Circles of Support and Accountability as a Desistance Model

This guest post comes from Kathy Fox, Dept. of Sociology, University of Vermont.

I recently completed an evaluation for the State of Vermont Department of Corrections (USA) on a program called Circles of Support & Accountability (CoSA), which is a reintegration model that originated in Canada. The U.K. has Circles, as does New Zealand, and a few places in the U.S. CoSA is a model for serious offenders who pose a risk of re-offense in part because of their isolation and lack of social support. The CoSA is a group of a few volunteers—just ordinary citizens really—who commit to meeting with the released “core member” for a year, offering support and guidance as he or she transitions back to the community.

Vermont is the only place as far as I know that uses CoSA many kinds of serious offender; other places use it only for sex offenders. Also this tiny state (of just under 700,000 people) has run more Circles than any place in the U.S.—close to 100 since 2005.

The evaluation was a qualitative one, designed to understand how CoSAs work, what the nature of the relationships is, and how members experience it. I interviewed 20 core members and 57 volunteers (each CoSA has three volunteers surrounding him or her). What I found was interesting and I will briefly summarize here:

  • CoSA is effective in part because it is unpaid, nonprofessional volunteers. Many core members were moved by the fact that ordinary people would invest time in them; this created a sense of mutual obligation. Core members didn’t want to let their team down.
  • CoSA volunteers balanced support and accountability but seem to work best when it’s longer on support, especially initially. Volunteers often would help core members stay within their conditions of release by giving them rides places, helping them shop, etc. And they would encourage them in their job pursuits. In addition, they would remind them of what they liked to do, like fish or bike, and participate in those activities with them. I found that the deeper and more socially involved the team became with the core member, the greater moral authority they had when they had to call them on a risky behavior.
  • A function that CoSA served that I had not expected was to de-institutionalize people who had served long sentences. Because the core members were released on supervision conditions, most could not drive, and some had restrictions on where they could live. They were often, lonely, overwhelmed, and unsure how to live in a world that had changed. CoSA helped them take small steps, like learning how to use a cell phone, and to get a bus pass.

A major function that CoSA serves is to communicate to ex-prisoners that they belong in the community and that ordinary citizens think they are worthwhile. This generated a sense of optimism—reminding me of the necessary components for desistance that Maruna (2001) talked about and “secondary desistance” that Maruna and McNeill discuss, which is the more abiding desistance based on a newly emergent pro-social identity. It’s quite clear that prison alone and even prison with rehabilitation does not lead to a more optimistic self—that is a social process, one that a circles model can assist. Over the coming months, I (along with Dr. Robin Wilson, who evaluated CoSA in Canada and found it reduced recidivism dramatically) will be conducting a recidivism study with the 100 Vermont CoSAs. The U.S. federal government seems interested in expanding CoSA (which was funded in Vermont by the Second Chance Act, which is a federal reentry fund), recently releasing a solicitation for programs to run circles.[1] Interestingly, the newest federal solicitation out of the SMART office (Office of Sex Offender Sentencing, Monitoring, Apprehending, Tracking, Registering and Tracking) refers to CoSA as a “supervision” strategy, whereas the original model out of Canada was called an adaptation of restorative justice.

Maruna, Shadd. 2001. Making Good: How Ex-convicts Reform and Rebuild their Lives. Washington, DC: American Psychological Association.