In early February, Massachusetts launched the largest social impact bond (SIB) in the short history of this highly touted funding innovation—a $27 million effort to keep young men who've been released from jail from going back. The allure of the SIB is that it lets government agencies pay only for programs that deliver documented results, while private investors who back such programs make a profit only if they succeed.
In the view of proponents, these factors combine to create the defining benefit of SIBs, which is, as a 2011 McKinsey & Company report put it, "to get proven solutions to scale with no risk to public budgets."
But if the SIB concept is to be judged on whether it ultimately engages sufficient private capital to fuel large-scale expansion of evidence-based social programs, then I'd bet that SIBs will eventually be judged a failure. The reason is straightforward: once a program generates promised outcomes and delivers significant cost savings, it's hard to see how governments will be able to politically justify continuing to reward private investors—especially well-known Wall Street firms—with risk premiums for something that's no longer judged to be risky.
No, it seems increasingly clear that social innovation bonds aren't primarily about directly scaling what works. Their real potential contribution is more likely to be demonstrating what works and changing how government itself works—which then could lead to dramatic scaling of effective programs.
The big Massachusetts social innovation bond illustrates both aspects of this value proposition. The service provider, a nonprofit named Roca, has a well-developed model with a good track record of reducing incarceration rates among the highest risk young men. But Roca has never implemented the model with so many men (well over 900) and with the kind of resources that this SIB provides. "This project can be viewed as a laboratory," said the state's secretary of administration and finance, Glen Shor. "We are testing and evaluating the types of interventions to prove their worth [and] quantify their impact." Here, Massachusetts is using the SIB in a very public way to validate whether a promising model can generate good outcomes for these young men and, by reducing prison costs, savings for taxpayers. Because governments face practical fiscal and political constraints to taking on these risks, the SIBs and their private investors are adding real social value at this early stage.
The second potential value of the SIB model is that it might actually be a powerful lever to improve government performance. Again, look at how Massachusetts has launched its SIB: the fact sheet for the program reads like a business plan. Success isn't described in lofty generalities but in concrete terms: a 40-percent decrease in days of incarceration. At this level of impact, the project would generate budgetary savings to the state equal to the cost of delivering services. At higher levels, it would actually produce a net savings for taxpayers—as well as profound social value in the lives of many young men. A rigorous third-party evaluation will determine whether the program is succeeding in meeting its well-defined goals.
This is where it gets really interesting for the government. If the SIB meets its goals, the private investors deserve their contractual premium—and our thanks—for the initial deal. But in terms of long-term scaling, the ongoing value of the SIB and its private investors would be minimal at best. Governments won't need the investors' money—they should have plenty of that since the savings would presumably be flowing. What governments need—and will already have received—is the readiness of the private parties to incur the risk of proving the program. As a result, I see governments being under great public pressure to take on the scaling challenge and to continue to deliver documented results against these very specific and proven goals.
In sum, a successful SIB may be one which both validates the ability of a promising and well-defined model to generate good outcomes, and spurs government to improve on its performance—by thinking hard about which program model it wants to bet on, setting concrete public targets, and evaluating whether the program meets those targets.
And if the SIB succeeds on these terms, what then? Governments, employing the same kind of rigorous management and monitoring techniques used for the SIB, can and should go on to directly fund the scaling of these proven programs provided by outstanding nonprofits like Roca. To be sure, accomplishing this will require that governments make changes to how they fund and manage programs, but I like the odds of success, especially with such extraordinary potential upside for our public agencies and the people they serve.
Perhaps government might even go on to apply what it's learning from these SIB experiments to other programs. If so, SIBs could end up transforming the delivery of social services and enabling truly profound social impact. That, not the ongoing use of private capital to keep growing proven programs, is where the most compelling value of SIBs may lie.