Study Review

Domestic violence perpetrator programmes

Domestic abuse perpetrator programmes in the UK intended to reduce repeated episodes of violent behaviour.

Outcome Overall
effectiveness

This rating shows how effective the intervention is at achieving the evaluated outcome.

Click here for information about how effectiveness ratings are applied.

Strength of
evidence

This rating shows how confident we can be about a finding, based on how the research was designed and carried out.

Repeated episodes of violent behaviour Overall effectiveness: 1 (maximum 2) Strength of evidence : 0 (maximum 3)

Headline points

  • Despite widespread use of domestic violence perpetrator programmes there is limited evidence of their effectiveness.
  • While there are promising findings regarding the reduction of repeated violence in relation to the Duluth programme in the UK, differences between programmes and how they are implemented makes it difficult to draw firm conclusions.
  • Limitations regarding the methodological quality of domestic violence perpetrator programme evaluations mean that no definitive conclusions can be made regarding their effectiveness.
  • More robust evaluation studies are needed to identify which elements of the programmes are most effective and how they should be implemented.

Useful contacts

What is this?

  • A high proportion of families with child protection concerns experience domestic violence. The Children’s Commissioner for England (2019) Childhood Vulnerability report showed that in 2017, 831,000 children lived in homes that reported domestic abuse.
  • Domestic violence perpetrator programmes are aimed at reducing the incidence of domestic violence by changing the attitudes, behaviours and beliefs of perpetrators.
  • This summary focuses in particular on the four UK studies. All four studies adopted the Duluth model, with three of the four studies incorporating elements of two other approaches – two studies included cognitive-behavioural therapy, and one study used a pro-feminist psychodynamic model.

How is it meant to work?

Domestic violence perpetrator programmes are based on three prevailing approaches aiming to change the attitudes, behaviours and beliefs of perpetrators.

  • Cognitive-behavioural programmes where violence is attributed to learned behaviours aim to change patterns of thought to change behaviour.
  • Psychodynamic approaches emphasise personality and the emotional disposition of perpetrators, and aim to educate perpetrators so they are able to recognise the feelings that precipitate violence.
  • Pro-feminist approaches view violence as originating from patriarchal values about women’s roles.

All of the four studies from the UK were also based around The Duluth model, developed in the city of Duluth, USA, which has been used widely in the UK. It is underpinned by pro-feminist and cognitive-behavioural approaches.  The key components of the Duluth model include:

  • An emphasis on the perpetrator being accountable for their actions.
  • Domestic abuse being situated in a wider societal context, and the role of patriarchy.
  • A broad view of the role of power and control within domestic abuse.

What are the evaluated outcomes?

  • Repeated episodes of violent behaviour

How effective is it?

Repeated episodes of violent behaviour
Overall, domestic abuse perpetrator programmes tended to have a positive effect on repeated episodes of violent behaviour. However, this is based on very low strength evidence.

  • The summary focused on four studies from the UK.
  • The sample size across the UK studies varied between 115 and 262 participants, and dropout rates ranged from 32% to 63%.
  • The review acknowledges it is difficult to gain valid data on rates and incidence of domestic abuse.
  • Only one study employed a control group. However group allocation was based on court sentence, so it is possible that only less severe offenders may have been mandated by the court to complete the programme.

Where has it been studied?

The original review included 12 studies, of which four were in the UK, four were from Spain, and one from each of Cyprus, Finland, Germany and Sweden. This summary focuses on the four UK studies as the applicability of evidence from other European countries to a UK context may be limited.

Who does it work for?

Of the four UK studies, three of the programmes were court mandated. None of the programmes tailored the domestic violence perpetrator programme to the specific characteristics or learning styles of the participants. Therefore, it was not possible to identify what type of perpetrator the programmes worked or did not work for. Further, the attrition rate was relatively high. The summary suggests that participants who voluntarily referred themselves to the intervention may be more likely than their court-mandated counterparts to be motivated to change.

When, where and how does it work?

The review doesn’t include significant detail around when, where and how programmes work. One study noted that re-offending was not associated with risk, offender type, the therapeutic environment, program attendance, or variations in program implementation, but was significantly associated with a criminal history prior to treatment.

What are the costs and benefits?

No economic analysis was included in the study.

How is it implemented?

There is significant variation in how perpetrator programmes are delivered across the UK and globally, but the following are drawn from the four UK studies included in the review.

As noted previously there was a variety of models used in the implementation of perpetrator programmes:

  • All four studies used the Duluth model; Two combined this with cognitive-behavioural approaches, and another combined this with a pro feminist approach.
  • Three of the studies used a 24 session model. Session duration varied, with one study noted as 24 150-minute group sessions, and five 150-minute follow-up sessions, and another offering 24 two hour sessions over six months. Both group and individual sessions were used. The fourth study involved 10 individual CBT sessions and 30 loosely defined group sessions, over 42 weeks.
  • Participants were court-mandated to attend in three of the studies, and the fourth used voluntary participation.

Who can deliver it?

  • The review doesn’t provide significant detail on who delivered the programmes, but one study is noted as involving meetings with the police domestic abuse unit, social workers, and women’s and children’s workers.

What are the training and supervision requirements?

  • The review doesn’t provide detail on training and supervision requirements.

What supports good implementation?

  • As noted previously the issue of measuring implementation is difficult given the variety in perpetrator programmes, and how they are delivered, which means that firm conclusions aren’t able to be drawn from the four studies reviewed.

Case study

The study by Dobash et al. (1999) compared the effectiveness of two court-mandated perpetrator programs with traditional criminal justice-based sanctions (e.g., fines, probation, and prison) in Scotland.

  • There were two group programmes, of 24 and 27 weeks.
  • The programme was court-mandated and combined the Duluth model with cognitive-behavioural therapy.
  • There were 51 men and 47 women who were their partners in the treatment group, and 71 men and 87 women in the comparison group. In all there was 47% male and 40% female dropout in treatment group, and 51% male and 42% female dropout in comparison group.

The authors administered interviews to the participants at the beginning of the program (Time 1) and sent postal questionnaires at 3 and 12 months thereafter (Times 2 and 3). There were few differences between the two groups on key demographic, criminal, and attitudinal variables at Time 1, although there were significant differences pertaining to employment and marital status.

During the follow-up period, marginally more men in the treatment condition appeared in arrest and prosecution records than men in the control condition. Women’s reports of subsequent violence based on questionnaire data revealed that 33% of the men in the intervention condition and 69% in the control condition used violence at Time 3. This difference was statistically significant.

A similar difference was obtained when the authors compared the use of frequent violence between the two groups. The authors also observed reductions in the intervention condition of controlling and intimidating behaviours, both over time and compared with the control condition. Moreover, women partners of men in the intervention condition reported more positive and statistically significant improvements in quality-of-life measures such as feelings of happiness, contentment, and safety than women partners of men in the control condition.

In summary...

  • Despite widespread use of domestic violence perpetrator programmes there is limited evidence of their effectiveness.
  • Limitations regarding the methodological quality of domestic violence perpetrator programme evaluations mean that no definitive conclusions can be made regarding their effectiveness.
  • While there are promising findings regarding the reduction of repeated violence in relation to the Duluth programme in the UK, differences between programmes and how they are implemented makes it difficult to draw firm conclusions.
  • Domestic violence perpetrator programmes included in this review were not tailored to participant characteristics, including whether participants have volunteered to attend or have been court mandated to do so, which may affect how effective they are.
  • More robust evaluation studies are needed to identify which elements of the programmes are most effective and how they should be implemented.

Further resources