The tendency for former criminals to end up back in prison generates over $50 billion every year in corrections costs nationally. After Medicaid, it is the second fastest growing budget item in the US. Three years ago, Goldman Sachs, New York City, and then-mayor Michael Bloomberg’s foundation aimed to do something about this, and inked a $9.6 million deal to reduce the recidivism rate of youth offenders at Rikers Island Prison using cognitive behavioral therapy.
The transaction, known as a Social Impact Bond (SIB), was structured with no upfront cost to the city and let investors (Goldman) and philanthropists (Bloomberg) assume the upfront risk for the social programs provided to current and former inmates, while the government only had to pay back the investors for the programs that actually worked.
The Rikers Island Prison SIB is one example of fast-emerging interest and activity around these kinds of strategies, which are also known as pay-for-success financings. SIBs create packages for achieving social progress where government only pays when it saves money; the investor can receive higher returns for higher impact, and the provider of the service can grow.
Given the ability for SIBs to save money and deliver better social outcomes, they appeal to both fiscal conservatives and social progressives, and over $40 million has been mobilized to date in the US. In the three years since the Rikers Island SIB was initiated, four other SIBs addressing early childhood education, homelessness, and prison recidivism in the US have been implemented.
But on July 2, the independent evaluator of the Rikers Island SIB announced that the program had failed to reduce recidivism among the participants by more than the 10% minimum that would have required the city to make payment to the investors. Put another way, any change in the recidivism rate of the program participants compared to a control group was determined to be statistically insignificant.
Although the approach had been used with success on older youths, the specific program at Rikers had not been tested and was being implemented in the challenging setting of a prison. New York City will terminate the program at the end of August, and Goldman will receive $6 million of the $7.2 million it had so far lent to fund the program, due to a $6 million loan guarantee by Bloomberg.
The results of the Rikers Island SIB and the launch of the other transactions raise a host of questions about whether or not these structures can actually transform public finance and bring more capital to social services.
Uncertainty around the SIB value proposition is what led the Rockefeller Foundation to have us to speak with more than 90 investors and other stakeholders in 2012 to understand the SIB investment landscape, and determine the structures that were likely to attract incremental capital to this new market. The resulting report, Building A Healthy and Sustainable Social Impact Bond Market: The Investor Landscape, (pdf) served to unpack the opportunity for SIB investors and to highlight their potential concerns, preferences, and insights to further the creation of a healthy and sustainable SIB market.
The case for SIBs is strong. For one thing, prevention is harder to fund than downstream problems. Government is great at running an ambulance service at the bottom of a cliff for those who fall, but it does not often take the steps needed to prevent people from falling in the first place. One reason is that prevention has no clear constituency to lobby for budget—consider the prospect of prison operators and unions lobbying legislators versus organizing people who have not been victims of crime advocating for more effective prison release programs.
What’s more, our current system for funding social programs is not tied to outcomes. Because legislators fund (or cut) social programs based on legal mandates, pressure from taxpayers, or simple political expedience, activities are funded—not outcomes. Service providers are paid for inputs rather than for producing meaningful outcomes—e.g. turning around the lives of juveniles, or preparing children for success in school. It is easier to monitor how many juveniles are institutionalized and pay a per diem than to consider what is needed to keep a troubled youth with his family and community—even though institutionalization is a bad outcome for the youth and taxpayers.
Lastly, public finance investors get repaid based simply on taxing authority, not on program effectiveness. One goal of SIBs is to make investors take on the risk of achieving needed social outcomes. By building the capacity of investors to understand the performance risk of social services, it is expected that more capital will flow into the most effective interventions, a point we explore further in a new report released by the Money Management Institute entitled Bringing Impact Investing Down to Earth: Insights for Making Sense, Managing Outcomes, and Meeting Client Demand.
While the Rikers Island program attempted an innovative approach to reduce recidivism, other SIB candidates such as Nurse Family Partnership (NFP)—a visiting nurse program for low-income, first-time mothers—are proven with evidence based on randomized controlled trials. NFP would use SIBs to raise capital to scale its proven intervention and reach a larger number of new mothers and their children.
The bulk of SIBs have been in criminal justice, juvenile detention, or sheltering the homeless. These sectors use high cost strategies of institutionalizing people who would be more effectively served in de-institutionalized settings. Most people (and even elected officials) can see the benefit of spending less on prisons, shelters, and dysfunctional juvenile detention centers. But what about areas where more spending is needed, such as early childhood education or job training or mental health?
In most cases, government entities are responsible for paying if the desired outcomes are generated. Even if investors accept the counter-party risk of the government, the ability of governments to make these commitments is subject to budget constraints and requires a complete re-engineering of procurement processes.
However, there are promising SIB opportunities that do not rely on public payors, such as workforce development and job readiness programs in which private sector employers agree to pay for the program if it delivers qualified employees. In the health sector, hospital systems and insurance companies that are now responsible for managing the overall health outcomes of communities can also structure innovative contingent payment transactions.
SIBs transfer the performance risk of the service providers from the public sector to private investors in exchange for part of the upside. But just because a risk is shifted, it does not go away. Remember the mono-line insurance companies that created those AAA bonds that defaulted? Sometimes what appears to be a clear transfer of risk is not. In other words, could government really step away from a failed SIB and leave the people and communities without services, or will it step back in and spend public funds?
Even though Goldman lost only $1.2 million in the Rikers Island SIB, this result might cool investor appetite to provide catalytic capital in large amounts to these structures, and may increase the call for philanthropic support to cushion the risks to investors. Investors may also favor more proven programs to the detriment of innovation.
With enabling legislation being passed around the country, and federal grants arriving to cover development costs, more and more SIBs are coming down the pike despite the Rikers Island results. SIBs have created great value simply by bringing together many unlikely parties to tackle some thorny social issues, but the jury is still out on their long-term growth and impact.