Pay for Success Programs Will be Front and Center in 2014 | Governing

BY: Jonathan Walters

Pay for success: These are the three words that you’ll hear much more often in the human services world this year as governments look for new and creative ways to fund successful programs. You’ve most likely already heard these words in the context of “social impact bonds,” which both my Governing colleague J.B. Wogan and I have written about in the last year or so.

But there are more pay for success (PFS) models than just social impact bonds. Currently I’m working on a paper for the Annie E. Casey Foundation on other forms of PFS programs. In fact, PFS in general is a concept that has been around for nearly a decade--Maryland started experimenting with a PFS model known as “opportunity compacts” back in the mid-2000s.

All PFS models have two things in common: They involve a third party intermediary who brokers a deal involving upfront investment in some preventative program or another. And they involve closely monitoring and measuring program performance to assess whether the upfront investment is actually achieving hoped-for down-the-road savings.

Where PFS models differ is in the structure of the financing. Social impact bonds involve a third party financier who funds a program hoping to reap a return on that investment later on. The only social impact bond programs that have been road tested so far are ones involving recidivism among prisoners. Goldman Sachs and Bloomberg Philanthropies, for example, areinvesting $9.6 million in a program aimed at reducing recidivism among 16- to 18-year oldscoming out of New York City’s main jail, Rikers Island.

If successful, Goldman Sachs could make a little more than $2 million on the investment. To do that, the program will need to reduce recidivism by 20 percent or more over four years (the bank’s break-even point is at around a 10 percent reduction).

Opportunity compacts, on the other hand, involve governments themselves striking a reinvestment bargain through a third party intermediary. That is, governments essentially structure a deal whereby a department of social or children and family services, for instance, commits to a specific level of performance for a certain program. That commitment includes identifying savings achieved by upfront investment and then, through a third party non-governmental organizations, a contract with their department of management and budget that guarantees savings will be plowed back into the program instead of returned to the general fund.

School is still out on whether any one of these approaches is better than another. What they both offer by way of positives is a clear requirement that governments get serious about costing out services, measuring program impacts and doing the kind of hard, complicated longitudinal analyses required to ensure that the numbers being bandied about are actually real.

The main question around both social impact bond and other PFS schemes is how widely they might be applied. Recidivism, as complicated as measuring that can be, is relatively simple compared to measuring the impacts of investing in things like early childhood education, domestic violence prevention, adult mental health interventions or job training.

And to piggyback off that question: Is there really the money and willingness out there in the private sector to invest billions -- or even mere millions -- in social programs with such uncertain payoff potential? With social impact bonds, investors stand to lose everything if the prevention program they’re backing doesn’t perform well.

The good news, though, is that there are now robust experiments in a variety of PFS efforts, which ought to at least lead to a more sophisticated approach overall to investing in human services, including -- one hopes -- a much more serious approach to doing high-quality, long-range program evaluation.