Obama's social impact bonds: private money, public benefit

And now for something completely different: an exotic financial instrument designed to help vulnerable Americans — not drive them into foreclosure.

Today, the White House plans to ask Congress for permission to conduct a $100 million test of "social impact bonds," a promising, experimental investment scheme out of England designed to get better results from publicly funded social services by harnessing the discipline of the private market.

Under a social impact bond arrangement, investors put up the money to run privately managed social programs. If the programs do what they're designed to do, like keep people drug-free or out of jail, the government pays the investors back, plus interest or performance bonuses. But if the programs fail to meet performance targets, taxpayers are off the hook.

It has the makings of a rare public policy trifecta — good for taxpayers, good for users of social services and good for private investors.

Joshua Sharfstein, Maryland's health secretary, says the idea is worthy pursuing, "if we can figure out ways to engage the private sector in some of these challenging problems without having to take the risk."

Washington has a poor track record of funding preventive social programs that can prove good outcomes. Nine of 10 federally funded social programs scientifically evaluated for impact since 1990 "found weak or no positive effects," policy analysts Isabel Sawhill and Jon Baron reported in Education Week last year. Among them: Job Corps teen job training and Upward Bound high school prep.

That's in part why the social bond model is attracting long looks from U.S. foundations — like Baltimore's Annie E. Casey Foundation — that believe more private-sector involvement can accelerate the development of better ways to fund and prevent stubborn social ills like homelessness and drug addiction.

Christa Velazquez, director of social investments at the Casey Foundation, is optimistic: "The hope is that this will unleash large amounts of investor capital."

Ms. Velazquez was part of a U.S. delegation that visited England in August to see the first social impact bond trial under way at Her Majesty's Prison in Peterborough. The government last year contracted with a nonprofit called Social Finance to provide vocational, housing and other services to 3,000 released prisoners over the next six years. The nonprofit is raising $7.9 million to provide the services and stands to earn for its investors as much as a 13 percent profit if the program sufficiently reduces recidivism from its current average of 60 percent.

"What the government gets in return is improved outcomes," wrote Harvard University public policy professor Jeffrey Liebman, in a paper about social impact bonds published this month by the Center for American Progress. "In some cases, improved outcomes may result in cost savings for the government that offset some or all of the expense of delivering the services."

Social Finance estimates that if its anti-recidivism program works and is scaled up across the country, the savings from reduced incarcerations could cover the government's performance-based payments.

Of course, there's no evidence yet that the Social Finance experiment in Peterborough will work — as Ms. Velazquez acknowledges. And there are other questions surrounding the idea of social impact bonds.

As Robert Embry, president of Baltimore's Abell Foundation, points out, such an unproven concept is unlikely to appeal to investors looking primarily for financial gain. "The likelihood of getting your money back is a lot less than buying a bond," he said. But charities like his might welcome even a small chance at recycling their endowments, Mr. Embry said: "it holds the opportunity of us getting money back for something we might do anyway."

The Obama administration is proposing a $100 million test of the social impact bond model in seven pilot programs, focusing on areas such as job training, juvenile justice and education. The money would come out of existing department budgets and be spent only if the programs meet performance targets.

If approved by Congress, the innovative model still faces a variety of challenges, among them the risk that investors or nonprofit service providers will abandon projects midstream if they appear unlikely to meet performance targets.

For that reason, the government should probably avoid outsourcing to the private sector "core" services such as running schools, day care centers or prisons. Professor Liebman, a former senior Obama administration official in the Office of Management and Budget, says better candidates for social impact bond funding are kindergarten-readiness programs or employment services for high school dropouts and welfare recipients.

Both Mr. Embry and Dr. Sharfstein expressed concern about the cost of collecting the kind of good performance data the bond model requires. "It may reduce the amount of money going to help people and increase the amount of money going to evaluation," Mr. Embry said.

On that front, Baltimore and Maryland may have a national advantage, said Dr. Sharfstein. So-called "data-driven government" initiatives like CitiStat and StateStat mean public agencies here are already collecting reams of program-outcome data and have become experts at analyzing the impact of government programs on people's lives.

"Maryland has not been a stranger to concepts like this," Dr. Sharfstein said. "This would be an interesting place to see some of these things in action."

Gadi Dechter, a former Sun reporter, is associate director of government reform at the Center for American Progress, a nonpartisan research and educational organization in Washington. His e-mail is gdechter@americanprogress.org.

Copyright © 2021, The Baltimore Sun, a Baltimore Sun Media Group publication | Place an Ad