http://orcid.org/0000-0003-1793-2093
Abstract
Purpose: We investigated whether psychopathy-associated personality traits and behavioral styles affect the manner in which homicides are committed or the motives underlying them.
Materials and methods: Using three nationwide registries and an in-house homicide database based on court verdicts, we identified all cases of homicide in Sweden during the years 2007, 2008 and 2009. In 72 male offenders who had undergone assessment using the Psychopathy Checklist – Revised (PCL-R), the manner of homicide was categorized as instrumental or expressive, and the motive as belonging to one of five categories: (1) intimate-partner or family-related homicide; (2) homicide occurring during altercations, (3) robberies or burglaries, or (4) criminal conflicts; or (5) sexual homicide.
Results and conclusions: Offenders who had committed homicide in an instrumental manner or with a sexual motive had higher scores on PCL-R factor 1 than offenders displaying an expressive manner or other motives, suggesting that partially adaptive personality traits influence the crime-scene behavior of the former type of offenders more than maladaptive behavioral styles.
Introduction
In a recent meta-analysis examining the relationship between psychopathy and homicide in 2603 homicide offenders from 29 unique samples [Citation1], psychopathic features, as measured by the Psychopathy Checklist – Revised (PCL-R) [Citation2], were found to be much more common in homicide offenders than in a community sample of non-offenders. Aside from a large overall effect size (Pearson’s r = 0.68), after stratification for type of homicide, effect sizes were even greater for sexual homicide (r = 0.71), sadistic homicide (r = 0.80) and serial homicide (r = 0.74). Further, in the entire sample, homicide offenders had significantly higher scores on PCL-R factor 2, which comprises items related to an irresponsible lifestyle and antisocial behavior, than on PCL-R factor 1, which comprises items related to personality-based features of interpersonal and affective function. Although, in the meta-analysis, associations between homicide types and scores on PCL-R’s component factors were left unexplored, a recent study of subtypes of homicide offenders [Citation3] did not provide support for a distinction between primary and secondary forms of psychopathy – constructs associated, respectively, with factor 1 and factor 2 [Citation4].
Acts of violence such as homicide are traditionally categorized according to the manner of their commission, as being either instrumental or expressive. Instrumental (or predatory) acts of violence are premeditated attempts to exercise control over others or gain possession of something coveted; whereas expressive (or reactive) acts of violence are spontaneous reactions to perceived provocations [Citation5]. In a recent meta-analysis examining the relationship between psychopathy and instrumental and reactive violence in 8753 subjects from 55 unique samples [Citation6], aspects of social deviance captured by PCL-R factor 2 were found to be more important for reactive violence than instrumental violence. However, only 40% of the samples inventoried had been recruited from the general offender population; in addition, the included studies made use of a range of scoring instruments and outcome measures, resulting in marked between-study variability. Indeed, several studies in the meta-analysis based on offender samples reported associations between psychopathic features and instrumental violence. For example, in a study of 82 Dutch male inmates convicted of a violent or sexual offense [Citation7], instrumental violence was associated only with interpersonal traits belonging to PCL-R factor 1.
In the present study, in a sample of 72 male offenders convicted of having committed homicide during the years 2007, 2008 or 2009, we investigated the influence of scores on PCL-R factors 1 and 2 on crime-scene behavior. The aim of the study was to determine the degree to which personality traits and behavioral styles associated with psychopathy affected the manner in which homicides had been committed and the motives underlying them. Consistent with recent findings suggesting that most homicide offenders do not fulfill diagnostic criteria for psychopathy [Citation8], categories were not compared with regard to percentages of offenders with a total PCL-R score exceeding a stipulated cutoff.
Materials and methods
Homicide was defined as murder, infanticide, voluntary manslaughter or involuntary manslaughter by means of assault. Three national registries and an in-house homicide database compiled using court verdicts and police files were used to catalog information on offenders and victims in all solved cases. Personal identity numbers provided to residents of Sweden enabled linkage of registries to the database. First, the registry of forensic autopsies administered by the National Board of Forensic Medicine (NBFM) was used to identify victims who had died of homicide. Second, data on offenders’ criminal histories were collected through the National Council for Crime Prevention’s Crime Registry, which includes information on all criminal convictions since 1973. Third, a registry of forensic-toxicological findings administered by the NBFM was used to investigate whether offenders and victims had been under the influence of alcohol or narcotic substances during the index homicide. Fourth, from the homicide database, information was collected allowing categorization of the manner of each homicide as either instrumental or expressive, and the motive of each homicide as belonging to one of five categories: (1) intimate-partner or family-related homicides; (2) homicides occurring in the context of altercations, (3) robberies or burglaries, or (4) criminal conflicts; and (5) sexual homicides. Categories were mutually exclusive; in instances in which more than one category were theoretically applicable, the fact that the database contained extensive information from court verdicts and police files – moreover, solely from solved cases – allowed identification of a principal manner or motive.
Between 1 January 2007 and 31 December 2009, 264 homicides were committed in Sweden, 228 of which were solved by the following means: court verdict (n = 213); suicide by offender in connection with the offense (n = 13) or while in custody (n = 1); or death of offender before police arrest (n = 1). Because some of the homicides involved multiple perpetrators, the total number of offenders in solved cases was 257. A total of 73 offenders, all male, had undergone assessment using PCL-R – in all cases, by means of both interview and review of collateral information – either during the course of a forensic-psychiatric evaluation (n = 8) performed by the NBFM, or in conjunction with a psychological assessment (n = 65) conducted at the National Reception Unit of the Swedish Prison and Probation Services. After exclusion of one offender, whose offense (an attempted spree killing resulting in a single homicide) featured a motive that did not belong to one of the five categories, ultimately, 72 offenders were included in the study.
In two separate analyses, scores on PCL-R factors 1 and 2 were compared across subgroups of offenders after categorization according to homicide manner and homicide motive using, respectively, an unpaired t test and one-way analysis of variance. Two-sided probability values less than .05 were considered statistically significant.
Results
Characteristics of offenders and index offenses are presented in . Mean total PCL-R scores, as well as mean scores on component factors 1 and 2, after categorization of offenders according to homicide manner and motive, are presented in . Offenders whose homicide manner was categorized as instrumental had significantly higher scores on factor 1 than offenders whose homicide manner was categorized as expressive (9.8 (SD = 4.2) vs. 5.1 (SD = 4.5); t(70) = –3.9; p = .0002); whereas, for scores on factor 2, the groups did not differ significantly (10.9 (SD = 5.5) vs. 9.2 (SD = 6.3); t(69) = –1.0; p = .3). Regarding homicide motive, scores differed significantly between groups for both factor 1 (F(4, 67) = 4.5; p = .003) and factor 2 (F(4, 66) = 3.9; p=.006). Post hoc tests using Bonferroni’s method demonstrated that offenders who had committed sexual homicide had significantly higher scores on factor 1 than both offenders who had committed intimate-partner or family-related homicide (12.6 (SD = 2.3) vs. 3.3 (4.0); p = .002) and offenders who had committed homicide in the context of an altercation (12.6 (SD = 2.3) vs. 5.8 (SD = 4.6); p = .02); and that offenders who had committed intimate-partner or family-related homicide had significantly lower scores on factor 2 than offenders who had committed homicide in the context of an altercation (3.7 (SD = 4.2) vs. 10.8 (SD = 6.1); p = .006) or a criminal conflict (3.7 (SD = 4.2) vs. 11.2 (SD = 5.2); p = .03).
Table 1. Characteristics of 72 male homicide offenders and index offenses.
Table 2. Mean total Psychopathy Checklist – Revised (PCL-R) scores and mean scores on component factors 1 and 2, after categorization of 72 male homicide offenders according to motive and manner.
Discussion
In this registry-based study, we compared scores on PCL-R component factors between groups of male homicide offenders categorized by the manner and motive of their offenses. Mean scores on factor 1 of sexual-homicide offenders were significantly higher than the scores of offenders who had committed intimate-partner or family-related homicide or homicide in the context of an altercation. Also, the mean scores on factor 2 of offenders who had committed intimate-partner or family-related homicide were significantly lower than the scores of offenders who had committed homicide in the context of an altercation or a criminal conflict. Thus, our findings appear to be in line with those of the study of violent and sexual offenders mentioned above [Citation5], in which psychopathy-associated personality traits captured by factor 1, rather than maladaptive behavioral styles captured by factor 2, predisposed offenders to a predatory manner of violence.
Interestingly, sexual-homicide offenders also had the highest mean score on factor 2, possibly in line with results of a recent study of sexual offenders [Citation9] in which impulsivity was identified as a personal characteristic that appeared to explain why some offenders had killed their victims. Yet, whereas, in that study, impulsivity was defined as the presence of a clinical diagnosis of an impulse disorder, in the present study – in which item-level score comparisons were not performed – the characteristic was captured by the factor 2 lifestyle item termed impulsivity. That, in the present study, sexual-homicide offenders’ mean score on the lifestyle-related and behavior-related items of factor 2 did not differ greatly from mean scores of offenders who had committed homicides in the context of altercations, robberies or burglaries or criminal conflicts is consistent with the conclusion of a recent study [Citation10] that, based on their criminal careers, sexual-homicide offenders ‘should be considered more as murderers than sex offenders.’
A novel finding of the present study was that psychopathy-associated traits and styles also appear to influence motives underlying homicide. Predictably, perhaps, maladaptive behavioral styles were of less importance for homicides committed against intimate partners or family members than for other homicide motives. Indeed, the gradient of mean scores on PCL-R factor 1 across homicide motives – from 3.3 for intimate-partner or family-related homicides and 5.8 for homicides committed in the context of altercations, to 8.7 for homicides committed in the context of robberies or burglaries and 12.6 for sexual homicides – suggests a corresponding gradient, across motives, of emotional detachment and narcissistic dominance. It is these latter two features of the psychopathy construct that, in the triarchic model of psychopathy proposed by Patrick and coworkers [Citation11], constitute the partially adaptive phenotypic facet ‘boldness.’ It has been suggested that factor 1’s boldness-related traits – specifically, items belonging to the factor’s first, interpersonal facet such as glibness or superficial charm and a grandiose sense of self-worth – may differentiate the psychopathy construct from antisocial personality disorder, an entity that, according to conventional definitions, overlaps greatly with PCL-R’s factor 2. Indeed, the discriminative value of facet-level PCL-R score comparisons has been empirically demonstrated in two recent studies. In a study of violent offenders [Citation12], the second, affective facet of factor 1 was found to be associated with problems with emotion regulation. In a second study of offenders that made use of canonical correlation analysis [Citation13], elevations on the interpersonal facet of factor 1 and the antisocial facet of factor 2 were associated with symptoms and signs of paranoid, narcissistic, histrionic and antisocial personality disorders; while, at the same time, high levels on the antisocial facet, in combination with low levels on the interpersonal facet, were associated with phenomena related to borderline personality disorder.
Thus, an obvious limitation of the present study is the lack of facet-level score comparisons. A second limitation is that only roughly one third of offenders in solved homicide cases – none of them women – had undergone PCL-R assessment and were thereby eligible for inclusion in the study. Also, assessments had been performed in correctional settings, each by one member of a staff of clinicians who, at the time of each assessment, was aware of the circumstances surrounding the index homicide. Additional limitations include the use of a single instrument to assess personality traits and behavioral styles, as well as the lack of a comparison group composed of non-offenders. A strength of this study is, on the other hand, its population-based design, whereby, through the use of nationwide registries and an in-house database, all cases of homicide committed in Sweden during a three-year period were identified. Moreover, because the database had been compiled using court verdicts and police files, we were able to restrict the study to subjects who had been convicted, rather than merely suspected, of having committed homicide. Finally, our study is part of a growing body of scholarship [Citation14,Citation15] examining psychopathy-associated personality traits and behavioral styles in sexual offenders in Sweden.
Disclosure statement
The authors declare no conflicts of interest.
Additional information
Notes on contributors
Sara Rodre
Sara Rodre, MSc Psych, is a psychologist, Swedish National Board of Forensic Medicine.
Jonatan Hedlund
Jonatan Hedlund, MD, PhD, is a forensic psychiatrist, Swedish National Board of Forensic Medicine.
Jenny Liljeberg
Jenny Liljeberg, MSc Psych, PhD, is a psychologist, Swedish National Board of Forensic Medicine.
Marianne Kristiansson
Marianne Kristiansson, MD, PhD, is a forensic psychiatrist, Swedish National Board of Forensic Medicine.
Thomas Masterman
Thomas Masterman, MD, PhD, is a forensic psychiatrist, Swedish National Board of Forensic Medicine.
Joakim Sturup
Joakim Sturup, PhD, is a crimonologist, Swedish Police Authority.
References
- Fox B, DeLisi M. Psychopathic killers: a meta-analytic review of the psychopathy-homicide nexus. Aggress Violent Behav. 2019;44:67–79.
- Hare RD. The Hare Psychopathy Checklist – Revised. 2nd ed. Toronto (Canada): Multi-Health Systems; 2003.
- Pajevic M, Batinic B, Stevanovic N. Subtypes of homicide offenders based on psychopathic traits. Int J Law Psychiatry. 2017;55:45–53.
- Skeem J, Johansson P, Andershed H, et al. Two subtypes of psychopathic violent offenders that parallel primary and secondary variants. J Abnorm Psychol. 2007;116:395–409.
- Adjorlolo S, Chan HC. The nature of instrumentality and expressiveness of homicide crime scene behaviors: a review. Trauma Violence Abus. 2017;18:119–133.
- Blais J, Solodukhin E, Forth AE. A meta-analysis exploring the relationship between psychopathy and instrumental versus reactive violence. Crim Justice Behav. 2014;41:797–821.
- Declercq F, Willemsen J, Audenaert K, et al. Psychopathy and predatory violence in homicide, violent, and sexual offences: factor and facet relations. Leg Criminol Psychol. 2012;17:59–74.
- Boduszek D, Debowska A, Willmott D. Latent profile analysis of psychopathic traits among homicide, general violent, property, and white-collar offenders. J Crim Justice. 2017;51:17–23.
- Healey J, Beauregard E. Impulsivity as an etiological factor in sexual homicide. J Crim Justice. 2017;48:30–36.
- Beauregard E, DeLisi M, Hewitt A. Sexual murderers: sex offender, murderer, or both? Sex Abuse. 2018;30:932–950.
- Venables NC, Hall JR, Patrick CJ. Differentiating psychopathy from antisocial personality disorder: a triarchic model perspective. Psychol Med. 2014;44:1005–1013.
- Garofalo C, Neumann CS, Velotti P. Difficulties in emotion regulation and psychopathic traits in violent offenders. J Crim Justice. 2018;57:116–125.
- Klipfel KM, Garofalo C, Kosson DS. Clarifying associations between psychopathy facets and personality disorders among offenders. J Crim Justice. 2017;53:83–91.
- Dåderman AM, Jonson C. Lack of psychopathic character (Rorschach) in forensic psychiatric rapists. Nord J Psychiatry. 2008;62:176–185.
- Sturup J, Rodre S, Karlberg D, et al. Male-on-female sexual homicides in Sweden, 1990 to 2013: a population-based controlled study of incidents, victims, and offenders. Int J Offender Ther Comp Criminol. 2018;63:1557–1574.