In Part I will examine the history and critique of social impact bonds (SIBs). In Part II, I will discuss SIBs in relation to PFS initiatives in federal policy.
For the past decade, there has been growing interest in Social Impact Bonds (SIBs) to address social ills. Private investors work with local state governments and social service providers to provide the initial capital for social projects — and are rewarded once the projects have met certain predetermined outcomes.
Proponents believe that all parties are in a win-win situation in a successful SIB. Service providers could now access the capital needed to fund their program from investors and were incentivized to adopt innovative measures. Governments would shift the investment risk to private investors, who would then be rewarded based on a pre-negotiated set of “triggers” for their return. Proponents emphasize SIBs ability to raise money upfront, to focus resources, to focus on outcomes rather than inputs, to improve incentives, and to build in flexibility. SIBs could be oriented towards preventative initiatives as opposed to ameliorative ones. To most critics, SIBs fail to address nuances and believe that SIB’s movement leaders fail to recognize the dangers of financialization in public services.
Defining and Evaluating Success
The definition of a SIB’s success can be complicated by policy changes and a lack of standard evaluation measures. A SIB’s intervention success can result in a correlation versus causation problem. In interviews conducted by Professor Williams with intermediaries, many expressed uncertainty about “policy risk” in SIB evaluations. One commented, “who should own the fact that either government could change the rules and not send people to jail as much, or that something could change in the real world, the opioids epidemic, or an increase in gun violence, or an economic drop or an economic improvement?”
The very first SIB, Peterborough, was launched in the UK in 2010 to reduce recidivism rates. Five million pounds, raised from trusts and foundations, were invested. For five years, men released from short-term jail sentences received comprehensive assistance from One Service, for a period up to 12 months post-release. The SIB reduced the recidivism rate by 9% against the national control group’s rate of 7% and the target of 7.5%. Investors were paid back in full along with a 3% per annum return. However, the project ended prematurely because the UK government had launched Transforming Rehabilitation, its own initiative for reducing recidivism rates. This launch effectively removed the ability to measure the Peterborough SIB’s results against an independent control group. Thus, while SIB proponents can deem the government’s implementation of policy changes as success, such policy changes simultaneously dilute and undermine the integrity of the SIB’s results.
[Understanding Peterborough: RAND Europe evaluated the Peterborough SIB and produced the above image in their report.]
Furthermore, throughout the SIB formation process, much time and resources are spent attempting to identify the appropriate triggers and evaluative measures to define success. In the film, Invisible Heart, director Nadine Pequeneza follows service providers in the time-consuming process of determining the “trigger” point. In an interview, Sir Ronald Cohen, a strong proponent of SIBs, remarks, “it takes eighteen months for a social impact bond to be launched. It should only take three.”
While SIBs can be grouped based on the level of financial risk exposure, there are no universal evaluation requirements. Some SIBs do not even employ a control group for evaluation, thus, raising questions to the integrity of their “success.” The Utah pre-school SIB, funded by Goldman Sachs ($4.6 million) and the J.B & M.K Pritzker Family Foundation ($2.4 million), had a threshold for full repayment of principal at 90% avoidance of special education for students at risk. In lieu of a control group to evaluate success, the SIB employed the Peabody Picture Vocabulary Test. The test determined that 99.9% of identified at-risk students no longer needed remedial education. Ultimately, according to Social Finance, Goldman Sachs and the J.B & M.K Pritzker Family Foundation made 95% of $281,550 — which was the amount of cost avoidance saved because only one child of 110 ultimately required special education in the first cohort. This astounding 99.9% success rate suggests that investors were overpaid.
SIBs also have potential for creating perverse incentives and agency costs. Professor Warner wrote, “public actors must operate in the context of norms regarding open government and the rights of citizens, but private financial investors adhere to values regarding profit and return.” The investor’s need for metrics and quick returns does not align with the complex problems that SIBs attempt to address. Professor Graham, from York University, took issue with the short evaluative time frame. He believes that SIBs are most successful when the social-project outcomes are correlated with easy-to-measure metrics. These of course, do not fully grasp the complex issues such as recidivism, homelessness, and drug abuse that SIBS attempt to resolve.
Even worse, potentially, the combination of policy and SIB structures could create perverse effects in the US. In the UK, the removal of access to public housing created a housing crisis. St. Mungos, a charity, which provided housing for the homeless through a SIB, also received funding for identifying and deporting foreign rough sleepers. In effect, St. Mungos SIB investors were “profiting from a policy crisis.”
The Track Record
According to Social Finance, out of the over 100 SIBs launched, “ten projects have reported with return on initial investor capital with return on investment. An additional eight SIBs have reported making payments based on successful social outcomes.” Social Finance also reports that there are now over 121 SIBs worldwide with over $413 million capital raised.
There is no requirement of disclosure for SIBs and only one has been reported to fail — Rikers Island SIB. The SIB attempted to reduce recidivism rates in nonviolent youths through cognitive behavioral therapy (CBT). The intervention was unable to reach the 8.5% reduction rate that would allow for investors to receive their full investment. Upon reaching its reduction rate, the project would generate funds for future endeavors. After Vera Instituted determined that the CBT intervention was unsuccessful, Goldman Sachs and other initial investors withdrew prematurely in order to receive their guarantee from Bloomberg Philanthropies.
Regardless, some still hailed the Rikers Island project as a success because the SIB had operated exactly as intended — based on the repayment structured agreed upon by all the parties. But to many, like Professor Warner and Nadine Pequenza, the high transaction costs for implementing SIBs structures do not sufficiently address complex social issues.
Professor Warner notes that, “in fact, SIBs may be more important as a new form of venture philanthropy than a new form of private equity investment.” Perhaps one should not look to SIBs as the silver bullet, but rather as tool for moving policies. One could view SIBs as useful tools for pushing policy. For example, after its initial success, Utah pre-school SIB expanded. In a discussion at Cornell between Nadine Pequeneza and Professor Warner, they noted that after this expansion, Utah began allowing public funds to be put towards providing state-education funding for preschool. If the goal of the original SIB was to emphasize the cost savings of preschool access to policymakers, then this SIB could be considered a success. J.B Pritzker, one of the original SIB investors, successfully ran for governor of Illinois this November. His five-point plan for childhood education proposes increases to public funding but does not mention SIBs.
There is no evidence that SIBs induce innovation, and investors are still inclined to back programs with a history of success. Less than ten years have elapsed since the Peterborough SIB was first implemented and there is still much room for change. The passage of SIPPRA, which will be covered in the next post, could bring about a new mix of policy risks and a new wave of SIBs.
Acknowledgements: I would like to thank Nadine Pequeneza and Professor Warner for their input. I would also like to thank Professor Milstein for his moderation during the discussion.