Op-Ed: What We Learned From the Nation's First Social Impact Bond

The op-ed below by James Anderson and Andrea Phillips was published by The Huffington Post on July 2, 2015

Nearly half of the youth released each year from Rikers Island, New York City's largest jail, return within 12 months. That's an unacceptably high number and one we aimed to dramatically lower when we started a robust new program for youth in the jail back in 2012. As we learned recently, the program, an evidence-based cognitive behavioral therapy that's been effective in reducing recidivism in many other correctional settings, did not work at Rikers Island. As a result, the program will be discontinued.

Still, we're encouraged that the innovative public-private partnership that allowed all of us to try to help these young people in need and do something new -- called a social impact bond -- worked. That may sound counterintuitive. It's not.

To understand why social impact bonds are impactful, and why they worked here, we need to consider how they are structured and what they are supposed to do.

With a social impact bond, a private investor funds a program that, if it achieves set goals, will address a policy priority, produce positive social results and save the government money in the long term. A portion of those savings is returned to the investor in the form of profits. If the program doesn't meet itstargets, the investor loses their money with no burden on taxpayers.

In the case of Rikers, Goldman Sachs invested $7.2 million resulting in a $1.2 million loss. $6 million in grant funds from Bloomberg Philanthropies guaranteed part of the loan in the event the program failed to meet its targets.

So why label the social impact bond financing mechanism a success?

Shifts Risk Away from Taxpayers

First, New Yorkers didn't spend a penny on the program. The beauty of social impact bonds is that they allow the government to avoid paying for programs that don't work. That's in stark contrast to the status quo in government contracting today. The federal government expects to spend over $1 trillion in 2015 for human services programs to deliver everything from early childhood education, to healthcare for the elderly and special transport options for the disabled. The vast majority of these programs are funded regardless of whether they achieve their goals. Government, in most cases, simply pays for the services provided. The social impact bond financing structure turns the old model on its head. Targets are set, impact is measured and government only pays for success -- with private capital carrying all the risk.
From Good Intentions to Good Data

Second, the social impact bond was structured to maximize the ability of government, investors and most importantly, non profit service providers, to understand what was and wasn't working. Too often, government lacks the resources and ability to mine existing data for insights. Because data collection and analysis is fundamental to understanding whether a project worked -- and whether the investors get paid -- social impact bonds provide government with insights they would not have otherwise found following traditional approaches. In this case, an approach that had a strong evidence base of successfully reducing recidivism in other settings didn't work for the young people at Rikers.

New Approaches, Even When Public Budgets are Tight

Finally, the social impact bond allowed the government to take serious steps to help find the right mixture of policy prescriptions to address a pressing policy priority in a time of budget constraints. It can be terribly difficult for government to secure financing for new programs -- let alone for prevention programs, which are often the last funded and first cut. We're proud to have opened up a new pool of capital to enable governments to bring more promising social service approaches to young people in need. And we're equally energized by the increased attention to social impact bonds and "pay for success" strategies more generally.

In fact, since this first U.S. social impact bond was launched in 2012, the appetite for these strategies has grown. Goldman Sachs invested in three more social impact bonds and early indications show encouraging progress with further results expected later this summer. Numerous other financial institutions and private investors have followed suit. In Washington, the Senate and House are considering bipartisan legislation to target federal dollars to social impact bonds. Relatedly, Bloomberg Philanthropies is investing in new contracting strategies that help city governments to ensure their funds are spent on programs and services that deliver results through its new $42 million What Works Cities initiative.

While we hoped this therapeutic program would have greater impact for youth, the social impact bond proved to be a highly useful tool. Government got to try a promising approach to a thorny problem. Taxpayers spent no money and avoided paying for a program that fell short of its intended goals. We learned more about the challenges of implementing evidence-based approaches. And we can now say that this unique financing model made it all happen -- and stands ready to help public sector leaders finance more efforts to better serve people who need support and better results.


James Anderson leads the Government Innovation portfolio at Bloomberg Philanthropies in New York City.

Andrea Phillips is a Vice President in the Urban Investment Group at Goldman Sachs.