Women convicted for violent offenses: Adverse childhood experiences, low level of education and poor mental health

A substantial increase in female offending, e.g. in England and the United States over the past two decades [1, 5, 6], has spurred the interest of researchers into learning more about the characteristics of female offenders, the kind of crimes women are more likely to commit, the circumstances which drive women to offending, and the factors that determine the risk for repeat offending (e.g. Steffensmeier [7]).

The origins of female offending have been explored from various perspectives and within different disciplines. The emancipation theory put forth by Adler [8] argues that as the gender roles of women change and become more and more similar to those of men, so will their participation in criminal activity - a theory not supported by criminal statistics. From a different angle, South Richardson and Hammock [9] examined whether the social role men and women have in society influences male and female aggression in connection with criminal behavior. Campbell, Muncer and Bibel [10] took an evolutionary approach and argued that the scarcity of resources drives women to perpetrate crimes. From a cognitive point of view, Bennett, Farrington and Huesmann [2] explained that women have lower rates of offending due to the fact that they have better prosocial skills than men and because they acquire these social cognitive skills earlier on in life, in comparison to their male counterparts. Hollin and Palmer [11] discussed the criminogenic needs of women and whether they are specific to women only.

One of the main questions being asked in the literature is to which extent a gender-neutral approach would suffice and for which areas gender-specific aspects would be necessary in order to create a model for female offending. Many of the researchers point out that the origins of male offending cannot simply be transferred to female offending without necessary investigation as entirely different determinants might be responsible for it [1‐3, 11, 12]. Other researchers argue that these gender-specific theories do not account for the full range of female offending and that gender-neutral theories would suffice to explain both female and male criminality [13, 14].

When reviewing the literature on male and female offending, it becomes apparent that there are similarities, as well as differences between the both, regarding the characteristics of the offense, the offender and the victim. It seems that men and women mostly commit the same types of crimes, but that the rates of female offending are substantially lower than male offending across all crime categories [7, 13, 15]. For both male and female adult offenders, we can see the predominance of property and drug offenses [12, 13], but it appears that the gender gap is greatest for violent offenses: Women generally seem to have much lower crime rates for violent offending than men [3, 7, 15‐18]. While some researchers have found an indication towards an increase of female violent offending [4, 19] others claim that the rate of female violent crime has remained stable [3, 5].

The current literature supports the hypothesis that female violent offenders display a different offense pattern in comparison to male violent offenders. Female violent offending is often characterized by the victims of women often being people close to them, the crime a result of interpersonal conflict - this includes parents, husbands and boyfriends, as well as children [9, 10, 13, 18, 20‐22]. In comparison to men, women are more likely to commit a violent crime at home. Weizmann-Henelius et al. [18] concluded that female violent behavior more often leads to the death of a close victim than that of an acquaintance or stranger. In a study by Roe-Sepowitz [22], it was documented that females mostly offended violently during the perpetration of another crime, such as robbery. There are, however, some similarities in male and female violent offense patterns: Both men and women commit crimes within the context of an argument and the victim is more likely to be male [20].

Concerning socio-demographic characteristics, both female and male violent offenders were often found to have a history of adverse childhood domestic constellations (absent and/or criminal parents, neglect), poor socio-economic backgrounds, equal rates of previous psychiatric hospitalization, to be less educated, and under- or unemployed at the time of offense [6, 7, 14, 21]. In comparison to male offenders, female offenders were more likely to be older, married and have children. They were also more likely to be presently victims of physical, psychological and sexual abuse, as well as childhood physical, psychological and sexual abuse. They showed also elevated levels of substances abuse, and were more frequently suffering from physical and mental illness [3, 5, 7, 11, 18, 23]. Yourstone [21] established that female offenders were less likely to be foreign nationals than male offenders. Female offenders have been found to have less prior convictions [5, 21], have lower rates of psychopathy (see the Hare Psychopathy-Checklist-Revised [24]) [14, 25, 26] and are less likely to recidivate than male offenders [7].

The aim of the present study was to assess the prevalence of female violent offending in a representative population of Swiss violent offenders. A second aim was to assess socio-demographic, psychiatric and offense-related characteristics of female violent offending in Switzerland and to compare these female offenders with their male counterparts.

The sample comprised all female and male violent and sex offenders that were convicted between August 2000 and December 2002 in the Canton of Zurich, Switzerland. The Canton of Zurich is an urban, sub-alpine area with a population of 1.2 million inhabitants. The concrete inclusion criterion was: Sentenced to a minimum prison sentence of ten months due to a violent or sex offense, or to court-ordered therapy by a court in the Canton of Zurich between August 2000 and December 2002. Two hundred and three offenders matched these criteria and have been included in the study sample.

Data was collected from correctional and court files by five research assistants, all trained psychologists holding the degree of Master of Science (MSc). There was no direct contact with the offenders. The files contained comprehensive personal details, including criminal history, the exact type and circumstances of the offense, as well as personality characteristics and psychiatric diagnoses (if any). Based on this information, socio-demographic characteristics (e.g. gender, age at the time of the index offense, level of education, marital status), childhood conditions (e.g. sexual and/or physical abuse, living in a foster home), psychiatric characteristics (e.g. diagnoses - if any, previous hospitalization, family history of psychiatric disorders, substance abuse), and offense-related characteristics (e.g. number and type of previous offenses; characteristics of the index offense: relationship to the victim, number of victims, degree of injury of the victim, type of offense and sentence) were collected. The interrater reliability was satisfactory (Kappa > 0.65).

Descriptive and bivariate logistic regression analyses were carried out with STATA 10.0.

The sample of the present study is a subsample of a large epidemiological study conducted on convicted offenders (inmates as well as offenders on probation) in the Canton of Zurich in the year 2000, which was approved as a whole by an external Ethics Committee (Kantonale Ethikkommission Zürich). In agreement with the committee, no informed consent needed to be obtained as there was no contact with any of the study subjects. All data was collected entirely from the subjects' files and anonymized before further analysis.

With 187 (92%) of the 203 offenders, there were distinctly more male than female (7.9%, N = 16) offenders. The mean age of the sample was 32.5 years (SD = 10.8) with a range from 18 to 65. Male and female offenders did not differ significantly concerning their age (female: 31.5 years, male: 32.7 years).

Fifty-five point five percent (N = 112) of the offenders were Swiss nationals. Female offenders were 3.8 times more likely to be Swiss nationals than male offenders (OR = 3.8, p < .05): Among the female offenders, 81.3% (N = 13) were Swiss and among the male offenders 53.2% (N = 99).

Furthermore, there were differences regarding the civil status at time of the offense: The odds of a female offender being married was 2.9 times higher than for a male offender (OR = 2.89, female: 50%, N = 8; male: 26%, N = 48). Thirty-nine percent (N = 78) of the offenders had at least one child. Male and female offenders did not differ with respect to this criterion.

Female offenders were consistently less likely to finish their school education (OR = 0.18). Only one third (31.3%, N = 5) of the female offenders completed their school education, whereas 72% (N = 126) of the male offenders hold the equivalent of a high school diploma. However, male and female offenders did not differ significantly concerning their vocational education (female: 43.8%, N = 7; male: 51.7%, N = 93).

Table 1 gives an overview of the socio-demographic characteristics of the study sample stratified for gender as well as the results of the bivariate logistic regression analysis using gender as the dependent variable.