‘Pay for Success’ in the UK and the US (Oxford Government Review)

By Jeffrey Liebman

A new way for governments to purchase social services, pay-for-success (PFS) contracts backed by social impact bonds (SIBs), is being developed on both sides of the Atlantic. This new approach is being applied to a wide range of policy areas, including prisoner re-entry, homelessness, prenatal care, workforce development, early education, and child welfare.

Pay-for-success contracting combines two tools – a performance contract and an operating loan. Under the former, the government contracts for social services for a specific target population. Instead of paying directly for the quantity of services delivered, the government pays based on the outcomes that are achieved by the services – for example, the number of ex-offenders prevented from returning to prison, the number of unemployed individuals who find stable employment, or the reduction in low-weight births.

Most social-service providers do not have the financial capacity to deliver services, wait several years for performance to be assessed, and only then receive repayment for the services that were delivered. And most are not positioned to absorb the risk associated with a large portion of reimbursement based on performance. For this reason, many PFS projects include an operating loan from private funders who provide upfront capital in exchange for the lion’s share of the government payments that become available if the performance targets are met. If the targeted level of outcomes is achieved, the loan is repaid with interest from the government’s performance payments. If the minimum outcomes are not achieved, the investors can lose all of their principal. This loan is a SIB.

Governments in both the UK and the US have been testing this approach in an attempt to make more rapid progress in addressing challenging social problems. The first PFS endeavour was the prisoner re-entry project in Peterborough (in the UK), and there are now more than 30 PFS projects in the UK. In the US there are 11, with five more expected to launch in late 2016.

The model offers three main benefits for governments. Firstly, it helps them reorient budgets toward prevention: governments spend large sums paying for the consequences of bad outcomes – putting people in prison, providing unemployment benefits, paying for medical care – but find it a struggle to afford the investments that can prevent these bad outcomes. PFS contracts, by offering taxpayers what is essentially a money-back guarantee if outcomes are not met, are encouraging governments to make greater investments in preventive services. Secondly, PFS contracts are enabling governments to sustain multi-year, outcome-focused partnerships with service providers to re-engineer systems to produce better results. Data is being used in real time to ensure the right services are being delivered to the right clients, and that clients are progressing through the service-delivery model successfully. Because payments depend on outcomes, there is a much greater urgency to solve implementation problems than occurs with ordinary social service contracts. Lastly, the PFS approach can help government learn which programmes work. In particular, some PFS contracts are comparing the outcomes for people referred to services to those of a control group of people who are not being served; this rigorously determines the impact of the social spending. Rigorous evaluation in a PFS project helps solve what I like to call the ‘immortality problem’ in government budgeting – that once a programme gets in the budget, it tends to receive funding year on year, regardless of effectiveness. In contrast, if a PFS project fails to achieve its target outcomes, it will fail quite visibly, and it is highly unlikely that those services will continue to receive funding in the future.

The New York State prisoner re-entry initiative provides a good example of how the PFS model can lead to a re-engineering of the systems that connect target populations to services. Under this contract, New York is obtaining training, transitional jobs, and job placement services from the Center for Employment Opportunities (CEO) for individuals being released from prison, with the goal of increasing employment and reducing re-incarceration. The state is using data in four innovative ways in this project. Firstly, the state is using a predictive model to identify the individuals being released from prison who have the highest probability of re-incarceration and is referring only those high-expected-cost individuals to CEO’s relatively intensive services. Second, the state is making approximately 700 referrals per year to CEO and holding CEO accountable for the outcomes of all 700 individuals regardless of whether or not they receive CEO’s services, thus creating a strong incentive for the provider to track down and enrol as many of the individuals in training as possible. Third, an operations committee with representatives from both the state and the provider meet regularly to review data on the percentage of former prisoners who make it to job training within a week or two of release. This committee can then take immediate action whenever the numbers lag below targets. Fourth, because there are not enough slots in the programme to serve all of the high-risk individuals, the state is using a lottery to determine which eligible individuals receive referrals to CEO. The lottery allows the impact of CEO services to be rigorously evaluated by comparing the outcomes for those referred to CEO to the outcomes of those who were not referred to CEO. The success payments in this project are based on the results of this randomised controlled trial (RCT).

The PFS model has evolved somewhat differently in the two countries. In the UK, there have been a large number of relatively small contracts, with the typical project requiring investment of approximately £2 million, while in the US, there are fewer projects, but many of them receive more than $10 million. Additionally, UK PFS projects tend to measure outcomes for the individuals being served and make payments based on those outcomes, without explicit comparison of the outcomes to a counterfactual scenario. In contrast, seven of the eleven SIBs in the US are being evaluated with RCTs.

In the US, I often hear PFS practitioners lament that they wish our PFS market was evolving to be more like the UK’s – with governments issuing ‘rate cards’ listing the prices the government is willing to pay per outcome achieved and a high volume of very similar transactions. In contrast, pricing in US PFS projects has been determined separately for each project, resulting in more complex negotiations and bespoke deals. In light of this lament, I was amused during my recent visit to Oxford to hear UK PFS experts say that they are hoping their PFS market evolves to be more like the US’s, with larger projects and more rigorous impact evaluations.

In both countries, there is clear excitement around the possibility that the lessons being learned from PFS projects will spill over into how other government-funded social service programmes are managed – even those that don’t pay based upon outcomes. In particular, PFS is teaching governments to actively manage their social-service contracts by meeting regularly with service providers to review data on programme operations, and jointly determine how to improve the results being achieved for vulnerable populations.

While at Oxford, I was asked what advice I would give to governments that are in the initial stages of considering the pay-for-success approach. Here is my answer: governments should view pay-for-success as a leadership tool that allows them to sustain intensive work with service providers to improve systems for serving vulnerable populations. All of the financing and contracting is simply the admissions ticket that provides entry to the main show, which is the re-engineering of systems. In choosing among the many possible policy areas to which pay-for-success can be applied, governments will have the greatest impact if they select an issue area where the potential benefits from outcomes-focused systems re-engineering are large.

Professor Jeffrey Liebman is the Malcolm Wiener Professor of Public Policy at the Harvard Kennedy School. During the first two years of the Obama Administration, he served at the White House Office of Management and Budget in a variety of roles.