How did Connecticut expand Family-Based Recovery using a social impact bond?

Rather than removing children from their families and providing services to their parents separately, Connecticut’s Department of Children and Families (DCF) has been institutionalizing the agency’s commitment to keeping families together when safe and possible to do so.

As part of this effort, DCF:

  • Adopted a family strengthening practice model.
  • Implemented a teaming continuum that allows staff to “hit pause” when it is likely that a child will enter care so that staff can work with families and their support systems to try to mitigate the need for an out-of-home placement.
  • Established clear performance expectations around efforts and commitments to keeping children safely at home.
  • Developed and implemented the Family-Based Recovery (FBR) program,1 which aims to prevent entry into care or support reunification by decreasing parental substance use and other risk factors through treatment and parenting support provided in the family’s home.

Why did Connecticut explore a social impact bond?

Pay for Success (PFS) is considered a “social impact bond,” which is a contracting approach that ties payment for service delivery to the achievement of pre-identified and measurable outcomes. PFS typically is used to either pilot innovative approaches to service delivery or to scale up a program or model that has been proven effective. PFS generally includes a rigorous evaluation and multiple stakeholder participation that includes public, private, and nonprofit partners, with private partners investing the initial capital needed for implementation.

Given the success of the FBR program in Connecticut, DCF decided to expand it through the PFS model. The new program, called the Connecticut Family Stability-Pay for Success Project, allows DCF not only to scale up the FBR program to serve more families (about 500 additional families over the next 4.5 years2), but also broaden some of FBR’s initial eligibility criteria. The broadened criteria includes: expanding services initially created for families with children birth to 3 years old to also include families with children ages 3 to 6; and extending the period of use of substances from within the past 30 days to the past 45 days.

What strategies did DCF employ to develop and implement PFS?

DCF was one of the first child protection agencies in the country to advance a social impact bond for a child welfare intervention. DCF issued a Request for Information as a precursor to developing the PFS model to generate feedback regarding how PFS might be a good fit for the state’s programming and services. The agency received substantial feedback from providers, intermediaries, and funders. The agency also conducted a feasibility study through Harvard’s Kennedy School of Government’s Government Performance Lab to determine which programs or models would be a good fit for PFS, and to determine the state’s level of readiness to support a PFS expansion.3

DCF also contracted with an intermediary organization, Social Finance, which has helped launch several PFS projects successfully. Social Finance subsequently contracted with an independent evaluator and a model developer. Social Finance also raised the capital needed to initiate the PFS program. Then, the model developer (Yale Child Study Center) established relationships and initiated subcontracts with individual providers to deliver FBR services across the state.

What are some of the lessons learned?4

Early lessons learned from implementing the PFS model include:

  • Take time to create an effective fiscal model. While there was initial enthusiasm about implementing a PFS model (many partner agencies were excited about the prospect of receiving additional funding), most stakeholders were unaware of the heavy lift required to develop it. Staff and partners had to learn how to effectively monetize outcomes and design contracts around the PFS model’s rigorous algorithms for assessing success and determining related payments. DCF also focused intensively on clarifying and confirming enrollment parameters in order to ensure that they were referring all eligible families into the program, and they were engaging them in treatment in a way that allowed for the PFS approach to be most successful. In addition, during early implementation, the state continually reiterated to its project partners and other stakeholders that, on average, it took two years to evolve a program from service launch to full capacity.
  • Engage private funders in supporting a new paradigm. Some funders did not initially understand the importance of keeping children with their parents while the parents received substance use treatment. To them, it seemed counter-intuitive to their understanding of risk and safety. DCF staff had to explain to funders and other stakeholders not only the legal obligations of child protection agencies, but also their moral responsibilities to children and families. DCF also educated partners about the agency’s 10 demonstrable years of successful FBR outcomes, and informed them about research supporting the benefits of keeping children safely with their families. This sharing of knowledge combined to help funders and stakeholders see the social value of expanding the FBR program using PFS.
  • Choose the right program for funding and expansion. DCF had to be thoughtful about selecting the right program to champion as a good fit for PFS. Some of the initial questions the agency considered included: Where is there unmet need? Where is there enough evidence to show that certain outcomes embedded in the PFS model are achievable? Will the model developer be willing to adjust the program model to fit the requirements of the PFS approach?
  • Normalize PFS through consistent messaging and strategic communications. DCF leadership invested a significant amount of time raising awareness around PFS. Agency leaders worked with staff to both launch this new service and understand the shift in overall practice needed to implement PFS successfully. DCF found it particularly challenging to normalize the PFS approach and generate support from staff and other stakeholders in implementing randomized control trials (RCT) to test the success of the program. The change to a new referral process for FBR and other substance use treatment services resulted in initial confusion on how to access services to meet the treatment needs of families. Therefore, the individuals championing the PFS model and the evaluators had to clearly explain how RCTs work and reinforce that the families that do not receive PFS still were able to receive the same suite of services that would normally have been available to them.
  • Hire and retain employees who believe in the mission. Given the population served by FBR and its unique approach, providers focused on hiring individuals who shared their holistic philosophy in serving children and families, and who could commit to the work and the provider long-term. Providers recruited staff who could agree to the shared risk and responsibility of keeping children in their home during their parents’ substance use treatment, be comfortable going into clients’ homes to deliver services, and align themselves with the program’s approach to treating substance use as a medical condition.

1 Please see What is Connecticut’s Family-Based Recovery Program
2 Connecticut Department of Children and Families. (n.d.). Connecticut Family Stability Pay for Success Project Fact Sheet. Retrieved from
3 Harvard Kennedy School Government Performance Lab. (2017). Helping Families Combat Substance Use: Connecticut Family Stability Project. Retrieved from
4 Interview with Elizabeth Duryea, Chief of Staff, Connecticut’s Department of Children and Families, Kristina Stevens, Deputy Commissioner, Connecticut’s Department of Children and Families, and Mary Painter, Clinical Director of Substance Abuse Services, Connecticut’s Department of Children and Families, on April 9, 2018.