Using Private Money for Public Good

Story appears in the 2013 March/April issue of Capitol Ideas

More than 3,000 adolescents go through Rikers Island prison in New York City each year. Half of them will reoffend and end up back in prison within a year after their release.

That’s an expensive problem. According to the Vera Institute of Justice, each inmate cost New Yorkers an average of $60,000 in 2012.

In January 2013, Rikers Island kicked off a new program called the Adolescent Behavioral Learning Experience, aimed at reducing the number of 16- to 18-year-olds who reoffend.

“What we’re doing is a form of cognitive behavioral therapy,” said Kristin Misner, chief of staff to New York City’s deputy mayor for health and human services. “There’s been tons of research done in the criminal justice arena around the use of cognitive behavioral therapy to reduce recidivism for offenders.

“It helps kids realize the choices they make have consequences and results. They can make different choices; they are not victims of circumstance. They can take control and ownership of their lives so when they leave Rikers, it makes them think more about the structured decision making that leads to positive outcomes for them.”

New York City is paying for this $9.6 million, four-year program by becoming the first entity in the country to enter into a social impact bond. It’s a new financial tool that may be able to make an impact of some of the country’s most intractable problems.

What It Is

Proven, effective programs that save money for states or local governments in the long run can be hard to finance, said Kristina Costa, research assistant and leader of the social impact bond work at the Center for American Progress. Sometimes the people a program targets are politically unpopular, Costa said, and sometimes it takes time to see the program’s effects.

Social impact bonds are a way to bring new money into old problems. Costa said governments select a population and set goals—such as creating a 10 percent reduction in the number of homeless people sleeping on the street six months or more—and decide what they’re willing to pay for the results.

An outside group, called an intermediary, sets up the contract language, finds investors willing to finance the project and manages the service providers. The government repays the money only if the goal is met, as judged by an independent evaluator. If the program works better than anticipated, the funder earns a small profit on its investment.

“Foundations were really an early supporter of this,” Costa said. “It’s a recognition they don’t have enough money to solve all the problems. This is a way to bring private capital into the public good.”

Goldman Sachs is funding the Rikers Island project. If the project can reduce adolescent recidivism by 10 percent, which would allow the prison to close a wing, Goldman Sachs will earn its money back. If the project reduces recidivism more, the bank will earn a profit, topping out at slightly more than $2 million for a 20 percent or more reduction.

“We invest the firm’s capital in what we think of as double bottom line investments, those that have both social and financial returns,” said Andrea Phillips, vice president of the Urban Investment Group at Goldman Sachs.

Because New York’s bond structure will not pay a return unless recidivism is reduced a minimum of 10 percent, Bloomberg Philanthropies provided a $7.2 million grant to the intermediary that will limit Goldman Sachs losses should the project fail.

Phillips said it was important to the bank to have that reassurance, or credit enhancement, since this deal is the first of its kind in the United States.

The first social impact bond is in year three of a six-year program at a prison in Peterborough, England.  Intensive support is being provided to 3,000 short-term prisoners--while in prison and after release—in an effort to reduce recidivism rates. The investors will be paid only if the recidivism rate drops by at least 7.5 percent relative to similar prisons.

“I hope the need over time for credit enhancement will decrease as use of these financial instruments becomes more commonplace,” Phillips said. “We wouldn’t have done it if we didn’t think it would work. Our real hope, in part spurred by our investment, is that other folks become interested in making similar types of investments that can really open up a new pool of capital for social interventions.”

It’s Not for Everyone

Misner and Costa stressed that while social impact bonds may be good financing tools for some social ills, they won’t work in every situation. Costa said the interventions need to actually save money—not just prevent bigger expenditures in the future—over a short period of time. While preschool education may have many benefits, especially for at-risk youth, she noted, those benefits are spread out over the child’s lifetime.

“I will say it is a tool in the toolbox,” Misner said. “If there are programs we know 100 percent are going to work, this is not the right tool to use. The government will fund that themselves.

“There has to a sweet spot where there is sufficient evidence an intervention is likely to work, but not enough confidence on either the budget or city side to spend new dollars when the city or state is already spending money on treatment.”

Costa said social impact bonds do have potential pitfalls. The Center for American Progress recommends they not be used for any essential services.

“The biggest risk actually is say you have a four or five year long social impact deal,” Cost said. “Say in year three, they (the investors) know they’re just not going to succeed, not going to reach the outcome threshold. … They are heavily incentivized to walk away from the deal. That’s the risk for government and there needs to be some kind of orderly wind down agreement in place upfront, so they don’t wind up leaving people hanging.”

Misner said the Rikers Island project in New York City avoided that potential pitfall in two ways. The project itself, she said, is a new service to adolescents at the facility. If the program fails and is discontinued, the youth aren’t in a worse position than they were before it began. The second protection is in the design of the contracts.

“We said they (Goldman Sachs) actually can’t pull out until the end of year three,” she said. “We’ll look at one-year (recidivism) rates after year three, then there will be an option to fund year four. … We felt really strongly that you can’t put all this work and effort in and pull out after three months or six months.”

Wave of the Future?

David Merriman, acting deputy chief of staff for health and human services in Cuyahoga County, Ohio, said the county is exploring whether a social impact bond—also called pay for performance—could work for it. Merriman said the two most obvious choices for services right now are foster care and adolescent mental health.

Although the bonds are an exciting idea, county officials are proceeding cautiously, he said.

“We are very concerned about any … proposed intervention or way to serve the community differently. That intervention has to be at least as good as what is in the system and probably a whole lot better,” Merriman said. “We’re not talking about walking away from core services to pay for success financing. We’re not talking about walking away from any traditional role as a county government.

“What we tend to speak about is prevention: What are the interventions out there that prevent the client from falling into that safety net? Frequently, those interventions don’t exist in the community or if they do exist, they exist in such a small way that you never get ahead of the problem.”

George Overholser, CEO of the intermediary firm Third Sector Capital Partners, said social impact bonds will not be constrained by money in the marketplace, but rather by interventions that are scalable and proven. His company is working with Massachusetts on a bond that targets reducing the number of nights in prison for young people aging out of the juvenile corrections and probation systems.

“The reason people like it is because it combines an evidence-based way of reaching social goals with an economical way of saving taxpayers money at the same time,” Overholser said. “For a long time, there used to be a tradeoff. This way, we can be both progressive and frugal at the same time and that’s attractive to people.”

View Capitol Ideas Archive