Private Money, Public Impact



A creative new “pay-for-success” funding approach for social programs is gaining steam—but also detractors.

By Jane Hoback

Everybody is talking about social impact bonds—an innovative yet controversial approach to reducing homelessness, helping ex-offenders stay out of prison, giving young children the opportunity to go to preschool, training old workers for new jobs, and expanding preventive health care for low-income people.

States are striking deals with investors to deliver results—reducing prison recidivism by 8 percent within five years, for example—and save taxpayers’ money in the long run.

If programs meet their goals, investors, who put up millions of dollars, are repaid with interest by the state agency overseeing the project. If they don’t, the foundations, investment banks and others are out their money, and don’t reap a return. Financial heavy-hitters like the Rockefeller Foundation, Goldman Sachs and Bank of America Merrill Lynch are in the game.

Proponents say the financing method—with a track record of about five years—presents little risk to taxpayers and ultimately saves states money. But critics caution it is too soon to tell, arguing the new model is unproven, risky and expensive, and potentially excludes social programs not attractive to investors.

Millions in Massachusetts

The idea’s popularity has grown fast. Today, some seven states and local governments have launched projects, and Massachusetts has the biggest—a $27 million, seven-year “Juvenile Justice Pay for Success Initiative,” begun in January 2014.

The state contracted with Roca, Inc., a nonprofit whose mission is to help former offenders get back on their feet. Roca is working with young men in the Boston area who have left prison or are on probation, and its goal is to cut the number of days these men are re-incarcerated by 40 percent.

Massachusetts officials estimate that would save the state as much as $41.5 million.

Depending on it success, at about the fourth year of the program, investor Goldman Sachs will be repaid at 5 percent, Kresge and other funders at 2 percent. Roca and Third Sector Capital Partners, which oversees the project, will be paid service fees for target results. But if the results exceed goals, Roca and Goldman can get up to an additional $1 million each and Kresge and other funders up to $300,000 each. If Roca fails to deliver, the government pays nothing and investors lose their money.

The initiative won strong bipartisan support from the legislature. “The members are interested in seeing that we can make these young people’s lives better,” says Massachusetts Secretary of Administration and Finance Glen Shor. “They know the state’s risk is diminished because success means shared savings. If the initiative doesn’t prove out, the government is not on the hook.”

Shor believes the financing method “gives us a chance to try something new and innovative. The investors take the risk. And the state is either doing something impactful or we learn important lessons about what works and what doesn’t. It’s a win-win either way.”

Although pay-for-success initiatives won’t replace traditional funding for all programs, according to Shor, “I do hope that some of the methods we use in social innovation finance will bring a new level of rigor in how government interventions are successful.”

New York, New Targets

New York also launched a pay-for-success program in January 2014. Its $13.5 million initiative to reduce recidivism and train and employ former prisoners is funded primarily through Bank of America Merrill Lynch, the Robin Hood Foundation and the Rockefeller Foundation, which has agreed to guarantee up to $1.3 million of investors’ principal should the project fail to repay 100 percent after five years. In addition, the U.S. Department of Labor awarded a $12 million grant for the project.

The program’s goal is to reduce recidivism by at least 8 percent and/or increase employment by 5 percent. If it succeeds, New York could save up to $12.8 million.

 “I wanted to find creative ideas to solve problems,” says Senator Kevin Parker (D), the initiative’s sponsor. “I also wanted to focus on something that has not been working—and that’s juvenile justice issues. We’re prosecuting a lot of young people as adults and holding them in adult facilities. That’s not where we ought to be. So we want to look at some new ways. Some nonprofits have some innovative ideas, and we want to reward that innovation.”

The 100-year-old Rockefeller Foundation champions social impact bonds as part of its mission to “promote the well-being of humanity throughout the world.”

“Publicly funded attempts to help poor or vulnerable populations are often remedial, ineffective and expensive,” the foundation wrote in Building a Healthy and Sustainable Social Impact Bond Market: The Investor Landscape. “Social Impact Bonds or ‘pay for success financings’ could … [link] the government and the private sector in an unprecedented partnership to save taxpayer dollars.”

Others Don’t Buy the Hype

Others are more skeptical. Jon Pratt, executive director of the Minnesota Council of Nonprofits, says the concept “comes from a good place.” But he argues that measuring results is a complex task. “If you’re contracting for highway construction, you know the organization you pay to accomplish this task is completely responsible. Did they do it or not?” Pratt says. “But if you have a prisoner who’s leaving prison, maybe he’ll be in chemical dependency treatment, he’ll be in community housing, he’ll have job coaching, he’ll have all these other activities. Which organization caused that?”

Pratt emphasizes that funding a program through a social impact bond is not additional money, but a “displacement” from other funding. He also worries that programs with easily measured results will be more attractive to investors and will have a competitive advantage over other necessary programs.

The council’s public policy director, Susan Brown, testified at a Minnesota legislative hearing on social impact bonds and cautions they carry a “high price tag. There’s a lot of intermediary activity that increases the cost—movement of money, establishing expectations around the outcomes, then doing the various evaluations that are needed to determine whether the outcomes have been met,” she says.

Some Just Say No

Governors in two of the nation’s largest states aren’t sold on social impact bonds. New Jersey Governor Chris Christie (R) pocket vetoed the New Jersey Social Innovation Act sponsored by Assemblyman Angel Fuentes (D). It would have established a $15 million, five-year pilot program to decrease chronic emergency room visits by providing preventive health care for low-income and uninsured patients. Fuentes reintroduced the bill, which “has wide bipartisan support, it has the support of the business community and we have a lot of interest from investors and foundations. It will save potentially millions of dollars and it’s not going to waste taxpayer money,” Fuentes says.

California’s Social Impact Partnerships Pilot Program also got the thumbs down from Governor Jerry Brown (D). It would have authorized the governor to solicit partnership agreements and submit them to the Legislature for funding. Contracts would be paid through the Social Innovation Financing Trust Fund in the state treasury.

Brown said the state budget was not appropriate for pay-for-success contracts and let the bill expire. He did, however, sign a similar Assembly bill authorizing the California community corrections board to provide grants to three counties to use the pay-for-success model to reduce recidivism.

Not Sold on the Idea

Not all state legislators are on board with this idea. In testimony at U.S. Senate Budget Committee hearings, Maryland Delegate Mark Fisher (R) said that although social impact bonds were “well-intended,” they “unnecessarily bloat bureaucracies.”

He cited Maryland’s Department of Legislative Services’ analysis that social impact bonds increase budgetary pressure, don’t save costs, and could exclude new providers and programs without established records of success that draw investors.

A House bill requiring the state board of education to issue requests for proposals for social impact bonds to improve public education was withdrawn.

 In Washington, House Appropriations Committee Chair Ross Hunter (D) blocked a bill to finance the development of a pilot program to improve social services and health care with social impact bonds. He believes state governments should fund social programs by raising taxes or issuing state-backed revenue bonds because social impact investments are risky and interest rates are high.

“As a private investor, what kind of interest rate are you going to ask for? Eleven percent? Nine percent? By contrast, interest rates on revenue bonds can be as low as 4 percent. If early learning is a good idea, I can issue [government-backed revenue] bonds to pay for it,” says Hunter.

“I’m not convinced this is a good idea. No one has shown me data where the finances work.”

A Bond?

Social impact bonds are not actually bonds but rather contractual arrangements between government agencies and nongovernmental organizations—usually private nonprofits—to provide social services. The government promises to pay the organization a set amount when it achieves specific, measurable results. Investors—commercial, philanthropic or a combination—fund the project initially and are repaid with interest when the goals are met. If the goals go unmet, the government pays nothing.

Is a Social Impact Bond in Your Future?

This new way of paying for and running social programs can be a powerful tool for improving government services. But making a successful transition requires a large commitment of time. Before making the big leap, the Harvard Kennedy School Social Impact Bond Technical Assistance Lab recommends that projects contain the following four essential ingredients:

  1. Enthusiastic support from leaders | Confirm that leaders are willing to be champions for the approach and to dedicate their time and energy to making it work.
  2. The potential to grow large | Ensure the project’s scale is big enough to justify the change and receive a noticeable cost savings.
  3. Alignment with other goals | Allow for the benefits of the project to spill over and support a broader reform effort.
  4. Sufficient interest from investors | Verify early on that there are enough potential investors, both philanthropic and commercial, to give the project a go-ahead.


If moving to a social impact bond model appears likely, the assistance lab at Harvard recommends taking these steps to ensure a smooth transition.

  • Conduct a benefit-cost analysis.
  • Build a potential payment schedule.
  • Develop a financial cash flow model.
  • Engage potential investors early on in the planning work.
  • Obtain authority from the legislature to select partners in an open and transparent procurement process.
  • Decide how to choose service providers for the project.

Source: Social Impact Bond Technical Assistance Lab at Harvard Kennedy School, “Social Impact Bonds: A Guide for State and Local Governments,” 2013.

First English Effort Failed

The United Kingdom established the first social impact bond in 2010 for a program aimed at reducing recidivism. A recent study showed that the first phase did not reduce the number of reconvictions sufficiently to trigger payments to investors. A second phase is scheduled for 2016.

In the United States, the Obama administration set aside $300 million to help fund pilot projects. New York City launched the country’s first social impact bonds in 2012 with a $9.6 million project to reduce recidivism among young men leaving Rikers Island prison. The project is subsidized by a guarantee from Bloomberg Philanthropies, and Goldman Sachs is the primary investor.

Several state legislatures are in various stages of developing initiatives. The Harvard Kennedy School Social Impact Bond Technical Assistance Lab, with support from the Rockefeller Foundation, conducts research on social innovation financing and provides assistance to state and local governments to develop pay for success contracts.

Jane Hoback is a Denver freelance writer and former business journalist.

Additional Resources

NCSL Resources

Other Resources