My professional background (though not my personal, I hasten to add) is in offending and it remains an area which fascinates me. A lot of the research in the area is obviously dedicated to reducing offending, be that in terms of severity or frequency, and a significant amount tries to identify factors which increase the risk of offending, i.e. ‘risk factors’. Recently, an article was published in Psychology of Women Quarterly. The paper investigated the effect of trauma and mental illness in imprisoned women using a mixed methods design. As a general rule, research is divided into ‘quantitative’ (typically employing questionnaires or other measurable outcomes which can then be analysed statistically to ascertain whether there has been a change in whatever has been investigated) and ‘qualitative’, which often uses interviews or focus groups. Crudely, one involves numbers and the other involves words. Both have their pros and cons but it seems to me that quantitative research has more adherents and I sometimes worry that by focusing more on numbers we lose the richness of qualitative research, which can provide extremely in-depth data and which is equally valid, despite the lack of statistics.
This study is relatively unusual in that it combines qualitative and quantitative methods; thus using a ‘mixed methods design’. Up to 90% of women prisoners experience physical or sexual violence in the year preceding imprisonment and many have experienced chronic lifetime abuse. Those abused or neglected in childhood are more likely to offend and to be arrested as adults. For many women, delinquent behaviour can be directly linked to the traumatic event, e.g. running away, using drugs, etc. It’s well-known that trauma experiences are common in those who offend (but, crucially, not all those who experience trauma will go on to offend). Secondly, trauma and mental disorder, be it depression, psychosis or personality disorder, are strongly associated. This paper, then, looks at the effects of both trauma and mental illness (MI) in female inmates.
The authors found that 85% of their sample had a history of substance misuse disorders and that half had a history of post-traumatic stress disorder (PTSD). PTSD can only be diagnosed if you have experienced life-threatening violence or a natural disaster, so rape at knifepoint would warrant a diagnosis but chronic, non-violent sexual assault would usually not. In this case, half the women had had PTSD but even more would have experienced acutely traumatic events without developing PTSD. Half had also experienced a MI such as depression or psychosis, both of which are also associated with trauma experiences. Conclusion: major trauma is very common in female prisoners. Specifically in this sample, 86% had experience sexual violence, and 77% had experienced domestic violence. Only 2 of the sample (of a total of 115) had not been victimised in any way at any point in their lives.
This is all interesting enough, but it’s simply associations; it doesn’t tell you what causes what. Similarly, there’s an association between age and height, but without statistical analysis it’s impossible to see the direction of the relationship. Regression analyses are designed to do just this. They look at factors as predictors and outcomes to show you what the biggest predictor is. They also allow you to control for baseline factors. So for example you could look at the relationship between mental illness and gender (predictors) on reoffending (outcome) but since people who have a history of offending a lot are more likely to reoffend a lot you can might also adjust the analyses for baseline rates of offending to even the analysis out.
Unsurprisingly, women with substance problems were more likely to engage in sex work and driving under the influence of alcohol or drugs. Domestic violence was associated with property crime, drug offending, and commercial sex work and it looks as though these men acted as partners, drug dealers, partners in crime (as it were) and pimps. Witnessing violence was associated with property crimes, engaging in assaultive behaviours and use of weapons, sometimes in self-defence and sometimes in defence of others, such as parents or grandparents. Those who had experiences of caregiver violence were much more likely to run away during adolescence; 4 times as likely as those who were not maltreated. Whilst for some this will allow them to escape being at home, we also know that living on the street is associated with substance use and other offending behaviour, as well as placing the person at greater risk of violence.
This paper looked at a number of mental illnesses but not at anxiety or personality disorders. Up to 70% of inmates have a personality disorder (PD) and we know that trauma is implicated in the development of such disorders. The fact that it wasn’t included means that it’s possible that it s PD which is the underlying factor and not necessarily the trauma experience. Nevertheless, it seems that several life events significantly increase the risk of offending and incarceration, including abusive childhood experiences, violent relationships, criminal environments and self-medicating with substances. It also seems that life events such as running away from home following maltreatment are critical.
So pragmatically, what does this mean? Well, for a start, it seems that MI is common in prison, therefore there need to be more funds allocated to assessing and treating it. Secondly, if people have histories of abuse, neglect and maltreatment, how is prison going to care for them? Will it provide consistent care to try to rebuild people’s abilities to form positive relationships? Given that early experiences are so crucial to adult functioning, how can we identify maltreatment more effectively and intervene to reduce the impact? Finally, how will social care, education, housing and substance misuse services join together to create ‘packages’ of care and support for those at the greatest risk of MI and offending? We try to implement such care packages in the UK (it is of note that this study used an American sample) but we still have a significant female offender population and we know that many of the issues raised in the paper are pertinent on this side of the pond. It has long been known that women offenders have specific needs; it just seems that we’ve still not found a way to attend to those needs as well as we could.