One initial premise of the Pay for Success (PFS) field was that the expense and time required to develop projects would decrease over time. NFF convened its portfolio of project stakeholders funded through a grant from the Corporation for National and Community Service’s Social Innovation Fund to explore this idea. Some are working on their second initiative in a new jurisdiction, while others are working on their first project in a market with prior PFS experience.
In six years, the PFS field has come a long way. More than a dozen US projects have launched, and more than 50 others are in development. As stakeholders gain experience with PFS, they are sharing what they've learned with each other, and project partners report that there have been some efficiency gains. However, the specific characteristics and local context of each project have more influence on its trajectory than market precedent. So, while it is getting easier to launch a PFS project, it is still hard. But the care and rigor that go into developing a project specific to the needs of local communities ensures PFS achieves its ultimate goal of better serving people in need.
Here are some reflections from our partners in the field:
Project and contract development remains highly customized, despite the ability to borrow thinking and contract language from prior initiatives. Center for Employment Opportunities (CEO) and Nurse Family Partnership (NFP) are developing projects in Pennsylvania and Michigan, respectively, following the successful launch of projects elsewhere. Both organizations agreed that the idiosyncratic approaches and processes of each state in which they have worked limit their ability to apply lessons learned from one jurisdiction to another. CEO noted that while there was some information shared between those involved with its project in New York and those in Pennsylvania, “It wasn’t government-to-government sharing. Despite some efficiency in moving from one market to the next, it’s challenging and you do need to adjust your expectations for what you can accomplish.” In NFP’s case, substantial differences in state approaches to the project, including differences in the administration of Medicaid, further complicated its ability to standardize its processes.
Additionally, several governments are now developing their second or third Pay for Success projects – notably the state of Massachusetts and Salt Lake County. A project with Jewish Vocational Service and Social Finance in Massachusetts marks the third for that state. Despite the historical knowledge developed among some government employees, these project partners highlighted the need to educate new stakeholders because of transitions in government. Given the complexity of PFS project design and the unique stakeholder ecosystem of each project, supporting these partners to reorient around outcomes, an important component of PFS, remains an important part of the work done during the transaction structuring phase.
Similarly, project partners working on the two Salt Lake initiatives relating to reducing recidivism and homelessness reflected on the substantial press, both positive and critical, that surrounded the first launched transaction in the jurisdiction. “Certainly, many were familiar with Pay for Success because of the Utah High Quality Preschool project, but given how different our initiatives are we needed to continue to educate the market on the value and goals of PFS, and reenergize key stakeholders’ commitment to adapting it to new issue areas. While we did achieve some efficiency in developing two projects side-by-side, each project was unique in its successes and challenges, which did mean that synergies were harder to realize than we had hoped.”
Evaluation design is a highly specialized process. Many of the service providers acknowledged the value of a Randomized Control Trial (RCT) in building the evidence base and providing an opportunity to learn how to best deliver an intervention. They also raised concerns that the first generation of projects may have set an evaluation precedent and that the RCTs they had undergone, or were required to develop as part of project design, demanded greater resources than they anticipated, and perhaps were capable of sustaining, over the long term. Ultimately, service providers agreed that evaluations need to be designed based on factors unique to the project in development.
NFP emphasized the importance of “RCTs in building the strong evidence base for our intervention. When model innovations are being tested or the families and communities that we propose to serve are unlike those we have previously served, then we are eager to undergo new RCTs. We are optimistic that given the extent of the evidence that exists, we will not have to use RCTs in all subsequent projects, which would greatly speed the development of less expensive PFS projects that can serve more people in need.”
As Hillside Children’s Center shared, after three years of exploration, planning, and significant steps toward project execution, New York decided not to move forward on contracting for its proposed project. “New York’s conception of PFS included an RCT because the first projects to launch in the state included one. However, the data we needed for an RCT was less accessible at the state level than it is in the counties in which we currently work, which made it difficult to develop the project as envisioned. We remain committed to serving the youth who were the intended service recipients of this initiative and look forward to working with the state to do so outside of a PFS project.”
Increased project development expertise leads to more efficient and effective project structuring. In NFP’s case, during the development of its first project, it identified outcomes that are “affordable, measurable and applicable” to projects in other states, which means that it will not have to repeat the analysis for each new geography. Service providers also shared that they were more focused on establishing referral pathways for program participants earlier in the transaction structuring process, which are critical to project success. This focus can reduce the time it takes to get to project launch and helps to ensure that recipients are receiving all the services they need once it does. Second timers also noted that the intense collaboration required to structure a PFS transaction prepared them to work more effectively with both old and new partners in a new jurisdiction, and enabled them to provide the support needed to partners just entering the field to develop a project that better serves those in need. Perhaps most importantly, service providers shared that developing a PFS project encouraged them to make investments in their systems and capacity which are important to undertaking and replicating this work. As outcomes-based funding becomes more prevalent, participation in these projects has many implications beyond the PFS context.
Project partners are better equipped to lay the groundwork for continued partnership once the PFS project has ended. Service providers share the concern about what will happen to programs after the PFS project ends. Providers who have been through the process before report that they are more focused on thinking about these long-term questions and that they have more confidence in raising those topics with potential end payors up front. According to CEO, “We want to continue to serve those in need long after this transaction is finished. We are raising those questions now so that we are prepared to continue our partnership with the state in the future.”
PFS stakeholders agree that prior experience has largely had a positive impact on project development. They now understand and can plan for the resources required to develop a robust PFS project, meaning they are better able to reach those in need with effective services. Given the importance of considering local context and developing appropriate evaluation strategies, PFS projects may always require significant time investment. But partners will continue to learn by doing, shrinking the time needed to get to launch when it is possible to do so.