Crime and Personality: Personality Theory and Criminality Examined
Concerning the fear that policy implications would neglect social problems, much to the contrary, personality theory’s focus on the interactional effects of individuals with their environment has led to the careful development of research implications involving the family, the school environment, as well as boarder community and societal institutions (Andrews & Wormith, 1989; Eysenck, 2003; Moffitt, 1993, 2006, 2007; Farrington, 2002; Farrington & Jolliffe, 2004). Childhood dysfunctional is attributed to exposure to parental violence, poor parental-child attachment, and abuse (Caspi et al., 1994; Moffitt, 1993). Childhood vulnerabilities are considered to be resultant of maternal drug use, poor nutrition, exposure to toxins, and deprivation of affection (Caspi et al., 1994; Farrington, 2002; Moffitt, 1993, 2006). Moffitt (2007) predicted higher prevalence of delinquency among ethnic minorities due to an increased likelihood of experiencing predicted risk markers (i.e., poor prenatal care, exposure to toxins, family adversity), which minorities are subjected to at elevated levels due to poverty and institutionalized discrimination. Moffitt (2006) also emphasized the importance of the prevention of entrapment of adolescents in detrimental snares, such as incarceration, teen pregnancy, or drug addiction. Therefore, policy implications generated from personality theory have not been limited to interventions at the individual level, but policy recommendations have been multi-dimensional, including interventions at the family, school, and community levels (Caspi, Lynam, Moffitt, & Silva, 1993; Lynam et al., 2000; Farrington, 2002).
However, vigilance is recommended regarding stigmatization resulting from labeling individuals with deviant personality disorders. In England, the presence of such diagnoses justifies non-engagement in treatment by mental health professionals with such clients based on the Mental Health Act of 1983, with its so-called “treatability” criterion (Appelbaum, 2005, p. 397; Feeney, 2003). Modifications in criminal law in England now permit indefinite detention of persons who are thought likely to represent a serious threat (Appelbaum, 2005; Seddon, 2008). Such persons must have a severe personality disorder diagnosis that makes her or him “more likely than not to commit an offence that might be expected to lead to serious physical or psychological harm from which a victim would find it difficult to recover” (Appelbaum, 2005, p. 398; Seddon, 2008). In clinical professions, “the most controversial issue has been whether a connection exists between serious mental disorder and violence” (Melton, Petrila, Poythress, & Slobogin, 2007, p. 318; Webster & Hucker, 2007). Yet, the courts in England are now requiring such evaluations and risk assessments from mental health professionals (Appelbaum, 2005; Seddon, 2008). Sexual predator statutes that have been passed in several U.S. states are similar to these new laws in England but have a much more limited scope (Appelbaum, 2005).
Regarding fears of undue leniency from the justice system due to labeling criminals with personality disorders, the pendulum of justice seems to be swinging in a more punitive direction. For example, the acquittal of John Hinckley in 1982 set off a groundswell of criticism against the insanity plea (La Fond & Durham, 1992; Melton et al., 2007). As a result, 39 states made dozens of changes in their laws regarding insanity pleas, Utah and Idaho completely abolished the defense at criminal trials, and the U.S. Congress passed legislation that made changes in the way the insanity plea may be used in criminal court (La Fond & Durham, 1992; Melton et al., 2007).
Beyond criticisms regarding the underlying assumptions or potentially negative impacts of personality theory on policy, criticisms were made in the past regarding research methods in personality theory research (Miller & Lynam, 2001). Researching personality traits was considered by many as a mere circular blaming exercise, condemning criminals for their own criminal propensities (Glicksohn, 2002). Trait theory was characterized as hypothesizing, “delinquents are delinquent because they have delinquency within them” (Jones, 2008, p. 12). In a review of personality research, Miller and Lynam (2001) and Roberts (2009) noted key methodological weaknesses in early personality research that planted doubts among criminologists regarding its reliability and validity, including: the construct validity and reliability of personality trait measures, doubts about the validity of projection tests, and concerns with scale construction through “criterion keying” (p. 766). Caspi et al. (1994) noted that critics of previous personality research highlighted problems with the measurement of personality and delinquency and biased sampling. Past methodological questions regarding measures of personality through the use of criterion keying specifically centered on the tautological problem with measuring the personality dimension denoting delinquency (Capsi et al., 1994; Miller & Lynam, 2001). As several measures, such as the Minnesota Multiphasic Personality Inventory (MMPI) and the California Psychological Inventory (CPI), were designed to differentiate delinquents from nondeliquents, items in the scales were tautological and limited to demonstrating that the most delinquent individuals assessed were in fact the most similar to the definitions of delinquency in the scales (Caspi et al., 1994; Miller & Lynam, 2001).
Past critics of personality research also were suspicious of the use of official reports for the offending outcome measure that could have created an underrepresentation of offenders who do not get caught or even that such undetected offenders may be inadvertently placed in the nonoffending control groups (Caspi et al., 1994). Also, criticisms were made of the use of self-report measures of offending for including trivial items, plus fears were raised concerning the underreporting of offending (Caspi et al., 1994). Although efforts have been made to correct these weaknesses in measurement by using multiple offending indicators (Caspi et al., 1994), these issues do not uniquely belong to personality researchers, but present a universal difficulty in criminological research (Cantor & Lynch, 2000; Geerken, 1994; Huizinga, & Elliott, 1986; Thornberry & Krohn, 2000). In fact, the Gluecks, renowned yet marginalized for their research on criminal careers, were collecting multi-informant indicators for delinquency in addition to using official records in their very earliest research (Laub & Sampson, 1991).
Criticisms regarding biased samples were due to the primary use of incarcerated populations that only represented offenders who were detected and sentenced (Caspi et al., 1994). Differences in the characteristics of detected and nondetected offenders were presumed to be spurious or confounding variables that could have influenced the study findings (Caspi et al., 1994). Also race, gender, and class biases in samples were noted (Caspi et al., 1994).
Overcoming the Methodological Weaknesses in Current Research
In their meta-analytic review of studies of personality and ASB, Miller and Lynam (2001) concluded that these previously mentioned weaknesses within personality research have been overcome (see also, Roberts, 2009). As far as the construct of personality, Miller and Lynam (2001) concluded that the structural models of personality that they included in their review (see Table 1) have been sufficiently tested and “warrant confidence in their reliability and validity” (p. 768). In regards to the use of projection tests, such tests are no longer viewed as reliable measures of personality (except by the few psychoanalysts in the field) and are no longer used in research (Miller & Lynam, 2001). To avoid the problem of tautology, Caspi et al. (1994) used assessment measures that were not designed to differentiate delinquents from nondelinquents, but instead they utilized instruments that assessed an array of traits that may be linked to ASB. Miller and Lynam (2001) corrected the tautological issue by including the criterion-keyed scales only as measures of delinquency, not as a personality trait. Moffit, Caspi, Rutter, and Silva (2001) similarly used conduct disorder as a measure of delinquency not a personality predictor.
In the meta-analytic review of over 60 studies investigating the relationship between personality traits and ASB conducted between 1963 and 2000 based on the four structural models in Table 1, the largest effect sizes (greater than .30) were found for psychoticism, agreeableness, and novelty seeking (Miller & Lynam, 2001). Interestingly, variations across gender and sample type were noted. For example, neuroticism was only weakly related to ASB in female samples and extraversion was more strongly related to ASB in nonprisoner and self-report samples (Miller & Lynam, 2001). As evidenced by this review and other reviews, variation in sample type has increased within these types of studies (Caspi et al., 1994; Miller & Lynam, 2001). In summarizing the results of the review (using FFM as a common denominator for all four models), these findings suggested that individuals low in agreeableness (i.e., those hostile and indifferent to others) and low in conscientiousness (i.e., those with poor impulse control and lacking in motivation) are more likely to engage in ASB (Miller & Lynam, 2001).
The personality trait of impulsivity or the FFM dimension of low conscientiousness (see Table 1) is the trait that has been the most consistently liked to ASB (Farrington & Jolliffe, 2004; Farrington, 1992, 2002; White et al., 1994). Advances in brain imaging have allowed researchers to investigate the connection between frontal lobe executive functioning and impulsivity (Moffitt, 1990; Eme, 2008), suggesting that hyperactivity and a poor ability to perceive potential consequences of actions may result in school failure (Farrington, 2002). Although Gottfredson and Hirschi (1990) are sociologists by training, they also argued that people differ in an underlying criminal propensity, i.e., low self-control. According to self-control theory, the formation of the stable trait of self-control occurs via the parent-child socialization or the reciprocal bonding process from birth to age eight (Gottfredson & Hirschi, 1990; Gottfredson, 2007). From these early attachments or social bonds, an individual gains self-control or “the tendency to delay short-term gain for long-term personal and collective interests” (Gottfredson, 2007, p. 537). Gottfredson and Hirschi’s theory has been empirically supported through numerous studies across ages, groups, cultures, crimes, behaviors related to crime, and time periods (Gottfredson, 2006; Pratt, 2009). A meta-analysis of the theory found a combined effect size from 21 studies of .20, regarded as one of the strongest documented micro-level correlates of crime (Pratt & Cullen, 2000). An exception to the link between low self-control and crime has been noted in white-collar criminals, with findings suggesting that they are high in conscientiousness, but low in integrity (Blickle, Schlegel, Fassbender & Klein, 2006).
Low empathy also has been thoroughly investigated as a predictor of ASB, but empirical evidence relating measures of empathy to ASB is not conclusive or consistent (for reviews, see Jolliffe & Farrington, 2004; Lovett & Sheffield, 2007; Varker, Devilly, Ward, & Beech, 2008). In a review and meta-analysis of 35 studies, low empathy was found to be strongly related to violent offenses, but relatively weak to sex offenses. Yet, when controlling for intelligence and SES, the relationship between low empathy and offending disappeared (Jolliffe & Farrington, 2004). In a review of 17 studies by Lovett and Sheffield (2007), no relationship between low empathy and aggression was found in children, yet a relationship was found with adolescents.
Although the issue of tautology is still remains in the Hare Psychopathy Checklist-Revised (PCL-R) due to its behavioral component, psychopathy as measured on the PCL-R has been consistently linked to offending and recidivism (Hemphill, Hare, & Wong, 1998; Porter & Woodworth, 2006; Salekin, Rogers, & Sewell, 1995). Based on a review of psychopathy and crime, Hare (1993) estimated that psychopaths commit more than 50% of all crime, making up a high percentage of violent offenders, law enforcement killers, serial rapists, and wife batterers. This finding is congruent with studies suggesting that a relatively small proportion of offenders account for a large proportion of crimes (Farrington, Ohlin, & Wilson, 1986).
Also supporting Hare’s proposition of the complex clinical condition of psychopathy, numerous studies using data from magnetic resonance imaging of the brains of those classified as psychopaths have revealed significant abnormalities in brain structure (Herba et al., 2007). For example, studies based on both community and offender samples have found reduced connectivity and abnormalities in the amygdala-orbitofrontal cortex network (Yang, Raine, Colletti, Toga, & Narr, 2009) and damage to the prefrontal cortices (Craig et al., 2009). Such defects in brain structure have been shown to result in aggressive behaviors, loss of normal anger and fear responses, decreased inhibition, disruption in decision making, and impairment in processing the emotional and social contexts (Craig et al., 2009; Yang et al., 2009).Continued on Next Page »