Paying for success: On the cutting edge in early childhood education
When Salt Lake City’s Granite School District wanted to expand its successful early childhood education program it dialed up one of the world’s largest investment banks for a loan.
Goldman Sachs said yes, launching the first “social impact bond” aimed at early childhood education. The program will be administered by the United Way of Salt Lake, and the state will repay investors based on carefully measured costs of avoided special education expenses down the road.
If the results aren’t achieved, Goldman Sachs gets nothing back.
Social impact bonds, or as they are often called, “pay for success loans,” are a new concept of blending public purposes with private resources and approaches.
The idea is to set out a measurable and achievable social objective. Private investors fund the project, while an experienced nonprofit organization implements it. A third party measures the results, and if results meet or surpass expectations investors get paid.
The Salt Lake bond is the world’s first focused on early childhood education. The two previous bonds — in England and New York City — both focused on reducing prisoner re-offending.
While some are skeptical about whether injecting complicated third-party deals into government social policy can save money, the idea’s supporters see it a way to test and develop ideas that otherwise get lost or mishandled in a bureaucratic morass.
“Social impact bonds completely disrupt the normal pattern of funding,” said John Hoffmire, director of the Center on Business and Poverty at the University of Wisconsin and director of the Impact Bond Fund at Saïd School of Business at Oxford University.
“They gives us an opportunity to change the role of government. We are no longer taxing people to pay for social programs,” Hoffmire said. “Instead, investors fund social programs, and if they continue to work they will continue to fund them.”
Building on success
In 2012, when the Rikers Island project was launched in New York, it was the first social impact bond project in the U.S. Last summer, the Salt Lake early childhood project became the second. Today, Hoffmire said, there are 18 SIBs functioning worldwide, and at least 100 more being considered.
The Salt Lake project builds on some solid data. From 2006 to 2009, Granite School District tracked its preschool program. It first proved that 33 percent of low-income students would need special education without the early education program. Then it established that with the program, only 5 percent ended up needing special education.
The Granite school program had a waiting list of 400 kids, and it needed to expand the program but didn’t have the money. If the program were fully implemented, school officials estimated the state would save $2,607 per child for 12 years.
But with state and local budgets stretched tight, the resources to fund the expanded program weren't there. In this case, even though the idea was proven to save money, the state remained skittish about funding expansion.
Goldman Sachs was impressed with the success of the current work in the Granite School District, said Andrea Phillips, vice president in the Goldman Sachs Urban Investment Group.
The Granite program uses a standardized pretest and then tracks kids from kindergarten through fifth grade: "They have had incredible results keeping these kids on track,” Phillips said.
The evidence was strong enough for Goldman to take the risk, Phillips said, helping the district scale its program and prove its results.
The big problems
“The big problems, like education and poverty, exist on a scale that is vastly larger than philanthropy. If we are going to do anything about these issues we are going to have to change how government addresses them," said George Overholser, a venture capitalist-turned philanthropist who is now the CEO of Third Sector Capital Partners, a firm focused on social impact investing.
The problem, Overholser argues, is that too often the way government procures social services “freezes the system and locks out innovations.”
In addition, he argues, the nature of social change means that programs that exist in one place and time, even if they were replicated perfectly, would often not continue to work the next place they were tried.
“Pay for Success is based on the simple premise that governments should pay for demonstrated results rather than for unverifiable promises,” agreed Harvard public policy professor Jeffrey Liebman at a U.S. Senate hearing on May 1.
“By focusing attention on achieving outcomes and evaluating impacts PFS has the potential to produce better results at a lower cost to taxpayers," Liebman said.
But others are not so sure. At the same hearing Maryland House Delegate Mark Fisher (R), argued that SIBs are “well-intended, but they unnecessarily bloat bureaucracies. Moreover, they have the potential of leading to crony capitalism, and as the Maryland Department of Legislative Services concluded, they do not save money.”
He was followed by Kyle McKay, an analyst for the Texas Legislative Budget Board who studied SIBs for his legislature before recommending against them. He told the Senate task force that the bonds were “expensive and risky” and that they “may also distract governments from a more comprehensive, sustainable approach to improving public policy.
“There is,” McKay concluded, “no evidence to suggest that simply throwing investors into the fray will resolve the ongoing limitations and problems. Instead, they may very well exacerbate the challenges.”
Success in England
Notwithstanding such doubts, social impact bonds loans have come a long way in a few short years, Hoffmire says.
The world’s first SIB was launched in 2010 at Peterborough prison in central England. It aimed to provide a social support net to short-term prisoners between 16 and 18 years old.
The Peterborough project helped released offenders with housing, getting IDs, getting government benefits, mental or substance abuse issues, family relations and more. Investors at Peterborough will be paid if the project reduces recidivism by 7.5 percent on a sliding scale that caps out at 13 percent.
Through the first three years, the project was quite successful. Short-term prisoners convicted of new crimes after release dropped 11 percent at Peterborough, while they climbed nationally by 12 percent.
The results were so persuasive that earlier this month the UK government announced that the experiment would be dropped going forward, and instead a similar program will be implemented nationally.
Some saw this reversal as a negative for the SIB concept, but supporters saw it as confirmation that the effort was working. It was, Hoffmire said, as if a controlled medical study was abandoned part way through because the results were so strong it became unethical to continue denying the treatment to the control group.
When the first U.S. “pay-for-success” program was launched in 2012 at Rikers Island, Goldman Sachs led a small group of investors to fund a proven “cognitive behavioral therapy” effort for youth offenders, helping them develop emotional and mental tools for dealing with frustrations and challenges in home and on the street.
Half of the kids who currently cycle through the jail return within a year of their release. Based on smaller scale experiments, believes they can cut recidivism by 20 to 30 percent using the new tools.
The New York project involves several layers of oversight. The funds are provided by Goldman Sachs, partnering with Bloomberg Philanthropies. The project is designed by MRDC, a respected education policy nonprofit, and is implemented on the ground by two other nonprofits. The Vera Institute of Justice will then evaluate the results to see if the investors will get paid.
The test will be quasi-experimental, using a baseline from previous years and then carefully controlling for any differences in individuals or the system as a whole that could skew outcomes.