Pay for success not services

Pay for performance has been around for a long time in government, there was a lot of hype surrounding pay for performance especially when Al Gore unveiled his Reinventing Government Plan in the 90s. Pay for success builds on pay for performance but alters it slightly. Instead of paying for delivery of services, for output, government only pays when providers delivers outcomes. Pay for success sounds almost too good to be true. So how do you make it work at your agency?

John Grossman is a partner at Third Sector Partners. He told Chris Dorobek on the DorobekINSIDER program that these procurements have the potential to change the way government buys.

“We have to figure out how to deliver services to the people that are relying on government in the most cost effective and the most fiscally responsible way possible. While we have always had that mandate it is becoming more and more clear that as resources to these places shrink and shrink pay for success is a solution that has real potential,” said Grossman.

Pay for Performance vs. Pay for Success

Pay for performance: “Where we have seen pay for performance contracts in the social services has been primarily in output projects. Government only pays when certain services are delivered. So if government is contracting to provide services for young men after they have left jail, they don’t pay in advance, they pay once the young men have received treatment.”

Pay for success: “What is happening instead is we are evaluating the delivery of that service over the course of months or even years and government only pays when they see the promised outcomes delivered. So the young men have left jail, government will only pay when the providers are able to prove in a scientific way that they are able to keep those young men out of jail at a greater rate than if they hadn’t received the intervention.”

Areas where pay for success works well?

“We are the beginning of experimenting in this space. It works fairly well in the space of recidivism. It works well there because there is large savings for the government to be had to keep people out of jail. There are also robust administrative databases that track whether people are in jail or not. There is lots of hope for this methodology in other spaces. At the same time it will never be a panacea and we don’t pretend that it would be. There are places where the savings aren’t as clear, the population isn’t as well defined and where the outcomes are even longer term then you see with recidivism, so they are not as investable propositions,” said Grossman.

What are the biggest challenges?

Pay for success requires three partners:

• A willing government.
• A provider that is willing to engage in this type of process.
• Funders

“So far we have found that actually the funders are the easiest part of the equation to find. Finding providers that have a track record for measuring outcomes, the capacity to scale and a cost effective program is definitely challenging. Finding governments that are ready, willing and sophisticated enough to get into this space is harder than we would have hoped,” said Grossman.

Ability to fail?

“The nature of this process is that all of these programs aren’t going to succeed. Early on all the players are trying to be very careful to pick interventions that are highly likely to succeed because we know the scrutiny on these early projects are going to be very high. When these projects become scrutinized the very nature of them, mandates that some of them will fail. That’s part of the exciting thing, because government doesn’t pay when the intervention fails. Government only pays when when the mandated outcomes are achieved,” said Grossman.

Why would contractors do this?

“For highly performing nonprofits this is hugely exciting. Right now if you know you are delivering outcomes you are competing against others who are frankly unable to prove that they are delivering outcomes or aren’t delivering outcomes at all. If you are a provider you want to scale and grow your programs based on outcomes,” said Grossman.

State and local level?

“This has been percolating at the state and local level. There have been projects going on in seven or eight states right now as well as a number of counties. The federal government has also been a proponent of this. The Department of Labor gave $25 million in grants last year to two projects, one is Massachusetts and one in New York. Some of the hurricane Sandy relief efforts have been mandated to be done with pay for success. So there is a growing interest on the federal level as well,” said Grossman.

What are the first steps?

• The first step is to figure out what is the actual outcome you are trying to achieve. Where we have seen it work so far is a reduction in incarceration, job readiness, foster care. Some people are doing some really good thinking about early childhood education. So figuring out where there are known interventions that we believe can deliver outcomes and could additionally deliver savings to government if we engage in those preventive efforts.
• A lot of governments have engaged in a RFI process (Request for Information) before they have gone to a formal procurement where they have looked for input from a variety of sources. They have taken that information, digested it, and used that to develop a RFP.

“We think the dialogue that it drives about how are we allocating resources? How are we procuring services? Are we procuring services based on good social science? Or are we doing it based on who has the best pitch? Are we looking at outcomes? That ought to be an integral part of every procurement,” said Grossman.