Therapeutic communities

A group-based approach to treat the effects of mental illness and substance abuse.
First published
Effect scale Quality of evidence
Effect Impact on crime Mechanism How it works Moderator Where it works Implementation How to do it Economic cost What it costs
Overall reduction, some rises





No information

Focus of the intervention

A therapeutic community is a participative, group-based approach to treat the effects of mental illness and substance abuse.

Therapeutic communities are used for both adults and juveniles.

Through the involvement of professionals and former drug users, they aim to promote lasting behavioural change and encourage the development of positive social identities.

Therapeutic communities can take place in custodial and non-custodial settings, and be offered as day or residential programmes.

While many communities request that participants stay for periods of between nine and 18 months, dropout rates are usually high.

This narrative is primarily based on two reviews. Review one, covering 42 studies, focused on therapeutic communities in custodial settings (for example, prisons).

Review two (covering seven studies) and Review three (covering 11 studies), focused on the use of therapeutic communities for addressing alcohol or drug dependency in civilian settings (for example, substance abuse clinics) or custodial settings. Review three only contributed to the mechanism section of this narrative.

All reviews focused on the effect of therapeutic communities on subsequent reoffending.

The primary studies across the reviews were based on evidence from the USA, Belgium, Spain and Germany.

Effect – how effective is it?

Overall, the evidence suggests that therapeutic communities have reduced crime, but there is some evidence that they have increased crime.

The evidence on the effectiveness of therapeutic communities used in this narrative comes from Review one, which is based on 42 primary studies.

A meta-analysis of outcomes from 35 of the studies, which addressed adults only, showed a statistically significant reduction in reoffending amongst participants who took part in therapeutic communities compared to control groups with no treatment.

There was no significant difference between effect sizes for studies of higher and lower methodological quality, or which were published or unpublished.
It is important to note that the two studies (from Review one) which showed statistically significant increases in reoffending after taking part in therapeutic communities were considered to be ‘poor’ by the review authors. The assessment of ‘poor’ was due to an internal appraisal (by Review one authors) based on the primary studies’ research methods.

How strong is the evidence?

Review one was sufficiently systematic that many forms of bias that could influence the study conclusions can be ruled out. 

Review one considered many elements of validity, conducted relevant statistical analyses and used quality assurance to ensure the accuracy of the information collected from primary studies. It also accounted for the potential effects of publication bias, and the possible effect of statistical outliers. However, it did not consider whether different study designs produced different effect sizes, or whether there were any potential unanticipated outcomes. One potential bias from the studies within Review one was their methodological quality – only three of the 42 studies had a quality rating of excellent, seven were rated as good, while 22 were rated as fair and 10 were rated as poor.

Review two and three did not contain enough information from the primary studies to provide a consideration of validity and effect outcome.

Mechanism – how does it work?

Taken together, the evidence suggested a number of mechanisms by which therapeutic communities might have an effect on crime.

  • All three reviews noted that the existence of a community and peer influence were key agents of change. By forming part of a community, participants learn social norms, develop effective social skills and build solidarity with others. Review one notes that this mechanism of change is partly brought about by staff (sometimes former participants) and other participants receiving treatment, reflecting an increase of personal and social responsibility.
  • Reviews one and two noted that therapeutic communities focus upon treating the individual rather than the drug use, and seeking to bring about behavioural change and a positive identity in the participant. Review two states that self-help is a fundamental principal of therapeutic communities, with the participants themselves being the main contributor to the process of change. Review one suggests that therapeutic communities help participants to internalise positive social values through role rehearsals. 
  • Review one suggests that the hierarchical nature of the community enables participants to respect authority by supporting the process of self-examination, the development of accountability towards authority and group interpersonal processes.

Moderators – in which contexts does it work best?

Review one notes that the effect of the intervention might differ by type of therapeutic community and age of participant.

Different type of therapeutic community

Review one separated the studies into different types of therapeutic community. Typical approaches involve role play and role rehearsal, for example, ‘acting as if’ with direct confrontation for rule breaches and failure to progress.

Milieu therapy varies from typical approaches by emphasising the use of traditional group counselling and psychotherapy (such as psychodrama).

Individual psychotherapeutic methods that differ from the therapeutic community are also used. They are typically more permissive, less structured, democratically organised (rather than hierarchically), use fewer confrontational methods and employ professionally trained staff (rather than promoting recovering addicts or ex-offenders).

When dividing the therapeutic community into different types and excluding statistical outliers which distorted the results, 15 studies of standard therapeutic communities for adults showed a statistically significant 2% decrease in reoffending in the experimental groups compared to control groups.

The eight studies that were referred to as Milieu therapy also showed a significant 13% decrease in reoffending among those who took part in therapeutic communities compared to control groups.

The implementation section below discusses in more detail the differences between types of therapeutic community.

Age of participant

Review one also separated studies involving adults from those for young people, finding that while adult participants in therapeutic communities showed a statistically significant decrease in reoffending, the decrease in reoffending seen in juvenile participants was not significant.

Review two considered a number of possible moderating factors, including:

  • the type of substance misused by participants
  • the reasons for attendance at therapeutic communities (voluntary or court ordered)
  • the treatment setting (inpatient or outpatient)
  • the duration of stay in the community

These comparisons were tested in single studies and the authors therefore conclude that that there is insufficient evidence that one type of therapeutic community is better than another.

Implementation – what can be said about implementing this initiative?

Review one noted that a typical therapeutic community is a community-based residence with a few professional staff, but primarily recovered addicts serving as staff.

Residents are asked to spend about nine to 18 months in residence, but the dropout rate or attrition rate is quite high – usually 60 to 80% of residents leave within the first three months.

When therapeutic communities are conducted within prisons, the dropout rate averages around 50%, compared to 70 to 90% in non-custodial settings.

Therapeutic communities in prison are also much more constrained by rules and policies of the prison setting, including security requirements.

A core characteristic of most therapeutic communities is the use of work as an organising therapeutic activity. This means residents are involved in all aspects of the community’s operations, including administration, maintenance and food preparation.

Therapeutic communities are hierarchically organised and staff and resident roles are aligned in a clear chain of command.

New residents are assigned to work in teams with the lowest status, but can move up as they demonstrate increased competency and emotional growth. Residents have an incentive to earn better work positions, associated rights and privileges, and improved accommodation.

Therapeutic communities increasingly provide additional services, such as family treatment and educational, vocational, medical and mental health services. Increasing proportions of professionals from the mental health, medical and education fields are being used.

Review one noted that the process of behavioural change begins within therapeutic communities in prison but must continue in the community if it is to have a genuine and lasting effect, as it takes about two years on average for behavioural changes to be adopted and a positive personal-social identity acquired.

When analysing the effect of dosage of treatment in therapeutic communities, the review found that longer programmes had significantly larger decreases in reoffending than shorter programmes.

This translated to a 6% differential in reoffending success rates with an increase in time in treatment from 5 to 11 months.

Review two noted implementation details of individual programmes, all of which occurred in prison.

Two of the programmes lasted for 12 months.

One of the programmes described how prisoners attended formal therapeutic community activities five days a week for four to five hours a day, with the rest of the time spent on prison work.

All of the programmes included an invitation to join a therapeutic community once released from prison.

Economic considerations – how much might it cost?

None of the reviews had any information about costs or benefits.

General considerations

The majority of the primary studies were based in the USA, so caution should be taken when applying to other geographical contexts.


Overall, the evidence based on two meta-analyses suggests that therapeutic communities have reduced crime.

Only two studies, included in the meta-analysis of Review one, suggest that therapeutic communities are associated with increases in crime.

Therapeutic communities are usually attended by drug using offenders, and can operate in custodial settings and within the community.

While statistically significant decreases in reoffending were seen in adult participants compared to control groups, the decreases in juvenile participants were not statistically significant.

Dropout rates are often high.

The evidence suggests that longer programmes see better results than shorter programmes, as it takes time for behavioural changes to be fully adopted.


Review one

Quality of evidence
Mechanism How it works Moderator Where it works Implementation How to do it Economic cost What it costs




No information


Lipton, D. S., Pearson, F. S., Cleland, C. M., & Yee, D. (2002) 'The effects of therapeutic communities and milieu therapy on recidivism', In J. McGuire (Ed.), Offender rehabilitation and treatment: Effective programmes and policies to reduce re-offending. Etobicoke: John Wiley Canada.

Review two

Quality of evidence
Mechanism How it works Moderator Where it works Implementation How to do it Economic cost What it costs




No information


Review three

Quality of evidence
Mechanism How it works Moderator Where it works Implementation How to do it Economic cost What it costs


No information

No information

No information


Summary prepared by

This narrative was prepared by UCL Jill Dando Institute and was co-funded by the College of Policing and the Economic and Social Research Council (ESRC). ESRC grant title: 'University Consortium for Evidence-Based Crime Reduction'. Grant reference: ES/L007223/1.

Was this page useful?

Do not provide personal information such as your name or email address in the feedback form. Read our privacy policy for more information on how we use this data

What is the reason for your answer?
I couldn't find what I was looking for
The information wasn't relevant to me
The information is too complicated