Connecticut took a sober look Tuesday at an intoxicating idea for these tough fiscal times: Use private capital to underwrite new, innovative programs that produce a return for investors only if they solve a social problem.
In other words, pay only for success.
Goldman Sachs is underwriting a $9.6 million effort to fight recidivism by young offenders in New York City's jails. If recidivism drops by 10 percent, Goldman gets its money back. A steeper drop could make the bank a $2.1 million return.
Whether they are called "pay-for-performance" or "social impact bonds," the concept is generating buzz in the normally staid world of philanthropy, with upbeat stories by NPR, The Economist and the New York Times.
A panel of experts told a Hartford audience of social-service providers and state officials, including Gov. Dannel P. Malloy and at least two state agency heads, that they should embrace the potential, but beware the hype.
"The limitations are not to be underestimated," said Tracy Palandjian, the co-founder of Social Finance. "The economics need to be there. The presence of proven interventions and able providers need to be there."
Investors essentially are betting on non-profit providers of social services to save the state money, most often with programs that cost a little up front, yet can avoid more expensive problems down the road.
Palandjian said recidivism and health-care costs are two areas with great potential: Keeping people out of prison or out of hospital emergency rooms can produce savings that are significant and, equally important, measurable.
"If we can more actively manage chronic illness like diabetes, like asthma, the payoff is very compelling in avoided hospitalization and emergency room utilizations," Palandjian said.
Her group is the affiliate of a British group using social finance to attack recidivism -- a model for efforts in the U.S. Some backers of the approach say they were highly skeptical
"Honestly, my first reaction was this is nuts. There can't be a way to unlock private capital to get something for nothing," said Jeffrey Liebman, a public policy professor at the Kennedy School at Harvard. "My Ponzi scheme antenna went up."
Liebman now is a believer, though one who says he remains realistic about the significant challenges of making these experiments work. He is a volunteer adviser to New York and Massachusetts, where social-finance experiments are underway.
The programs must be run by experienced providers with a good business plan that identifies results that are measurable, both in terms of value to clients and savings to municipal or state governments, he said.
Liebman and others say a benefit of the approach is that it requires a market approach to social services, requiring close monitoring to determine if an approach is a success.
With most social programs, a state might pay for outputs, but not outcomes.
Rep. Diana Urban, D-Stonington, a long-time proponent of results-based accounting in state government, said an output might be how many people completed a job-training course, while an outcome would measure how many people were hired and how their incomes improved.
Palandjian, Liebman joined George Overhosler, the chief executive of Third Sector Capital Partners, to talk about the potential and pitfals of social finance. It was one of several panels at a conference organized by the Connecticut Association for Human Services and the Capitol Region Council of Governments.
"This is an important conversation to have," Malloy said.
He cautioned, however, that social-finance programs offer an opportunity to test new programs, not replace the core functions of government -- an assessment shared by the experts who spoke Tuesday.
Liebman, who worked in the Obama administration's budget office for two years before returning to Harvard, said every state government relies heavily on private non-profits to deliver services, but few governments measure the quality of those services.
"We're not reallocating more money to the superstars, because we don't know who they are," Liebman said. "And we're not stopping spending on things that don't work, because we don't know which things are working or not."
By their nature, social-finance programs require real-time monitoring and testing, he said.
"Even if up front we decide to spend money on something that doesn't work, at least we won't keep doing it for 20 years, like we do with our normal spending," he said. "We'll know."
Officials said assessing the effectiveness of state agencies and non-profits on outcomes is a goal for government, whether or not Connecticut experiments with social-finance programs.
Effective social programs save money, they said. A failure by the Department of Children and Families today could mean a future client -- and expense -- for the state's social services or its prisons, officials said.