Commentary

Incentivizing ‘Pay-for-Success? Partnerships at the Federal Level

Interview with Indiana Congressman Todd Young

One of the policy innovations receiving major international attention over the past few years has been the concept of “social impact bonds”-known variously as pay-for-success contracts and social innovation financing-a new type of public-private partnership that generally refers to privately financed, evidence-based social interventions delivered by nonprofits on behalf of governments.

In a prototypical social impact bond initiative, philanthropic interests, foundations, financiers and other social investors raise private capital to cover the upfront costs of social interventions-in areas like recidivism reduction or chronic homelessness, for example-to be delivered by nonprofit organizations through performance-based contracts with government agencies. The contracts are designed to shift the financial and implementation risks of these initiatives to the private sector, in that investors only earn their money back if the privately financed interventions hit predetermined outcome targets specified by governments in the contract.

In short, if the contracts achieve or exceed their target outcomes-and ultimately lower costs relative to current government spending on particular social programs-then investors get repaid and potentially earn a modest return on investment. If the interventions fail to meet their outcome targets, governments do not repay the private investors.

The concept originated in the United Kingdom in 2010 with a social impact bond (SIB) focused on reducing recidivism rates in one prison, and since then it has spread quickly to other countries. As of today, there are now dozens of similar pay-for-success projects in operation or in development internationally, including several major projects launched in the United States-including state and local projects in New York City, New York State, Massachusetts, Illinois, Salt Lake City and Denver-over the past two years.

Though the bulk of the experimentation in pay-for-success initiatives has thus far occurred at the state and local level in the U.S., there is increasing interest in Congress in how the federal government can encourage the maturation of the social impact bond marketplace, with an eye toward lowering government spending on social interventions while improving social outcomes.

U.S. Representative Todd Young (R-Indiana) is one of the early congressional leaders on the subject, having introduced the “Social Impact Bond Act” (H.R. 4885) with co-sponsor Representative John Delaney (D-Maryland) in June. The Act would allow state and local governments to compete for a share of a $300 million federal funding allocation dedicated to social impact bond projects addressing recidivism, long-term unemployment, reduced welfare dependence, child abuse and neglect, and other social challenges currently targeted by federal programs. The goal is to fund projects that work, both in terms of delivering positive social outcomes and taxpayer savings.

Reason Foundation Director of Government Reform Leonard Gilroy recently interviewed Congressman Young on the social impact bond concept, the proposed legislation, the federal role in state and local social impact bond activity, and more.


Leonard Gilroy, Reason Foundation: There’s been growing interest and adoption of “pay-for success” contracting (a.k.a. social impact bonds) at the state and local level in recent years. What attracted you to the concept?

Indiana Congressman Todd Young: For me, it started at the ground level. Like so many congressional districts around the country, I represent a district where I encounter vulnerable and at-risk Americans on a regular basis. As I learned more about many of the federal government programs that they receive that help sustain these populations and aim to improve their lives, I discovered that the results are suboptimal in many cases, to put it charitably.

I think we can do a lot better. That suspicion was reinforced as I started to read about and visit various nonprofits and religiously-affiliated groups-and even some for-profit entities-that are out there figuring out creative ways to significantly improve outcomes for what are believed to be intractable social pathologies.

So I began asking members of staff to research what the evidence base was for many of our federal government programs, and we discovered so many instances in which the evidence base was non-existent with regard to various federal programs and pilot programs that had occurred in the past. Or, outcomes were hard to locate or unclear in terms of the results that had been achieved. That needs to change.

I also believe that we need to engage in more pilot programs with high-quality evaluations to figure out exactly what works across different populations and geographies within this social realm. And we need to scale up what does work.

With all of that said, we have to figure out where those resources are going to come from. The social impact bond (SIB) model-which has already been implemented in the United Kingdom and to a limited degree at the state and municipal levels in the United States-seems to provide a very appealing alternative to just throwing government money at things. It creates a market of sorts within the social realm to scale up evidence-based interventions to improve lives and improve outcomes, first and foremost.

Secondly, because the government only spends money when those lives and outcomes are demonstrably improved, it will save taxpayer dollars in the short term in some instances, and in the medium and long term in other instances. It also shifts the risk of implementation of certain interventions from the taxpayers to the nongovernmental sector, whether that’s philanthropists or capitalists.

On all of those fronts, I find the social impact bond model appealing.

Gilroy: Can you describe the basic components of the proposed Social Impact Bond Act?

Young: It really begins at the state or local level in communities around the country. A state or municipality will approach the federal government-typically partnering with a not-for-profit entity-and propose that a performance contract be struck between the various parties in which certain outcomes are to be reached among an identified population after a certain period of time.

If those positive outcomes are met and an agreed-upon level of government savings is realized, then the investors who pay for scaling up these evidence-based interventions will receive their principal plus a modest return on investment back. If instead, the positive outcomes in the contract are not met, then the investors will get nothing.

This creates a real market incentive and market discipline for nonprofits to work with a sophisticated investor class to ensure that the served population is actually being served well and their lives are being improved in the agreed-upon way. It will also impose that same sort of discipline to ensure taxpayer savings, which in many cases will be redeployed to invest in further scaling exercises and other second-, third-, and fourth-order efforts to assist at-risk and vulnerable Americans.

The process would be guided by the Department of the Treasury, working with the proposed Federal Interagency Council on Social Impact Bonds-which would be a new entity, but not a new department of government. It would be an interdepartmental council.

Gilroy: What is the role envisioned for the proposed interagency council?

Young: The reason it would exist is because so many social pathologies-and their associated social interventions-touch upon government programs and policies that cross different departments of government. So to get around that “stovepiping,” we want to treat a person and whatever challenges they’re facing holistically.

It’s probably useful to give an example. If you’re trying to address low birth weight babies and infant mortality in a major metropolitan area or across an entire state, and you want to implement the Nurse-Family Partnership, for instance-a proven intervention that dispatches nurses into low-income areas, works with expectant mothers to ensure that they’re taking appropriate care of the child in womb, and then continues to work with the mothers for the first few years of that child’s life-you would stand to save a lot on money within Medicaid at the state level. That’s one department of government.

But you could also save money in early childhood education with improved educational outcomes, so there would be savings that could be monetized at the federal Department of Education. And there could also be local savings with respect to local programs designed to assist this population.

You could aggregate all of these savings, which is a lot easier to do in this modern era of Big Data and integrated databases. At the federal level, we want there to be interagency coordination in terms of making sure that these savings are aggregated across different departments of government.

We also want to ensure that when the applications to participate in the SIB program are received by the Treasury Department, that it has the subject matter expertise at hand to select whichever applications may be the strongest. The Treasury deals with financing, money and these sorts of things. That’s why we locate the SIB program there, because we want tough negotiators that will be very sober-minded and negotiate hard with would-be investors and stakeholders in the program.

But at the same time, they need some subject matter expertise as they construct these contracts. The interagency council provides that.

Gilroy: How does your legislation address transparency with regard to state or local social impact bond projects that benefit from federal support?

Young: The bill requires progress reports on each project, every two years, which would be made public. The final report on outcomes-which includes the results of the evaluation and the conclusion of the evaluator as to whether the outcomes were met-would also be made public.

Markets only work perfectly when you have perfectly informed consumers and perfectly rational market participants. In the long term, one might argue that you can approach a perfectly functioning market, but you certainly cannot get there unless market participants are given the information to be well informed.

So our bill will try to create that market of sorts by making public all of these various pieces of information so that we subject the program to critical peer review, and any suboptimal outcomes that might be realized can be made more optimal, and we can learn from both the successes and the suboptimal outcomes in future interventions.

Gilroy: Social impact bond initiatives are complex endeavors involving an array of partners, both public and private. At this point, most jurisdictions have limited experience and capacity for navigating the process of developing these initiatives. How would your legislation address this “skills gap”?

Young: After speaking with numerous stakeholders, we noticed that there was a consensus around the notion that most state and local governments do not have the necessary capacity in order to form an effective SIB contract. They also have major difficulties in obtaining funding for early stage project development for SIBs.

Our bill aims to address this issue by providing up to $10 million for states and local governments to fully explore the possibility of using an SIB through a feasibility study. Feasibility studies aid states and local governments in determining whether an SIB model can be applied to a social issue by examining the project’s expected costs and savings, its strengths and weaknesses, its opportunities and challenges, and the resources required. The federal government covers up to 50 percent of the cost of this study.

Stakeholders are also able to draw from a constellation of existing, mostly non-governmental resources. For example, at the academic level, Harvard University has an SIB lab where they can provide pro-bono technical assistance to would-be applicants for social impact bond programs. They’ve been doing this sort of work with states and municipalities on SIB projects currently underway, and we worked with them as we put together this legislation at the federal level.

There are also philanthropic organizations that can provide technical assistance. A number of them already have the internal expertise. One that comes to mind is the Rockefeller Foundation.

And through our bill, the Department of the Treasury would also have some monies available-$1 million per year-for technical assistance. So we’re proposing that a little bit of money out of the $300 million account that we proposed establishing would be devoted to technical assistance. We just wanted to ensure that we cap it at a fairly modest level so that most of these monies would be set aside for outcome payments.

Then we have the private sector. I visited the United Kingdom some months ago at the invitation of the British Embassy just to study their SIB model and the various stakeholders associated with it, from the highest leaders in the government, to those that design the social impact bond programs, to the financiers, to the not-for-profits involved, all the way down to the beneficiaries of the programs. One of the things I was encouraged to discover is that you actually have the entrepreneurial instinct kicking in to such a strong degree in the United Kingdom. An opportunity has been realized by some of the not-for-profits.

Social Finance UK-one of the SIB intermediaries in the United Kingdom, or meta-project managers-are now employing people to scour municipal budgets to identify what the cost drivers are in these budgets. And trying to link those cost drivers to identified evidence-based interventions that already exist that might help bring down the costs. So it’s the intermediaries who see a business opportunity that are now approaching municipalities. Intermediaries have the expertise to help the municipalities navigate the application process in the U.K., and so both sides having their respective vested interests are working together to implement SIB projects.

Incidentally I would note that most of the interventions in the United Kingdom that are being adopted are actually social interventions that were conceived of and exist here in the United States in our vibrant and unprecedented civil society that De Tocqueville celebrated generations ago and has only grown since. The U.K. is borrowing from that, but they have figured out a way to finance these interventions in a unique way that saves taxpayer money and creates a new market.

Gilroy: State and local governments have been the primary drivers of “pay-for-success” contracting thus far, with some limited financial support from the federal government in a few cases. What do you see as the federal role in this space? In your estimation, how would federal taxpayers potentially benefit from successful state and local social impact bond initiatives?

Young: The real social savings are going to be realized at the federal level, making possible far more projects. That’s because by implementing an SIB program at the federal level, you can aggregate not just the local and state government savings, but also the massive federal savings for taxpayers in programs like Medicaid, the Supplemental Nutrition Assistance Program, and so forth, making what might not otherwise be a positive return on investment opportunity into one.

For example, the state of South Carolina wanted to use an SIB program to expand the Nurse-Family Partnership program. But because the federal-state split on Medicaid in that state is about 70-30, they had trouble showing a positive return on state savings alone. By allowing a federal SIB program to exist, South Carolina could bundle federal and state savings to demonstrate a much higher return.

Gilroy: Given that the only social impact bond projects underway are still fairly new and haven’t matured yet, the jury is still out on how successful the concept will ultimately be. Hence, some might fear that federal taxpayer money might be at risk if your legislation is enacted. How would you respond to this concern about risk?

Young: Investors are only paid back if taxpayer savings are realized, period. Investors are only paid back when the agreed-upon outcomes in the contract are met. We don’t offer taxpayer backstops at the federal level.

The risk to investors is one of implementation. They will be trying to scale up evidence-based social interventions that have proven effective across certain geographies or populations, and replicate them in other geographies and among other populations. Typically this would present an implementation risk to U.S. taxpayers, but this model shifts that risk to philanthropists and private investors, major financial institutions and others.

So the risk is shifted from the government to the private sector, federal taxpayers realize savings, and a portion of those savings will be shared with those who bear risk if outcomes are met.

Gilroy: We’re a long way from seeing final evaluations of the early SIB programs underway, but are you encouraged by the early results seen thus far?

Young: One would never want to judge the success of the SIB model based on one or two projects, because the whole idea of the concept is to learn from suboptimal outcomes and to improve upon those. But earlier this month, the Petersborough prison SIB project in the U.K.-which is well known within the SIB community as the first pilot project, one designed to reduce recidivism in the U.K.-released an intermediate progress report. And I will say that this has been a positive datapoint.

There has been an 8.4 percent reduction in the reoffending rate compared to a national baseline. Suffice it to say that they’ve met their benchmark and are on their way toward achieving their agreed-upon outcomes in the contract. The way that contract was structured, investors stood to realize a bonus payment if they would have hit a higher threshold above the agreed-upon outcomes. They didn’t hit that bonus payment in the most recent report, but they are providing improved outcomes in comparison with the baseline rate of reoffending, so it’s been successful so far.

Gilroy: The proposed legislation has a bipartisan group of sponsors in the House, and elected officials from both major parties have embraced the concept at the state and local level. Why do you think the social impact bond model has such a broad political appeal?

Young: It would improve outcomes for at-risk and vulnerable Americans. It saves taxpayer money at a time of constrained resources. It draws on private investment capital to scale up policies that have been proven to generally work. It shifts the risk of implementing new policies from taxpayers to other private parties.

I can’t find a Republican or Democrat who would disagree with any of those motivations or principles. That’s why I think the concept ought to have-and thus far, seems to have-broad political appeal.

Gilroy: What do you hope to see moving forward in the SIB space?

Young: Scaling is going to be interesting, and all of the different streams of revenue and business opportunities that will come out of this are fascinating. You’ve got major investment banks that are looking at investing in some of these SIB projects. You also have interest from management consultancies, which have the expertise to go into organizations to figure out how to make a project work and then to take that template and apply it in other areas. Replication is a real skill of management consultancies, continuously improving the model and adapting it to different circumstances and contingencies. And of course foundations and philanthropies have a real interest in the concept.

It will be fascinating to see how organically this market might evolve. The market discipline associated with these projects is as important as any other dynamic. When you have investors demanding intermediate progress reports-sophisticated investors who are either looking to pocket a positive return on investment or redeploy the monies to help other people-that imposes a discipline on even the best run not-for-profits that they’re not typically used to.

These SIB implementations also involve some of the best minds in America and beyond, which ought to be very exciting for those who care about at-risk and vulnerable populations and who want the best minds involved in solving or mitigating these social challenges.


U.S. Rep. Todd Young represents the 9th District of Indiana, a 13-county area in the southeast corner of the state. He currently serves on the House Ways and Means Committee, which has jurisdiction over taxes, health care, Social Security, Medicare, international trade and welfare. As a member of the Ways and Means Committee, he serves on both the Select Revenue Measures and Human Resources Subcommittees. Previously, he served on the House Armed Services Committee and House Budget Committee.

Young graduated with honors from the United States Naval Academy in 1995 and accepted a commission in the U.S. Marine Corps. After serving a decade in the military, in 2000 Todd was honorably discharged as a Captain. He then spent a year in England, earning an MA from the School of Advanced Study in London. He also earned an MBA with a concentration in economics from the University of Chicago.

Prior to entering Congress, Young worked several years as a management consultant, advising public and private organizations how they could implement business practices to provide their constituents and customers with more value, often by investing fewer resources.

Other articles in Reason Foundation’s Innovators in Action 2014 series are available online here.

Leonard Gilroy is vice president of government reform at Reason Foundation, a nonprofit think tank advancing free minds and free markets. He also serves as senior managing director of the Pension Integrity Project at Reason Foundation, which assists policymakers and other stakeholders in designing, analyzing and implementing public sector pension reforms.