social impact bonds – failure or success?

July 21, 2015

The impact investing world recently got the result of the first U.S. social impact bond (SIB) – and it was either good or bad, depending on how you think about the goals of this innovative financial product.

New York City structured a $7.2 million SIB back in 2012 to fund an effort at trying something new to reduce youth recidivism for prisoners at Rikers Island. Here’s how it was supposed to work: 

  • Goldman Sachs, the investor, loaned New York City $7.2 mil. 
  • The city directed these funds to a nonprofit called MDRC to implement a cognitive behavioral therapy program that had been effective elsewhere. 
  • After three years, a third party evaluator would determine if kids in the program reoffended at a lower rate than a control group of prisoners. 
  • Based on that evaluation, the city would either repay the investor its money, plus a profit, or it would owe nothing if the program had been unsuccessful in meeting its goals. 

Results of the three-year evaluation: the program did not work. There was no difference in the rates of recidivism between those that went through the program and those that did not. Goldman Sachs (and Bloomberg Philanthropies who guaranteed part of the investment) lost all $7.2 million. 

This would seem to scream “big failure!” Investors lost all their money, New York City did not discover a new effective program, and youth offenders did not manage to escape the revolving door of prison. But that’s not how the city or the investors feel. 

“Even though we didn’t get the result with the program that we all wanted and hoped for, we now know that definitively, thanks to the social impact bond structure that we put in place,” said Jim Anderson of Bloomberg Philanthropies. 

“The social impact bond let us test the program without taxpayer expense,” said Kristin Misner-Gutierrez, director of social services for the city’s deputy mayor for health and human services. 

These sentiments get at what many believe is the real goal of the social impact bond structure – shifting public policy towards a pay-for-performance mentality. Rather than allowing funding to continue flowing to incumbent programs without knowing whether they actually have an effect, governments would do better by insisting on measurement and proof. 

Partnering with governments represents one of the biggest opportunities for scaling successful interventions and creating positive social change. But for that change to be real, and for governments and taxpayers to be willing to take on the cost, there needs to be a better system of accountability and proof. Pay-for-performance could be that system. 

The SIB structure is but one mechanism for pushing governments in that direction, but it is not the only one. In fact, The Social Impact Bond Lab at Harvard Kennedy School (one of the originators and drivers of the product) is considering a name change to better reflect its underlying goal of pay-for-performance. 

So from this perspective, the Rikers island SIB was successful: It demonstrated the benefits of experimenting with an intervention and measuring outcomes to determine whether or not to continue the program. The intervention failed, but the structure worked just as planned. 

But the failure to return anything to investors raises a question of sustainability. While perhaps someday governments will be so fluent in the use and structuring of pay-for-performance contracts that SIBs will be unnecessary, we are not there yet. Currently, governments still need the enticement of having private money pay upfront costs of launching programs and measuring them. And while investors in these “bonds” understand that they are investing more for impact than for profit, there is probably a limit to their interest in these deals if they continue to go unpaid. Even if pay for performance is the ultimate goal of SIBs, getting to that goal may still require a more traditional definition of success for the product – namely that someone gets paid. 

The number of governments looking to set up SIBs and pay-for-performance contracts right now is exploding, while interest in funding them seems to be keeping pace. The deals have evolved and become more sophisticated since the Rikers Island deal as the industry has grown. But the results to come out in the next few years will likely have a big impact on whether the excitement continues, or whether the social impact world needs to return to the drawing board.

this post brought to you by third plateau summer fellow luke gilroy.