. Blair presented each youth with a task story and assessed each participant’s response to a transgression and the protagonist’s emotions. Psychopathic youth were significantly more likely than nonpsychopathic youth to consider moral transgressions acceptable if there were no formal rules prohibiting the transgression. Only 25% of the justifications provided by psychopathic youth for their judgments concerned the welfare of others, compared to 45% of the justifications provided by nonpsychopathic youth (p < .07). Psychopathic youth were also less likely to attribute guilt to task story characters, suggesting that they were less sensitive to or aware of this potential emotional response.
Studies to date support the notion that psychopathic youth function at a less ethically developed level than nonpsychopathic youth do, although findings of differences between psychopathic and other delinquent youth are less consistent. In general, the methodological limitations and mixed findings of the scant available research comparing moral reasoning in psychopathic and nonpsychopathic youth indicate a need for further research. Saltaris (2002) reviewed research suggesting that the capacity to feel empathy for others and to discern others’ emotional states (i.e., “perspective taking”) develops very early in life, varies greatly across individuals, and is potentially a key determinant of psychopathic (and altruistic) orientation. Longitudinal studies of perspective taking and empathy commencing very early in the lives of high-risk youth would contribute significantly to current knowledge regarding the developmental origins of the callousness and narcissism observed in adolescent and adult psychopaths.
Child Abuse and Poor Parenting
Several investigators have examined the role of adverse early life experiences in the development of psychopathy. Forth and Tobin (1996) found that a history of child abuse was highly prevalent among 95 psychopathic and nonpsychopathic incarcerated male youth, although rates in the two groups did not differ significantly. Psychopathic and nonpsychopathic offenders also did not differ significantly with regard to their histories of specific forms of abuse including physical, emotional, or sexual abuse or neglect. Experiencing or witnessing parental violence did not significantly predict PCL-R scores.
McBride (1998) found that parental antisocial characteristics and a history of physical abuse, but not other forms of abuse, were associated with psychopathy among 239 adolescent male sex offenders ages 12 to 18. A related investigation of 74 adolescent male offenders identified a significant association between a history of physical abuse and poor parenting, and adolescent psychopathy. McBride and Hare (1996) reported that PCL-SV scores were significantly positively correlated with a history of physical (r= .33) and sexual (r= .16) abuse. Together, a composite measure of parental deviance, a history of physical abuse, and a diagnosis of ADHD explained 22% of the variance in psychopathy scores, suggesting that these factors may play a role, but are not decisive, in the development of juvenile psychopathy. Burke and Forth (1996) found that a global index of family background variables, including sexual abuse or parental alcoholism, was significantly related to Factor 2 (antisocial lifestyle) psychopathy scores among a sample of 106 young male offenders, but not with PCL:YV total or Factor 1 (personality/interpersonal features) scores. None of the 10 family-background variables significantly predicted total, Factor 1, or Factor 2 scores among the young offender sample. Laroche and Toupin (1996) studied 60 adolescent male offenders and found that psychopathic youth participated in fewer family activities and were more poorly supervised than nonpsychopathic offenders. Campbell, Porter, and Santor (2004) identified a significant association between a history of physical abuse and total PCL: YV scores in a sample of 226 male and female adolescent offenders. A history of non-parental living arrangements, such as living in foster care, predicted PCL:YV scores in multivariate analyses. Forth and Burke (1998) reviewed studies of developmental correlates of juvenile psychopathy, concluding that “relatively little research has examined whether psychopaths have dysfunctional family backgrounds. The research that has been done, though, has shown no decisive link between family history and the presence of psychopathy in adults” (p. 223). Gretton (1998) reported that psychopathic youth offenders were separated from their biological mothers and fathers at significantly younger ages, on average, than were nonpsychopathic adolescent offenders, although no differences were found in the prevalence of childhood abuse.
Inconsistent findings about the role of parenting practices in the development of juvenile psychopathy may be due to failure to distinguish between correlates of the two factors thought to comprise the disorder. Wootten, Frick, Shelton, and Silverthorn (1997) predicted that youth high in callous-unemotional (CU) traits would not be substantially influenced by different parenting practices with regard to conduct disorder, whereas youth low in callous-unemotional traits would experience differential outcomes related to parenting practices. Wootten et al. found that “the association between ineffective parenting and conduct problems was moderated by the presence of C/U traits in the child … children with high CU traits exhibited high rates of conduct problems regardless of the quality of parenting they experienced … past studies may have ignored the important association between parenting practices and conduct problems by failing to distinguish between youth low and high in CU traits” (1997, p. 305). In a study of 136 adolescent Hispanic females, Vitacco, Neumann, Ramos, and Roberts (2003) found significant relationships between poor parental monitoring and inconsistent discipline, and the narcissism and impulsivity subscale of the Antisocial Process Screening Device (APSD), but nonsignificant associations with the C/U scale. Future studies should examine the effects of parenting on youth at the extreme ends of the callous-unemotional trait distribution.
Violent and Institutional Offending
Current research supports a relatively robust association between psychopathy and violent offending for adult males (Bodholdt etal., 2000). Edens et al. (2001) reviewed 11 studies of adolescent offenders evaluating this relationship that used a variety of psychopathy measures, research designs, and violence outcomes. Overall, findings across studies were remarkably consistent, indicating that total psychopathy scores are moderately associated with violence, with most correlations ranging from .20 to .40. Brandt etal. (1997) and Forth et al. (1990) found that PCL-R total scores were significantly related to time-to-violent reoffending and number of charges/convictions for violent reoffenses, respectively, among incarcerated delinquents released into the community. Five studies assessed the relationship of the modified PCL-R to institutional misbehavior and infractions. Brandt etal. (1997) found moderate associations between PCL-R scores and verbal (r= .31) and physical (r= .28) misbehavior. Edens etal. (1999) found a significant correlation of .28 between PCL-R total scores and a combined measure of verbal and physical institutional misbehavior in a sample of 50 adolescent inmates. Significant associations were also identified between PCL-R scores and institutional charges for violent/aggressive behavior (r = .46) (Forth et al, 1990), violent institutional infractions (r = .39 for African American youth) (Hicks, 2000), and physically aggressive institutional infractions (r= .28) (Rogers, Johansen, Chang, & Salekin, 1997).
The relationship of juvenile psychopathy to measures of verbal and physical aggression parallels that identified in the adult psychopathy literature in both direction and magnitude. Lhe longitudinal stability of psychopathic characteristics identified early in life has not been established. For this reason, Edens et al. (2001) cautioned against the premature application of psychopathy measures for purposes of long-term prediction or decision making with long-term consequences. Lhe respective independent and interactive roles of Factor 1 (core personality traits) and Factor 2 (antisocial lifestyle) (or Factor 3— narcissism) traits in violent offending remain to be elucidated by future investigations. Forth (1995) identified substantially stronger associations of Factor 2 traits with a variety of measures of criminal offending (e.g., age of onset of offending) compared to Factor 1 traits, although number (r= .24) and variety (r= .22) of offenses were significantly associated with Factor 1 (i.e., personality) traits. Future studies should explore qualitative assessments of differences in the nature of criminal offending by psychopathic and nonpsychopathic youth. For instance, they should examine the interactive effects of factor traits on criminal behavior, particularly violence, and examine whether or not adolescent psychopaths commit more serious or instrumental violence than other youth offenders who engage in similar classes of crime.
Other Criminal Behavior
Measures of juvenile psychopathy are associated with many aspects of juvenile offending. In a sample of 130 adolescent offenders, Brandt etal. (1997) found PCL-R total, Factor 1, and Factor 2 scores correlated significantly with younger age at first arrest and with the number of prior incarcerations and crime severity. Ridenour, Marchant, and Dean (2001) reported that PCL-R scores predicted future sentencing rates of a sample of adolescent offenders beyond baseline number of delinquency charges and a continuous measure of disruptive behavior, whereas conduct disorder diagnoses did not. Christian, Frick, Hill, Lyler, and Frazer (1997) found that a psychopathic group of children had higher rates of lifetime school suspensions (55%), police contacts (36%), and parental psychopathy (40%), than youth who had low scores on one or both scales. Vincent, Vitacco, Grisso, and Corrado (2003) performed a cluster analysis and identified a 3-factor model of the PCL:YV. Lhe authors classified 259 male offenders(M age = 17.0, SD= 1.3) into 4 groups: those with low scores on all three scales, those with high scores on all three scales (i.e., the Psychopathy group), and those with predominantly impulsive or callous-deceitful traits. In general, the impulsive and psychopathic groups had the most severe offending histories. The psychopathic group had a substantially higher rate of violent recidivism (50%) than did the other 3 groups when followed an average of 14.5 months. The authors concluded that there is a subtype of adolescent offender with high impulsivity and callous-unemotional traits that is at particularly high risk for persistent and violent offending. Lynam (1997) reported that among community youth at high risk for delinquency, childhood psychopathy scale scores were positively related to seriousness of theft (r = .26) and seriousness of violence (r = .32), rates of general delinquency at age 10 (r = .32), and to the variety (r= .19) and seriousness (r = .39) of delinquency at age 13. Significant associations of CPS scores with measures of impulsivity and aggressiveness were also noted. Other studies indicate that psychopathic youth experience an earlier onset of criminal offending (McBride & Hare, 1996), engage in more frequent criminal behavior (Gretton, 1998), and are more likely to engage in intentional self-injurious behaviors (Gretton, 1998) than nonpsychopathic adolescent offenders (Forth & Mailloux, 2000).
Clinicians are understandably reluctant to diagnose children or adolescents with personality disorders given the many developmental transitions youth pass through that can produce disturbances mimicking personality disorder and the uncertain stability of any identified perturbations. Of course, it is often far from clear which signs and symptoms of personality disorder in youth are likely to reflect the presence of a relatively enduring personality dysfunction.
One of relatively few studies to examine comorbid personality disorders in relation to adolescent psychopathy is Meyers, Burket, and Harris’s (1995) evaluation of 30 consecutive youth admitted to an adolescent inpatient psychiatric program (M age = 15.3, SD = .99). Each adolescent completed standardized semi-structured interviews for the assessment of DSM-IV Axis I and Axis II (i.e., personality) disorders (APA, 2000). Youth diagnosed with conduct (N = 21), narcissistic (N = 4), and sadistic (N=2) personality disorders had the highest PCL-R scores—scores that were significantly higher than those of study participants without such diagnoses. Meyers et al. (1995) noted that adolescents diagnosed with narcissistic personality disorder had the highest PCL-R elevations, commenting that “psychopathy and narcissistic personality disorder share common ground in the areas of lack of empathy, exploitativeness, grandiose sense of self, feelings of entitlement, and a need for attention or stimulation” (p. 437).
Meyers and Blashfield (1997) examined 14 juvenile sexual homicide offenders, reporting that they averaged 2.3 major mental disorders and 1.9 personality disorders per person, substance use disorders (43%); attention-deficit/hyperactivity disorders (21%); and schizoid (38%), schizotypal (38%), and sadistic (31%) personality disorders predominated. PCL-R psychopathy scores were elevated (i.e., > 20) in 12 of the 14 youth. Contrary to expectation, Cluster A personality disorders such as schizoid and schizotypal personality disorder were more prevalent than Cluster B personality disorders in this sample. This was reflected in the paranoid ideation, odd beliefs, and social withdrawal exhibited by many of these youth. Few, if any, studies have examined the relationship of psychopathy factor subscales to personality disorders in youth offenders. Daderman and Kristiansson (2004) found no significant associations between a PCL-based measure of psychopathy and scores on the Karolinska Scales of Personality, although the sample was small and power was low to detect significant relationships.
Many theorists speculate that psychopaths have low levels of fear and anxiety that impair their ability to learn from aversive experiences (Lykken, 1995). Lynam (1997) found that psychopathy scores were significantly inversely associated with anxiety and internalizing disorders in a large sample of high-risk community youth. Frick (2002) also found an inverse association (r = -.28) between symptoms of negative affect (including anxiety) and scores on the APSD callous-unemotional factor subscale, when symptoms of conduct disorder were controlled for. The correlation between conduct disorder and anxiety symptoms was r = .51.
Moeller and Hell (2003) identified a significant positive correlation between PCL-R scores and number of prior traumatic events experienced; however, none of the psychopathic inmates they studied metDSM-IV criteria for posttraumatic stress disorder. These findings suggest that psychopathic offenders may be unlikely to develop posttraumatic stress disorder, despite life histories marked by repeated trauma. Moeller and Hell (2003) also found lower rates of affective disorders and suicide attempts in psychopathic offenders than in nonpsychopathic offenders. Studies of the prevalence and etiology of posttraumatic stress disorder in adolescent psychopathic and nonpsychopathic offenders are needed (Newman, 2002).
Ample evidence exists to support the association of comorbid attention-deficit/ hyperactivity disorder and conduct disorder with juvenile psychopathy. McBride (1998) observed that “a pattern of HIA [hyperactiv-ity-impulsivity-attentional deficits] and CP [conduct problems] is associated with an early onset of disruptive behavior, aggression, and an offending pattern marked by versatility and chronicity … the pattern of offending appears to be topographically similar to that uniquely associated with psychopathy” (p. 83). McBurnett and Pfiffner (1998) speculate that the arousal deficits and neuropsychological impairments observed in individuals with comorbid attention-deficit/hyperactivity disorder and conduct disorder both might play a role in antisocial behavior. Vitelli (1998) found that adult inmates with histories of both attention-deficit/hyperactivity disorder and conduct disorder were significantly more likely to be diagnosed with antisocial personality disorder and psychopathy as adults, and to have a history of committing violent acts in childhood, than inmates with only a conduct disorder. Recent findings provide further support for the notion that comorbid attention-deficit/hyperactivity disorder and conduct disorder is a particularly disabling syndrome associated with psychopathy-like features and far poorer outcomes than either disorder in isolation (Lynam, 1996, 1997, 1998, 2002).
Current findings are difficult to interpret with regard to comorbid personality disorders observed in psychopathic adolescents. There is some evidence for elevated rates of Cluster B (the impulsive or dramatic) personality disorders such as narcissistic personality disorder (Meyers et al., 1995), and Cluster A (the odd or eccentric) personality disorders such as schizoid and schizotypal personality disorders (Meyers & Blashfield, 1997) among juvenile psychopaths. Research supports an association between comorbid attention-deficit/hyperac-tivity disorder and conduct disorder and juvenile psychopathy. Comorbid anxiety disorders may be inversely associated with psychopathy, although findings are difficult to interpret given the widely varying definitions of anxiety and instruments for its assessment.
A number of studies have examined antisocial adolescents’ ability to modify previously rewarded behaviors when such behavior is no longer rewarded (Daughtery & Quay, 1991). Some theorists, using a neurobiological model of personality, have hypothesized that
antisocial individuals would have a “reward dominant” style in which their behavior is more dependent on appetitive drives than on avoidance of punishment…. ne would predict that antisocial individuals would be more likely than nonantisocial individuals to persist in a previously rewarded response, even if the rate of punishment for this response increased. (O’Brien & Frick, 1996, p. 224)
Recent investigations lend credence to this notion. O’Brien and Frick (1996) asked 132 youths ages 6 to 13 (92 clinic children and 40 normal controls) to complete four computer games with three potential levels of prizes attainable based on cumulative point totals. Participants began each game with 50 points and had a point added or subtracted from their point total following each trial. Across the 100 possible trials, the rate of rewarded trials per 10 trials declined from 90% for the first 10 trials to 0% for the last 10 trials and was independent of subjects’ actual responses. The total number of trials played served as the dependent measure.
Children with high scores on the APSD callous-unemotional subscale and no comorbid anxiety disorder displayed the most reward-dependent response orientations, compared to several other groups of clinic and community children with varying constellations of anxious conduct, and no symptoms. More anxious youths with conduct problems or psychopathy displayed a significantly less reward-dominant response style than did comparable non-anxious youth. Consequently, anxiety disorders were found to moderate the relationship between conduct problems/psychopathy and reward dominance. Psychopathic youth without anxiety disorders persisted in their efforts to obtain rewards despite being punished longer than did any other subgroup of youth offenders. Using a similar experimental paradigm, Lynam (1998) found that youths with both hyperactivity-impulsivity-attentional impairments and conduct problems displayed significantly more reward dominance than did youth with only one problem or neither problem. Lynam (1998) concluded,
The present results are consistent with theories that identify deficits in response modulation … as the fundamental deficit in psychopathy. Although the response modulation hypothesis is a somewhat narrower conception than the reward dominance hypothesis, both suggest that the primary deficit in psychopathy involves dysregulation of behavior in the face of a strong set for reward, (p. 572)
At present, the origins of reward dependence/response modulation deficits are unknown. Dadds and Salmon (2003) noted that these dispositions are often seen as biologically based aspects of personality (i.e., temperamental traits). Their review of a substantial body of findings suggests that environmental factors can influence individual sensitivity to punishment.
Research over the past decade has supported earlier conceptions of juvenile psychopathy. In 1964, McCord and McCord contended that juvenile psychopaths are excitement seeking, impulsive, aggressive, and callous. Recent research presents a somewhat more refined portrait of “fledgling psychopaths,” but substantially more investigation of their clinical characteristics is needed. Current findings indicate that psychopathic youth are substantially more likely to present with comorbid psychiatric disorders such as attention-deficit/hyperactivity disorder, conduct disorder, substance abuse/dependence, and other personality disorders, than are nonpsychopathic youth offenders or adolescent nonoffenders. Psychopathic youth exhibit a moderately greater propensity to violence and institutional violence or misbehavior, and earlier and more persistent and varied criminal careers than do nonpsychopathic delinquents. Psychopathic youth may evince more “egoistic” and less developed moral reasoning than do their general population counterparts, but it is currently unclear how, if at all, their ethical decision making differs from that of nonpsychopathic youth offenders (Edens et al., 2001). Although there is some support for the relationship of early life experiences such as child abuse to the development of psychopathy, the respective roles of genetic and environmental factors remain to be elucidated. That is, although research characterizing the clinical features of juvenile psychopathy has produced some important findings to date, far less has been accomplished with regard to the development of a convincing etiological account of juvenile psychopathy.
Labeling Effects of Juvenile Psychopathy Diagnoses
Recent findings suggest that youth diagnosed with psychopathy may experience a variety of adverse outcomes. An analysis of 424 juvenile detention and probation officers’ attitudes revealed that many (53.7% of juvenile probation officers and 37.9% of juvenile detention officers) believed that youth labeled psychopathic were unchangeable (Cruise, Colwell, Lyons, & Baker, 2003). More than three-quarters of these juvenile justice workers felt that incarceration was the best intervention for youth described as psychopathic.
Edens, Guy, and Fernandez (2003) presented 374 university undergraduates with one of two scenarios describing the case of a young man who was on death row for a murder committed when he was 16. Undergraduates presented with a case scenario including trait descriptions of the defendant consistent with psychopathy were significantly more likely (36.3% vs. 20.9%) to endorse the death penalty for the defendant than undergraduates presented an identical scenario describing nonpsychopathic personality traits. Undergraduates who ascribed psychopathic traits to the defendant were also more likely to endorse the death penalty, irrespective of the scenario conditions they read.
Mock juror studies using case scenarios describing adult defendants have yielded findings consistent with those obtained using adolescent psychopath scenarios. For example, Guy and Edens (2003) found that female jurors participating in a mock civil commitment trial of a sexually violent predator who were presented with a defendant described as psychopathic were significantly more likely to support commitment than women who were presented with a case description in which the defendant was described as at “high risk” for reoffending but not psychopathic. Mock juror studies suggest that there are powerfully negative consequences for defendants when they are labeled or are otherwise described as psychopathic.
Future Research Directions
The construct of psychopathy is a promising, though currently problematic, contribution to the study of serious, violent, and chronic youth offenders (Petrila & Skeem, 2003; Walters, 2004). More research is needed, particularly studies that examine the ethical, prevention, and rehabilitation implications of valid and false-positive diagnoses of juvenile psychopathy (Reed, 1996) and the nature, pervasiveness, and consequences of the stigma that characterizes public and professional perceptions of the disorder.
Future investigations should evaluate the temporal and cross-situational stability of behavioral and affective characteristics thought to comprise psychopathy and the concurrent/predictive validity of psychopathy measures designed for children and adolescents, including cutoff thresholds for psychopathy diagnoses and associated sensitivity and specificity rates (Seagrave & Grisso, 2002). Studies should be undertaken to assess the interactive effects and temporal stability of the factor traits that together are thought to constitute psychopathy (Lilienfeld, 1998) and the similarity of adolescent and adult psychopaths with regard to psycho-physiological, neuropsychological, psychiatric, autonomie, and affective characteristics (Forth & Burke, 1998).
Epidemiological studies should examine the nature, prevalence, and developmental manifestations of juvenile psychopathy and associated psychiatric disorders among girls and women (Lynam, 1996, 1997) and the relationship of ethnicity to psychopathy in a variety of adolescent offender and community samples (Seagrave & Grisso, 2002). Studies of the prevalence of psychopathic traits and frank psychopathy in large community samples of youth, might also offer new insights into the disorder and the factors that moderate its expression (Forth & Burke, 1998). Additional studies are needed that are similar to Frick, Kimonis, Dandreaux, and Farell’s (2003) investigation evaluating the 4-year stability of psychopathy in youth and predictors of stability. Frick et al. (2003) found that parental ratings of juvenile psychopathy were highly stable across assessment intervals and noted that the factors that predict stability of psychopathic traits may differ from those that predict the initial development of the disorder. Genetic studies of long-term behavior are needed to describe factors associated with different trajectories to antisocial and violent behavior (Schaeffer, Petras, Ialongo, Poduska, & Kellam, 2003; Simonoff, Elander, Holmshaw, Pickles, Murry, et al, 2004).
Additional research pertinent to adolescent psychopathy assessment is vitally important to better identify and treat juvenile psychopaths. Studies examining the interpersonal behavior and validity of self-reports of psychopathic youth are especially needed. Investigations incorporating various psycho-physiological and neuroimagining measures might also help to better distinguish psychopathic and nonpsychopathic youth offenders (Lynam, 1996, 1997, 1998).
Studies should assess the long-term clinical and criminological outcomes of youth with various configurations of psychopathy factor traits (e.g., low Factor 1-low Factor 2, high Factor 1-high Factor 2, low Factor 1-high Factor 2, high Factor 1-low Factor 2). Investigations of the role of parenting practices and other experiential factors, particularly deficits in early attachment and factors that mediate affective bonding to parents and others, on long-term outcomes vis-à-vis psychopathy are particularly needed (Saltaris, 2002).
Policy analyses evaluating the extent to which measures of psychopathy are currently being used to make transfer, decertification, and sentencing decisions involving youth are critically important in determining to what extent the construct is being applied prematurely or inappropriately in the juvenile justice system (Steinberg, 2001).
Finally, clinical responses of psychopathic youth to a range of pharmacologie, psychosocial, and combined treatment/management interventions are needed to determine whether these efforts can be successful with this youth population given the current level of knowledge regarding the disorder (Steinberg, 2001). To some extent, the relative dearth of studies examining treatments for psychopathic youth may reflect the therapeutic pessimism that has traditionally accompanied the diagnosis of psychopathy.
Promises and Perils of a Psychopathy of Crime: The Troubling Case of Juvenile Psychopathy
Contemporary clinical, legal, and social service education and practice could profit substantially from greater appreciation of scientific research pertaining to mental disorders. The promise of juvenile psychopathy research, specifically, is that it may eventually allow for effective early legal, psychological, and social intervention with youth who might otherwise proceed inexorably to adult psychopathy and to personally and socially costly criminal careers. However, given historical conceptualizations of the disorder and recent research findings, it is apparent that unbridled application of the construct within the juvenile justice system could result in the “writing off” of a significant number of youth. The most prudent course currently, given the uncertain state of scientific knowledge and highly stigmatized nature of the disorder, would seem to be to restrict application of the construct to research settings, pending additional studies that assess the stability of the disorder over a person’s life and its amenability to a range of prevention, treatment, and management approaches.