Additional Resources:

Program Profile: Amity In-Prison Therapeutic Community

Evidence Rating: Promising - One study Promising - One study

Date: This profile was posted on June 10, 2011

Program Summary

Provides intensive treatment to male inmates with substance abuse problems during the last 9 to 12 months of their prison term. The volunteer participants must reside in a dedicated program housing unit during treatment. The program is rated Promising. Overall, participants had lower levels of reincarceration rates compared to the control groups.

This program’s rating is based on evidence that includes at least one high-quality randomized controlled trial.

Program Description

Program Goals, Target Population

The Amity In-Prison Therapeutic Community (TC) provides intensive treatment to male inmates with substance abuse problems. Participants volunteer for the program; all participants must reside in the dedicated housing unit during the last 9 to 12 months of their prison term. The Amity TC is located in a 200-man housing unit at the R.J. Donovan Correctional Facility, a medium-security prison in San Diego, Calif. Program residents are provided with a variety of treatment services to help prepare them for reentry into the community following release from prison.


Services Provided

The Amity TC treatment program uses a three-phase treatment process. The first phase (lasting 2 to 3 months) includes orientation, clinical assessment of inmates’ needs and problem areas, and planning of interventions and treatment goals. Most program residents are assigned to prison industry jobs and given limited responsibility for the maintenance of the TC.


During the second phase of treatment (lasting 5 to 6 months), residents can earn positions of increased responsibility through hard work and by showing greater involvement in the program. Residents also participate in encounter groups and counseling sessions that focus on self-discipline, self-worth, self-awareness, respect for authority, and acceptance of guidance for problem areas. During the reentry phase, which lasts 1 to 3 months, residents prepare for their return to the community by working with program and parole staff to strengthen their planning and decision-making skills.


The Amity TC program has a number of components that are not generally found in other community or prison TCs. For instance, the program uses a formal curriculum to complement informally mediated teaching sessions. The curriculum, which includes workbooks, teacher’s guides, and videotapes, focuses on topics such as “Basic Assumptions of a Teaching and Therapeutic Community” and “Understanding Violence Both Inside and Outside of Yourself.” The program also uses psychodrama groups that allow participants to reenact roles or situations that remain unresolved. Amity staff use video playback as a therapeutic tool to raise self-awareness and develop realistic self-perception. Participants can see themselves and make appropriate judgment as if they were watching the actions of another person.


Key Personnel

An integral part of the program is the presence of “lifer mentors.” Lifer mentors are highly committed, recovering substance abusers who have criminal histories. They are carefully selected, specially trained, and supervised by senior program staff. These mentors serve as role models for inmate participants because they have demonstrated great commitment and positive personal changes despite their criminal histories and substance abusing past. They support and ensure the Amity TC culture and environment by remaining available to counsel inmates 24 hours a day. They also work with inmates prior to release to help them develop a plan for life outside of prison.



Amity TC graduates are offered the opportunity to participate in a community-based, residential aftercare TC treatment program for up to 1 year. The community TC, called Vista, can accommodate up to 40 residents at a time. The aftercare residents must work to maintain the facility while they continue the program curriculum they began in prison. Vista residents are encouraged to participate in self-help groups (such as Narcotics Anonymous) and other community services as needed.


The aftercare program follows a continuity-of-care model by building on the foundation of the in-prison TC curriculum and individualizing services for each resident so participants can build on the progress they made while in the prison treatment phase. Vista also provides services for the wives and children of residents, as well as a drop-in center for Amity program graduates, which offers counseling and evening groups to encourage recovery and provide peer support.

Evaluation Outcomes

top border

Study 1


At 12 months post-release, Wexler, De Leon, Thomas, Kressel, and Peters (1999) found the no-treatment control group had significantly greater levels of reincarceration than the intent-to-treat group (49.7 percent versus 33.9 percent, respectively). The reincarceration rate was higher for the control group, and the rate decreased consistently across the four treatment study groups. Within the intent-to-treat group, inmates who completed both the prison and aftercare therapeutic community (TC) showed the lowest rate of reincarceration (8.2 percent).


This potential confounding effect of residential status is less of a factor in the 24-month follow-up period. At 24 months post-release, the no-treatment control group had significantly greater levels of recidivism than the intent-to-treat group (67.1 percent versus 42.2 percent, respectively). The aftercare TC dropouts showed an increase in reincarceration rates, but it was a non-significant reversal, and the group still maintained a success rate that was 7 percent better than the no-treatment control group. The inmates who completed both the prison and aftercare TC programs had a low reincarceration rate of 14 percent at 24 months, which was more than 12 months after completing aftercare TC treatment.


Multivariate analysis showed that 12 months after release from prison, the intent-to-treat group had a significantly lower reincarceration rate. Compared to the control group, inmates who had any level of participation in the prison TC treatment program were 48 percent less likely to be reincarcerated within 12 months after release to the community. At 24 months post-release, inmates who participated in TC treatment were 37 percent less likely to be reincarcerated following release from prison.


Days Until Reincarceration

Compared to the intent-to-treat group, the no-treatment control group had significantly fewer days until reincarceration. After 12 months at risk following release from prison, the control group had an average of 172 days until return to prison, compared to an average of 192 days for the intent-to-treat group. The greatest difference in days until reincarceration was found for the group that completed aftercare TC treatment compared to the no-treatment control group (274 mean days versus 172 mean days, respectively). Again, this result could be confounded by the fact that aftercare TC completers were in residential treatment for most of the 12 month post-release period.


As with the examination of the reincarceration rates, however, this factor is less of an issue over the 24-month period of risk. At 24 months, the aftercare TC completers had an average of 391 days until reincarceration, while the control group had an average of 216 days until reincarceration.


Overall, the reincarceration rates and days until reincarceration show a similar pattern of positive results. The full intent-to-treat group showed significantly more positive effects than the no-treatment control group. Within the intent-to-treat group, the post-release outcomes are generally related to increased involvement in treatment. The most favorable outcomes occurred for inmates who completed both the prison TC and the aftercare TC program.


Study 2


Wexler, Melnick, Lowe, and Peters (1999) found that at 36 months, the intent-to-treat group showed lower rates of reincarceration than the control group; however, the difference was not statistically significant. A significant difference does emerge when looking at the intent-to-treat subgroups. Only 27 percent of the aftercare completers were reincarcerated, in contrast to more than three fourths of the prison TC completers (with no aftercare), prison dropouts, and controls.


The study also found that when controlling for background variables (including age and the number of previous incarcerations), aftercare completers were 6.2 times less likely than the program completers and 4.3 times less likely than the program dropouts to return to prison. There was no significant difference in the likelihood of returning to prison between the prison TC dropouts and prison TC completers.


Days Until Reincarceration

Overall, the results showed that the number of days to reincarceration increased significantly across the intent-to-treat subgroups, indicating that increased amounts of treatment resulted in a greater number of days to reincarceration. It took significantly longer for the aftercare completers to recidivate than either the program dropouts or program completers. The aftercare completers took an average number of 578 days to return to prison, while program completers had an average number of 386 days and program dropouts had an average number of 305 days. By comparison, the control group had an average number of 295 days to reincarceration.


When controlling for background variables, there was a significant relationship between increasing the amount of treatment and the amount of time until reincarceration. Thus, the longer the treatment duration, the longer the inmate remained on parole prior to a first return to prison.


Study 3

Although Prendergast and colleagues (2004) found some significant differences between the groups when examining the outcomes of reincarceration using bivariate analysis, multivariate analysis did not find significant differences. Instead, other factors (age and postrelease treatment) became significant predictors of reincarnation. Also, there were no significant differences on measurements of heavy drug use and employment over the 5-year follow-up period. There were some significant differences on employment and reincarceration rates when looking at the subgroups, but no significant differences on heavy drug use.



There was a significantly smaller percentage of the treatment group that was reincarcerated compared to the control group. For the treatment group, 75.7 percent had at least one reincarceration over the 5 years, compared to 83.4 percent of the control group. The survival analysis, looking at the number of days from release to reincarceration, also showed a significant difference between groups. The treatment group spent an average of 6 months longer on the street following release than did those in the control group (809 and 634 days, respectively).


However, the Cox regression analysis found that group status (treatment group or control group) was not significantly associated with time to reincarceration. The two factors that predicted delayed incarceration were being older and the number of months attending treatment since release. The survival curves, after controlling for certain variables, for the treatment and control groups began to merge after approximately 2 years.


Heavy Drug Use

The percentage of treatment group and control group members who reported using drugs heavily during the year before the interview was not significantly different and nearly identical (24.93 percent for the treatment group and 22.55 percent for the control group).



Self-reported employment in the year before the interview was also not significantly different between the two groups (54.84 percent for the treatment group and 52.34 percent for the control group).


Subgroup Analysis

The prison program dropouts and the prison program completers had similar percentages of return to custody (87 percent and 86 percent, respectively). The VISTA dropouts had a slightly lower reincarceration rate (81 percent) but those who completed the continuum of treatment had a significantly lower reincarceration rate of 42 percent. The survival analysis found that Vista completers took 1,369 days to return to prison, compared to 618 days for the prison TC dropouts, 636 days for the prison TC completers, and 795 days for the Vista dropouts.


The results for drug use showed that participants in the four treatment subgroups reported using drugs heavily in the past year, although at relatively low rates. Over the 5-year period, 24.7 percent of the TC dropouts, 27.7 percent of the TC completers, 19.2 percent of the Vista dropouts, and 21.5 percent of the Vista completers reported heavy drug use in the past year. The differences were not significant. The differences between the groups were significant for employment. Seventy-two percent of the Vista completers, 40 percent of the prison TC dropouts, 56 percent of the TC completers, and 38 percent of the Vista dropouts reported having been employed.

bottom border

Evaluation Methodology

top border

Study 1

A prospective follow-up study by Wexler, De Leon, Thomas, Kressel, and Peters (1999) used an intent-to-treat analysis with a random selection of participants to evaluate the Amity In-Prison Therapeutic Community (TC). Volunteers were recruited from the general prison population. The California Department of Corrections (DOC) and Amity staff then reviewed inmate records to identify eligible inmates. Inmates were eligible to participate in the program if they met the admission criteria of having a drug problem and being 9 to 14 months from parole. Inmates who had been convicted of arson or sexual crimes involving minors were excluded. Eligible inmates left in the volunteer pool were randomly assigned to the treatment condition as bed space became available in the program. Inmates who were not randomly selected to participate became members of the no-treatment control group. The control group did not receive formal substance abuse treatment during their time in prison, although limited drug education and 12-step groups were available.


The study included a total sample of 715 inmates, including 425 in the intent-to-treat group and 290 in the no-treatment control group. The treatment group was 39.5 percent white, 28.4 percent Hispanic, 23 percent African American, and 9.2 percent “other,” with an average age of 31.2 years. The control group had an average age of 30.5 years, and was 35.3 percent white, 32.5 percent Hispanic, 21.6 percent African American, and 10.6 percent “other.” The only significant difference between the groups was in educational level, in which the control group had slightly more participants who had continued their education beyond the high school level.


The intent-to-treat group was categorized into four subgroups:


  • Inmates who entered the in-prison TC program but left early for disciplinary or personal reasons (prison TC dropouts)
  • Inmates who completed the in-prison TC but did not choose to participate in the aftercare TC available in the community (prison TC treatment completers)
  • Inmates who completed the in-prison TC and volunteered to attend the aftercare TC but dropped out in less than 3 months (aftercare TC treatment dropouts)
  • Inmates who completed the prison and aftercare TC programs (aftercare TC completers)

Outcomes included reincarceration, which was measured as returns to prison for either a parole violation or for a new arrest. Drug and alcohol temporary returns for less than 30 days, known as “dry outs,” were not included in the reincarceration rates. The study also included a measurement of days until first reincarceration. This provided information on the relative treatment effectiveness for inmates who were returned to prison. The study looked at the 12- and 24-month periods following release from prison when study participants were at risk for reincarceration. Follow-up was conducted on all study participants for at least 12 months following their release from prison. A 24-month follow-up period was observed for a smaller group of 263 study participants. Reincarceration data was obtained from the California DOC computerized data system. Data was analyzed using bivariate and multivariate analysis, including logistic regression and ordinary least squares regression.


It should be noted that findings could be confounded by the fact that the inmates who completed the prison and aftercare TC programs spent an average of 9 of their 12 months post-release from prison in the residential TC aftercare setting. It could be argued, therefore, that they were at less risk during the 12-month follow-up period than inmates who completed the prison TC program and were paroled directly to the community with no aftercare.


Study 2

The study by Wexler, Melnick, Lowe, and Peters (1999) is a follow-up study that built off of the previous research by Wexler and colleagues by looking at the effects of treatment 36 months after release. This study included outcome data on 478 subjects who were released from prison: 189 members of the control group and 289 members of the intent-to-treat group. The intent-to-treat group consisted of three subgroups:


  • 73 inmates who dropped out of the Amity TC (prison TC dropouts)
  • 154 inmates who completed the Amity TC but decided not to participate in aftercare, or who volunteered to participate but withdrew within the first 90 days (prison TC completers)
  • 62 inmates who completed the Amity TC and completed aftercare (aftercare completers)  

There were no significant differences between the control group and intent-to-treat groups on any of the background variables. The intent-to-treat group was on average 30.85 years old, and 31 percent African American, 28 percent white, 27 percent Hispanic, and 4 percent “other.” The control group was on average 30.48 years old, and 36 percent African American, 35 percent white, 22 percent Hispanic, and 7 percent “other.” Within the intent-to-treat group, inmates who completed aftercare were significantly more likely to be older, be white, have a history of injection drug use, report greater readiness for treatment, and have a greater number of incarcerations.


The outcome measures included reincarceration and time until reincarceration within 3 years following release from prison. Reincarceration included returns to prison for either a parole violation or for a new arrest. Return-to-prison data was collected from the DOC’s computerized Offender Based Information System. Statistical tests included chi-square, analysis of variance (or ANOVA), and logistic multiple regression analysis.

Study 3

A third study by Prendergast and colleagues (2004) attempted to locate the 715 inmates from the original sample in order to conduct a 5-year follow-up assessment. For those study participants that could be located, interviews were conducted in the community, in jails and prisons, and out of state. Of the original sample, 81 percent were interviewed for the follow-up study, 5 percent were confirmed to have died, 3 percent refused to participate, 2 percent had been deported and 10 percent could not be located. The sample for this study was 576 study participants, including 341 treatment group members and 235 control group members. There were no significant differences between the study participants who were interviewed and the study participants who could not be interviewed, except on race: the group of study participants who were interviewed had significantly fewer Caucasians than the group of study participants who were not interviewed.


The follow-up assessment focused on three major outcomes in assessing program effectiveness: reincarceration, heavy drug use, and employment. Reincarceration was defined in two ways: (1) any record of return to prison in the 5 years since release and (2) the number of days until the first incarceration within 5 years after release. Heavy drug use was defined as self-report of having used drugs (excluding alcohol and marijuana) several times per week in the 12 months before the interview. Employment was defined as self-report of having held a job (full time or part time) in the year before the 5-year interview.


To supplement the self-reported information on criminal justice system involvement, arrest records were also obtained from the California Department of Justice’s Automated Criminal History System (ACHS). Incarceration and parole records were also collected from the California DOC’s Offender Based Information System (OBIS).


Bivariate and multivariate analyses were performed on the outcome variables. The bivariate analyses included chi-square tests for categorical variables and independent t-tests (two-tailed) for continuous variables. For selected outcomes, the effect sizes were calculated as the standardized mean difference between the treatment group and the control group using the D-STAT program for meta-analytic data. Survival analysis was performed using the Kaplan-Meier method on the variable measuring “days to first reincarceration.” Reincarceration data was censored at 1,825 days (5 years). Multivariate analyses included backwards stepwise Cox regression analysis and Cox regression survival analysis.


The results were examined in two ways: intent-to-treat analysis, which compared the treatment group to the control group, and an analysis that compared the four treatment subgroups for comparison with the earlier studies on Amity TC.


Like the previous studies, the follow-up study also analyzed the outcomes by subgroups. The subgroups included:

  • Prison TC dropouts (n= 77)

  • Prison TC completers (n=159)

  • Vista dropouts (n= 26)

  • Vista completers (n= 79)

However, as with the other studies, the findings from the subgroup analyses should be viewed with caution because of the selection bias that is present in the four groups.

bottom border


top border
A cost-effectiveness analysis (CEA) examined the short-run returns (i.e., 1-year post-parole) of the Amity In-Prison Therapeutic Community (TC) and Vista aftercare program. The CEA relied on recidivism data from outcome evaluations described in the Evaluation Methodology section. The cost data for the Amity TC and Vista program was collected using the Drug Abuse Treatment Cost Analysis Program (DATCAP). According to estimates from DATCAP, the Amity TC program cost $8.57 per day per participant and the Vista aftercare program cost $16.26 per day per participant. The results from the CEA found that the intent-to-treatment group cost an average of $4,122 and had 36 percent less incarceration time (51.48 fewer incarceration days) than the average control group member, which translates into a cost of $80 per avoided incarceration day. Inmates who just completed the Amity TC (but no aftercare program) cost an average of $2,708, and generated 17 percent less incarceration time (23.9 fewer incarceration days) than the average control group member, which translates into a cost of $133 per avoided incarceration day, the highest cost savings of any other group. Finally, for inmates who completed both Amity TC and the Vista aftercare program, the average cost of treatment was $4,277, and led to 84 fewer incarceration days compared to the average control group member. According to the results, with the in-prison TC program already in place, the cost of avoiding an additional incarceration day through the aftercare component was $51. Therefore, the CEA found that for about $80 per day per offender, the California Department of Corrections could reduce reincarceration rates for substance abusing offenders. This translates to an additional investment of about $20 per day over the daily cost of housing an inmate (McCollister et al. 2003).
bottom border

Evidence-Base (Studies Reviewed)

top border
These sources were used in the development of the program profile:

Study 1
Wexler, Harry K., George De Leon, George Thomas, David Kressel, and Jean Peters. 1999. “The Amity Prison TC Evaluation: Reincarceration Outcomes.” Criminal Justice and Behavior 26(2):147–67.

Study 2
Wexler, Harry K., Gerald Melnick, Lois Lowe, and Jean Peters. 1999. “Three-Year Reincarceration Outcomes for Amity In-Prison Therapeutic Community and Aftercare in California.” The Prison Journal 79(3):321–36.

Study 3
Prendergast, Michael L., Elizabeth A. Hall, Harry K. Wexler, Gerald Melnick, and Yan Cao. 2003. “Amity Prison-Based Therapeutic Community: 5-Year Outcomes.” The Prison Journal 84(1):36–60.
bottom border

Additional References

top border
These sources were used in the development of the program profile:

Coalition for Evidence-Based Policy, Social Programs That Work. 2011. “Amity Prison Therapeutic Community.” Accessed Feb. 9, 2011.

De Leon, George, Gerald Melnick, George Thomas, David Kressel, and Harry K. Wexler. 2000. “Motivation for Treatment in a Prison-Based Therapeutic Community.” American Journal of Drug and Alcohol Abuse 26(1):33–46.

McCollister, Kathryn E., Michael T. French, Michael Prendergast, Harry Wexler, Stan Sacks, and Elizabeth Hall. 2003. “Is In-Prison Treatment Enough? A Cost-Effectiveness Treatment and Aftercare Services for Substance-Abuse Offenders.” Law & Policy 25(1):63–82.

Prendergast, Michael L., Elizabeth A. Hall, and Harry K. Wexler. 2003. “Multiple Measures of Outcome in Assessing a Prison-Based Drug Treatment Program.” In Nathaniel J. Pallone (ed.). Treating Substance Abusers in Correctional Contexts: New Understandings, New Modalities. Philadelphia, Pa.: The Haworth Press, Inc., 65–94.

Wexler, Harry K., Gerald Melnick, and Yan Cao. 2004. “Risk and Prison Substance Abuse Treatment Outcomes: A Replication and Challenge.” The Prison Journal 84(1):106–20.
bottom border

Related Practices

top border
Following are practices that are related to this program:

Incarceration-based Therapeutic Communities for Adults
This practice uses a comprehensive, residential drug treatment program model for treating substance-abusing and addicted inmates to foster changes in attitudes, perceptions, and behaviors related to substance use. The practice is rated Promising in reducing recidivism rates after release for participants in therapeutic communities.

Evidence Ratings for Outcomes:
Promising - More than one Meta-Analysis Crime & Delinquency - Multiple crime/offense types

Adult Reentry Programs
This practice involves correctional programs that focus on the transition of individuals from prison into the community. Reentry programs involve treatment or services that have been initiated while the individual is in custody and a follow-up component after the individual is released. The practice is rated Promising for reducing recidivism.

Evidence Ratings for Outcomes:
Promising - One Meta-Analysis Crime & Delinquency - Multiple crime/offense types

Rehabilitation Programs for Adult Offenders
This practice includes programs that are designed to reduce recidivism among adult offenders by improving their behaviors, skills, mental health, social functioning, and access to education and employment. Offenders may become participants in rehabilitation programs during multiple points in their involvement with the criminal justice system. This practice is rated Promising for reducing recidivism among adult offenders.

Evidence Ratings for Outcomes:
Promising - One Meta-Analysis Crime & Delinquency
bottom border

Program Snapshot

Age: 18+

Gender: Male

Race/Ethnicity: Black, Hispanic, White, Other

Geography: Suburban, Urban

Setting (Delivery): Correctional, Other Community Setting

Program Type: Alcohol and Drug Therapy/Treatment, Aftercare/Reentry, Group Therapy, Individual Therapy, Residential Treatment Center, Therapeutic Communities

Targeted Population: Alcohol and Other Drug (AOD) Offenders, Prisoners

Current Program Status: Active

Listed by Other Directories: Campbell Collaboration

Program Director:
R.J. Donovan Correctional Facility
480 Alta Road
San Diego CA 92179
Phone: 619.661.6500

Harry K. Wexler
Senior Investigator Emeritus
National Development and Research Institutes, Inc.
71 W 23rd Street, 8th Floor
New York NY 10010
Phone: 917.562.7273