Dog training and animal-assisted interventions delivered in prisons to reduce reoffending and improve social-emotional functioning.
Focus of the intervention
Prison‑based dog programmes (PBDPs) involve dog training programmes or dog‑assisted interventions delivered inside prisons. Participants care for and/or train dogs (for adoption or assistance roles) or take part in structured therapeutic sessions that incorporate dogs. These programmes aim to reduce reoffending and improve wellbeing among people in custody. They have been implemented with adults and young people, and with men and women, in various correctional settings.
What research is this summary based on?
This updated crime reduction toolkit summary is based on the findings of two meta-analytic reviews.
- The overall findings are taken from Review one, which covered 11 primary studies focusing on prison-based dog programmes (PBDPs). These include both dog training programmes (DTPs) and animal-assisted interventions (AAIs). Eight of the primary studies included in the review were conducted in the USA, and in three other countries (unspecified). The review distinguished a primary crime outcome (post‑release reoffending) and secondary non-crime outcomes relating to participant wellbeing/social-emotional functioning (for example, self-esteem, depression, anxiety and aggression).
- A second meta-analytic review (Review two) covered 12 primary studies focusing specifically on DTPs, which require offenders to train dogs for service purposes (for example, guide dogs or therapy dogs) or teach basic commands to shelter dogs to increase their chances of being adopted. This review investigated whether DTPs reduce reoffending, anti-social behaviour and drug and alcohol use, and increase self-control, self-esteem and overall mental health and wellbeing. Seven of the 12 primary studies included in the review were combined in a meta-analysis, and were conducted in the USA and Canada. Review two contributes to the Effect, Mechanism and Implementation sections.
Effect – how effective is it?
- The research suggests that prison-based dog programmes have reduced reoffending overall.
Review one
The meta-analysis in Review one (based on 93 effect sizes) showed a very small, but statistically significant effect of PBDPs compared to controls (participants receiving treatment as usual, alternative interventions or on a waiting list).
In relation to reoffending, the pooled effect was small‑to‑medium in favour of PBDPs. By contrast, there was no statistically detectable effect on social‑emotional outcomes.
Sensitivity analysis indicated that the review’s overall findings were not driven by any single study. Quasi-experimental designs tended to show stronger effects than randomised controlled trials (RCTs).
Review two
The meta-analysis in Review two found that DTPs led to a reduction in reoffending and reconviction rates in prisoners compared to those who had not participated in DTP.
The review reported a 36 percentage-point difference in reoffending rates in the treatment group compared to the control group (68% reconvicted in a control group compared with 32% reconvicted in a treated group). Of the four primary studies that specifically examined reoffending, two showed a statistically significant crime reduction effect, while the other two showed no effect.
The findings also suggest self-esteem, wellbeing and self-efficacy are increased when prisoners participate in DTPs. These outcomes are similar to those found with more traditional programmes such as education, employment and training programmes.
How robust is the review evidence?
- The quality of review methods used to assess the impact of prison-based dog programmes was very strong.
Review one used a comprehensive search strategy including unpublished literature, calculated appropriate effect sizes, was attentive to variability between studies and assessed publication bias.
However, while Review one was well conducted, the authors reported that the primary studies were generally of low methodological quality, with many quasi‑experimental designs, small samples, limited reporting on programme duration and intensity, and no implementation fidelity data. Further analysis by the review authors indicated the presence of publication bias in the sample which may have inflated the overall impact of PBDPs.
Mechanism – how does it work?
- The quality of evidence in the reviews to assess how prison-based dog programmes work was moderate.
Review two suggested several mechanisms by which DTPs in prisons might impact crime, however, none of these mechanisms were tested within the review. These include:
- inmates having to stay trouble free to be in the programme
- inmates learning patience with the dog which assists them to remain trouble-free
- participants feeling empowered to achieve goals
- increased empathy for dogs (and subsequently humans) making it easier to bond with people and increasing emotional intelligence and coping skills
Review one also set out plausible mechanisms for the wider category of PBDPs, but did not test them empirically. Proposed pathways included:
- reducing stress and providing social support through human–animal contact
- immediate behavioural feedback from dogs that helps participants practise self‑regulation
- attachment to the dog acting as a secure base that increases engagement with rehabilitation
- fostering responsibility and prosocial identity through caregiving and training
The review suggested that the lack of observed effect on social‑emotional measures means that other factors (such as identity change among prisoners or improved responsivity to other treatments) might explain the reduction in reoffending. However, these were not formally analysed.
Moderators – in which contexts does it work best?
- The quality of evidence in the reviews to assess the contexts where prison-based dog programmes might work best was moderate.
Review one examined several moderators using meta‑analytic subgroup analyses. These showed:
- larger effects for older participants in PBDPs
- stronger effects in adult justice settings than in juvenile settings
- no significant difference in effect by country setting, sex, addiction status or the intensity of dog contact (full‑time vs visiting)
- no significant difference in effect by programme type (DTP vs dog‑assisted intervention)
Implementation – what can be said about implementing this initiative?
- The quality of evidence in the reviews to assess how to implement prison-based dog programmes was low.
Review two gave some account of how DTPs could be implemented but implementation challenges were not discussed. Most programmes require offenders to be free of behaviour-related infringements for a specified period before admittance to the programme as well as continued good behaviour throughout the programme.
Programmes can vary in numbers of participants, criteria for participation, length of participation, and amount of time per day that the participants spend with the dogs. Many also include a vocational component.
Economic considerations – how much might it cost?
- The reviews provided insufficient evidence to assess how much prison-based dog programmes cost to implement.
Neither review provided cost data or formal economic analyses for prison‑based dog programmes.
Background literature in Review two highlighted that training dogs in prisons is cheaper than dog training outside of prisons ($4,000 when trained by prisoners, whereas it costs $10,000–$12,000 to train dogs outside of prison). However, the cost/benefit analysis was not specifically focused on in the main review.
General considerations
- Most primary studies were conducted in the USA. Caution is advised when applying findings to other contexts.
- Primary studies in Review one were often small and of low methodological quality.
- Outcome measurement and reporting varied, and social‑emotional outcomes did not show reliable improvements.
Summary
- Overall, the evidence suggests that prison-based dog programmes have reduced reoffending.
The meta-analysis in Review one pointed to a small overall effect of PBDPs largely driven by reduced reoffending. No reliable effects were found for social‑emotional outcomes. However, the review authors noted that publication bias may have inflated the overall effect of PBDPs.
Moderator analysis in the review showed that older participants tended to benefit more from programmes than younger participants.
Review two found that DTPs reduce reoffending overall and have desirable effects on the psychological outcomes of offenders. Moreover, training dogs inside prisons may be cheaper than training dogs outside of prisons.
Reviews
Review one
Reference
Duindam HM and others. (2020). 'Are we barking up the right tree? A meta‑analysis on the effectiveness of prison‑based dog programs'. Criminal Justice and Behavior, volume 47, issue 6, pages 749–767.
Review two
Reference
Cooke BJ and Farrington DP. (2016). The effectiveness of dog-training programs in prison: A systematic review and meta-analysis of the literature. Prison Journal, volume 96, issue 6, pages 854-876.