Pay for what works (US News & World Report)

By Katharine B. Stevens

A few days ago, a promising new program to reduce youth recidivism at New York’s Rikers Island Jail received the disappointing news that it will be ended. The Adolescent Behavioral Learning Experience program was launched in 2012 to keep recently-released teenagers from going back to jail through an evidence-based therapy approach that’s been highly effective in other correctional institutions. But although it’s worked elsewhere, at Rikers the program failed. Almost 50 percent of the youth released from the jail return within 12 months and the program made no dent in those numbers.

It’s not unusual that a government social program doesn’t work. In fact less than 1 percent of billions of public dollars spent annually on social services goes to programs that have evidence of actually accomplishing their goals. What’s very unusual, however, is that the ABLE program was rigorously evaluated by a third party and will be ended because of its findings. Other, more effective ways of helping young people avoid returning to Rikers will now be explored. And even more unusual is that this useful experiment didn’t cost taxpayers a penny.

That all happened because of an innovative public financing approach known as Pay For Success, or social impact bonds, which gives us a whole new way of funding social services that both saves money and, even more important, does a better job of helping people in need.

The key to the approach is that the purpose of social programs is to improve people’s lives, and, if programs are successful, save public money by reducing the need for more public spending down the line. Home visiting programs can save money by lowering costs associated with child abuse and neglect. Early education programs can save money by decreasing the need for remedial education in elementary school. Anti-crime programs can help people avoid living in prison on the public dime.

But while those are great goals, governments usually face two big problems in pulling them off. First, public spending is mostly directed at coping with expensive problems once they’ve occurred, rather than investing in solutions that keep problems from happening in the first place. So we spend much more on Medicaid than home visiting, more on special education than preschool, more on prisons than anti-crime interventions. We’re spending so much money fixing problems that we don’t have enough left over to spend on preventing them. Second, many social programs, like the Rikers Island example, sound good but actually don’t work in practice. So rather than saving public money, those programs just cost extra: First we pay for preschool, for instance, and then we pay for special education anyway because the preschool didn’t end up really helping kids.

Pay for Success provides a way to break this cycle of ineffective government spending, bringing three complementary partners together – the government, non-profit social service providers and private investors – to run needed social programs focused explicitly on producing the outcomes they’re intended to. This is how it works:

1. The government and investors work together to find a social service provider with a rigorously-documented track record of success that proves their program is worth investing in.
2. Investors pay that service provider to run a scale-up of the program, aiming to improve specific, agreed-on outcomes for a defined group of at-risk individuals.
3. After the program is implemented for several years, the results are evaluated to see if the program achieved its goals.
4. If, and only if, the program is proven successful, the government pays investors back their original investment plus “success payments” out of the taxpayer dollars saved by preventing expensive problems that would have occurred without the program.
5. If the program isn’t successful, like Rikers Island, the government pays nothing and the program is closed.

The Utah High Quality Preschool Initiative, a project launched in 2013, illustrates the model. In partnership with two private investors, Goldman Sachs and J.B. Pritzker, Utah found a great preschool program with strong data showing past success. The investors provided $6.8 million to scale the program to serve 2,600 at-risk 3- and 4-year olds over four years in Salt Lake County’s Granite School District. Over the coming years, Utah will pay the investors back for every child who doesn’t need remedial education services in elementary school, using the savings gained from lower special education expenditures. So Utah pays for effective preschool, not just any preschool, because the program ties public funding of preschool directly to its success in helping kids do well later in life.

There are only six Pay For Success projects in the U.S. so far. The Rikers Island program was launched as the first U.S. project in 2012, the Utah Preschool program was the second, and four more projects – in early education, child welfare and recidivism – are currently in operation. Dozens more projects are in the pipeline, though, and a lot of bipartisan momentum is building around this groundbreaking strategy for improving how government gets things done.

Pay for Success promotes evidence-driven government spending and brings much-needed private sector expertise to bear on making social service delivery more effective. It draws new funds to programs that work and ends programs that don’t. And through its focus on preventing problems rather than fixing them once they’ve already happened, it has exciting potential to save a lot of money for taxpayers while helping millions of Americans live healthier, happier, more successful lives.