women-drinking-FI
Russell Webster

Russell Webster

Criminal Justice & substance misuse expert and author of this blog.

The dynamic predictors of reconviction for women

Binge drinking is a major predictor of women's reoffending.

Do women & men have different risk factors?

HMPPS has just published a new analytical summary on the dynamic predictors of reconviction for women. The research was driven by a policy question about the extent to which rehabilitative services for women in prison or on probation need to be different to those provided for men: 

Do women and men have different risk factors for criminal behaviour that will demand different interventions?

The authors, Rosie Travers and Ruth Mann, examined the nature of the most prevalent and most predictive risk factors for reoffending in a sample of nearly 15,000 women and over 95,000 men on community sentences or leaving prison for whom  Offender Assessment System (OASys) risk and need assessments were available. They were interested in whether or not there are risk factors that predict reoffending specifically for women.

Key findings

The most prevalent criminogenic needs for women were poor problem solving, impulsivity, and unemployment which were also the most prevalent needs for men.

The prevalence of other needs varied to some extent by gender, but also by risk and offence type. For example, 74% of higher-risk women with a current conviction for acquisitive crime had a Class A drug problem compared to 54% of men in the same risk band and offence type category. This difference between men and women was not evident with those at lower risk of reconviction. 

The most prevalent needs were not always the most predictive of reoffending. Five needs emerged as the strongest dynamic predictors of any reoffending for women: 

  1. Unemployment 
  2. Binge drinking 
  3. Impulsivity 
  4. Spending time with others involved in crime
  5. Class A drug use 

Most needs influenced reoffending to a similar degree for men and women. Class A drug use and binge drinking were more strongly linked to reoffending for women than for men – but were also a risk factor for men. 

The strongest dynamic predictors of violent reoffending for women were: 

  • Lack of accommodation 
  • Temper control 
  • Being the victim of domestic violence 
  • Problem drinking 
  • Lack of closeness with family 
  • Binge drinking 

There were significant gender differences in the predictiveness of binge drinking, lack of closeness with family, and poor temper control (all more influential for women). The importance of binge drinking as a major predictor of women’s reoffending is a new finding which should translate into policy action.

 

Conclusion

Overall, the researchers found more similarities than differences between women and men. While there were gender differences in prevalence, they were generally small and a question of degree, rather than signifying the presence or absence of a problem. Some of the larger gender differences in prevalence were seen for need items that did not predict reoffending, such as the prevalence of psychological problems for women while most of the stronger predictors of reoffending were the same for men and women. Gender differences were again largely a question of degree, suggesting some issues are gender-salient rather than gender-specific.

The relationship between individual criminogenic need items and reoffending varied, to an extent, according to risk and offence type as well as gender. For example, the predictiveness of a lifestyle that encourages reoffending was the same at every level of risk for both men and women. In contrast, binge drinking had a stronger relationship with violent reoffending specifically, and its impact was apparent for women even at lower levels of risk. Further research could usefully explore how needs cluster together both within and across gender, risk bands and offence characteristics. This would help to identify distinct groups for whom a package of care could then be designed to be responsive to personal characteristics, and to reflect predominant needs and risk level.

Leave a Comment

Your email address will not be published.