Social Impact Bonds: A Conversation with Simon Jawitz

This article is featured in NPQ‘s upcoming issue on the changing landscape of nonprofit finance.

Editors’ note: Social Impact Bonds may be the latest and newest instrument for drawing private investment into entrepreneurial social programs. But with only one major example of Social Impact Bond financing, from the United Kingdom, and efforts to gin up Social Impact Bond initiatives in Boston and the Twin Cities having only just begun, there are as yet too few results to trumpet. NPQ contributing editor Jon Pratt sat down with Simon Jawitz, board member of and senior advisor to Growth Philanthropy Network, to discuss the concept’s prospects.

Jon Pratt: To introduce this topic to our readers, I would like to set the table with a working definition of Social Impact Bonds lifted from the Social Finance website. Here is how one of the early players describes them: “Social Impact Bonds provide up-front funding for prevention and early intervention services, and remove the risk that interventions do not deliver outcomes from the public sector. The public sector pays if (and only if) the intervention is successful. In this way, Social Impact Bonds enable a reallocation of risk between the two sectors.” There is no doubt that this is an interesting concept, and some people see enormous potential for Social Impact Bonds to bring for-profit investment money to the nonprofit space for everything from healthcare and education to affordable housing and human services. Where do things stand now?

Simon Jawitz: I think your characterization of people seeing huge potential is accurate. There is a lot of energy and enthusiasm for the concept of the Social Impact Bond. Somebody I spoke to recently, who is very actively involved, characterized the current level of interest as undergoing exponential growth.

But the level of actual transactional activity is a very different story. So far only one transaction has occurred, and it was overseas—the U.K. Social Impact Bond (at HMP Peterborough) between the Ministry for Justice and Social Impact Partnership. And this was a very small transaction in the scheme of things—$5 million pounds sterling. One can draw very different conclusions from that. One can draw the conclusion that the Social Impact Bond has been hyped beyond anything that’s reasonable; or alternatively, that people look at this and see huge potential, even if they recognize that there’s a very long runway, and it’s probably going to be a slow takeoff. There is a tremendous amount of dedication out there, not only to getting individual transactions completed but also to being very transparent about them. So I think one of the strategies the players are focusing on—folks like Third Sector and Social Finance—is to make sure that as these transactions get completed, information gets out very broadly to the public so that a base of knowledge builds up, making it easier for additional deals to get done.

JP: So in the case of the British example, where the Ministry for Justice is expecting their service provider to reduce recidivism by, say, 7.5 percent, have you heard about how it’s actually going?

SJ: Not yet. It may be too early.

JP: Who are the investors in the Peterborough Social Impact Bond?

SJ: The only one that is a matter of public record is the Rockefeller Foundation. Other than that, I’m not aware that the investors have been identified. What I do understand is that if you wanted to characterize them, you would characterize them as, like the Rockefeller Foundation, not your
traditional financial investor looking for a return; they’re more the “impact investors”—people who are looking for both social impact and, potentially, a financial return.

JP: So maybe the same types of institutions that might fund these types of services as a project, but here they’re using their assets rather than their grants.

SJ: That’s exactly right. I think that part of it is that the Rockefeller Foundation—and I don’t want to speak for them, but I’ve had many conversations with them on this point—are one of the many institutions in this country that see great potential in this product. They would like to see the field develop, and hence—and, again, I don’t want to speak for them—I think that was at least a part of the reason for their participation in the U.K. deal and a part of the reason why they have given grants to Social Finance and Third Sector Capital Partners. Other organizations have made similar grants. The Boston Foundation has also made a grant to Social Finance, as did the Pershing Square Foundation, established by the hedge fund Pershing Square Capital Management.

JP: Is the underlying premise of the Social Impact Bond that there’s to be a cost savings for government and this is why government would want to participate in this?

SJ: Yes. The underlying premise is that these interventions will, if successful, result in hard-dollar savings to the government. And, since the government pays only if, in fact, the interventions are successful, at the end of the day—so long as the government doesn’t end up paying out more than 100 percent of its realized savings—it comes out ahead financially and has also avoided taking on the risk up front. So, there are really two benefits for the government: avoiding taking on the risk that the intervention will be successful and only paying out of savings that it actually receives.

JP: So, potentially, say, in the British example, if there was no change in recidivism, it’s possible that the bondholders would just completely lose out.

SJ: It’s not a possibility—it absolutely would be the case. In fact, if they don’t hit certain benchmarks there is no payout whatsoever.

JP: It’s kind of this combination of new money, private money, and then demonstrable results.?.?.?. So what are the actual savings? For example, one area this has been promoted in is job training. Take, for instance, people leaving prison. If you could help them get employed they would be paying taxes, and this would reduce cost to the state. But is the state receiving money it would not otherwise have received if this employed person who pays taxes had not been trained by this organization? What if the employer had hired someone else? The state would have captured the same tax payments. So, after all, employers are hiring not out of charity but because they need a body to be doing the work. They’re satisfying a need that they have in the marketplace.

SJ: Right. If you wanted to focus on a group of unemployed people—and I’m just thinking about this as we talk—and you want to give them job training, the theory being that they would then go on to get a job and increase tax revenues .?.?. I think you might be able to develop something, but it would be extremely difficult, partly for the reason that you just identified: What’s the control group? What would have happened absent the intervention?

Let me use another example—supportive housing for the homeless. There are many studies and a great deal of data that demonstrate that homeless people are extremely high maintenance in terms of visits to hospitals, use of emergency rooms, and incarceration. So the theory there is if you can provide an alternative in the form of some kind of supportive housing, you can potentially avoid all these other high costs that go along with having to take care of that population on the street. Numerous studies have demonstrated the ability to quantify those savings.

JP: Right. But isn’t this the argument that’s commonly made for appropriations? That it is in the community’s interest to undertake this expenditure because there’s some other benefit that may well be an economic one?

SJ: You’re right. But the difference here is that in this particular case we’re not just arguing that we should do this because it may have this result—we’re saying, let’s put this structure in place. We believe that we can get the results, and we’re only going to pay if, in fact, we do get these expected outcomes and the resulting actual savings.

JP: Do you see any possibility that by creating this new stream of revenue, support—including political support—for current appropriations could be decreased?

SJ: Absolutely. You talked about two things earlier. You identified new money coming into the system. So that would be your third-party investors. But the other big thing is evidence-based programs and outcomes. Those are the two key components of the Social Impact Bond structure. I think the second, quite frankly, is more important than the first, meaning that yes, perhaps we can find new sources of capital to cover the upfront cost of these programs. Maybe initially it will be impact investors. Potentially, down the road, if the Social Impact Bond becomes established enough and well understood enough, you might find real, pure financial investors putting in the money for these programs up front. But, ultimately, you’re still limited by the amount of money that the government is going to be willing to pay out.

So while I think the concept of new money is certainly important, the more salient aspect is that we are taking another step along the road of moving philanthropy more generally in the direction of focusing on outcomes, and not outputs, and putting higher value on those programs and nonprofit organizations that have proven results. I believe that a lot of the people involved in Social Impact Bonds see a shift down the road where—whether governments are using this structure or not, whether they’re taking the risk themselves or not—governments are going to prefer to fund those programs with proven outcomes.

If I’m a nonprofit organization, and I’m hearing about this structure for the first time, what should I be thinking? Forget about all the structural complexities and the details. I should be thinking at the very least that here is yet more evidence that philanthropy is moving in the direction of funding evidence-based programs. I would want to start thinking more seriously about my own program, and what data I had and what data I could get, and position myself—whether for a Social Impact Bond or not—to be able to propound my program versus alternative programs that are out there. And, I think there’s going to be increasing pressure in that respect as we go forward. We all know about the current financial and economic situation and the budgetary constraints at all levels of government.

As a governmental official I would have thought that the idea of making sure that my dollars are being utilized effectively—more precisely, that the taxpayers’ dollars are going toward programs that actually generate results—is logical and compelling. Now, those results don’t always have to be dollar savings for me—as opposed to lots of other benefits and results that the government wants to support—but at least I know that with regard to money being spent, I’m getting the results that I’m looking for.

JP: This goal was part of the idea behind No Child Left Behind—that we should be able to measure the results and reward them. Is it possible, though, that the increased attention on payment conditions on meeting specific targets could have the effect of increasing pressure on the providing organization and also result in a change in the nature of the relationship between the service provider and the person served? In other words, do we want nonprofits choosing only the easiest clients so that they’re more likely to succeed, and avoid the most difficult populations?

SJ: The contracts have to be structured very carefully.?.?.?. One of the most important things is to make sure that the service provider is not cherry-picking, because then you’re really defeating the purpose. It comes down to the negotiation and the structuring of the contract between the state and the service provider. I think another very important aspect of all this is transparency.

My own view is that it’s absolutely critical that all aspects of this transaction be clear and well understood by all the parties. And that would include the relationship between the service provider and the recipient of those services. That has to be out in the open. That has to be vetted. That has to be understood. I think that will go a long way to making sure that we don’t inadvertently create perverse incentives that push organizations or people towards conduct that we don’t really want to encourage.

JP: What advice would you give to government officials developing these pilot projects? What should they pay most attention to getting right?

SJ: Oh boy, there’s a long list. In no particular order may I just try to give you some thoughts? Well, one is—as we’ve just discussed—making sure that all the incentives are lined up, that the transaction is properly structured. You have to make sure that you can measure the outcomes properly, which means some kind of control. You want to make sure that the service provider is not cherry-picking, so that you’re getting improvement in the population as a whole. If all goes well, and the deal is structured properly and the targets are met, the government will make a payout that will go to investors. That payout will by its very nature be greater than what the government would have paid if it had just contracted for the services upfront.

That should be clear to anyone. Investors are putting up the money that covers the cost of the program interventions. They will be looking to get that money back plus an appropriate return based upon the risks they are taking. So the equity kicker, so to speak, that the investor receives, is an incremental cost to the government that it would not otherwise be required to pay out. The government is paying out more than it would have paid out if it had done this the traditional way, which is basically just entering into a contract and paying for the services.

Now, the benefit of doing that—and it has to be weighed—is that the government is avoiding taking the risk that the program doesn’t work. But how valuable is that? How much real uncertainty is there in the program? If there’s so much uncertainty, then you have to ask why any investor is going to be willing to take the risk. So it requires a careful balancing of risk and an intelligent assessment of pricing in light of each party’s incentives and priorities. Obviously, the government will want to be able to assure taxpayers that their money is not being wasted—that while the government is actually paying a premium, it is getting real value in exchange.

There are many other considerations for the government. Are the necessary data available and reliable? Are the agreed outcomes measurable? Is there a clear agreement on how those outcomes translate into savings? Now, as currently conceived, the Social Impact Bond is structured with an intervention and then a measurable outcome. So, for example, in the case of the U.K. Peterborough deal, there is an ongoing intervention, and the outcome that’s being measured is recidivism. The payments, if any, to be made to investors will be determined by the reduction in recidivism. In a way, that is the end of the transaction as far as investors are concerned. Of course, they will have an interest in understanding how the government computes its savings, if only to assure themselves that they are getting a fair share.

But determining how a reduction in recidivism translates into actual cost savings is very likely to be the thorniest problem of all. It’s an extraordinarily complicated question, because you’re going to start getting involved in budgetary issues and allocations and overhead and who captures the savings and where it goes.?.?.?. But it’s obviously something that governments need to be focused on. I would advise governments looking to do this to make sure that they work with or hire advisors or intermediaries—people versed in how Social Impact Bonds work—and make sure that all the conflicts of interest inherent in putting these deals together are understood.

JP: What is the current level of Social Impact Bond activity in the U.S.?

SJ: Domestically, most of the current activity on Social Impact Bonds seems to be taking place in the Boston area. As you know, a few months ago the state issued a request for information, asking people in the field to educate them about their potential applicability. They received, I think, more than twenty-five responses—from academics, intermediaries, and nonprofits. I have not been involved in those discussions or meetings, but from what I have been told they have been working diligently since then to try to understand exactly what the concept is and how it could be used in Massachusetts. I think I’m not saying anything that’s not a matter of public record when I tell you that Professor Jeffrey Liebman from Harvard University, who is one of the driving forces in pushing this along in Massachusetts, has been engaged by the state to help them with that process.

I’m told that within the next month it’s likely that Massachusetts will be asking for a second round of information, and this will probably be closer to a request for proposals. Once they get that, they may identify and choose two pilots: one in the area of criminal justice (similar to the deal that was completed in the U.K.) and the other likely in the field of homelessness. At that point they’ll get down to the hard work of actually finalizing the deals to get these two pilots done. These are expected to be very complex, time-consuming transactions involving many parties, and necessitating involvement by the legislature because most of these transactions will require some kind of enabling legislation that will allow the state to enter into pay-for-performance contracts as well as to allocate funds for a multiyear purpose, which generally is not the way funds are allocated. So a lot is going on in Massachusetts.

I’m also told that other states around the country are looking into Social Impact Bonds; I know less about this, but I’m told that Virginia and California are also taking a hard look, and we know that something similar is being explored in Minnesota. New York City has been studying this for some time. And then of course there’s something in the proposed 2012 federal budget allocating up to $100 million for seven pilot programs in five different areas. But we will have to wait and see how that all plays out, obviously. Also, on October 21, the White House Office of Social Innovation and Civic Participation co-hosted an event with the Nonprofit Finance Fund to bring together representatives of state and local governments from around the country to share information regarding the potential of Social Impact Bonds and efforts currently underway. This is certainly a clear indication that while development of Social Impact Bonds is still in the early stages, their potential has caught the attention of both public and private sector players, all of whom are working hard to make Social Impact Bonds a significant and meaningful part of the U.S. landscape.

Could Social Impact Bonds Increase Pressure on Client Relationships?

Social Impact Bonds are expected to change the relationship between service providers and government, but that’s not the only relationship they may change. By increasing the stakes to achieve performance goals based on changes in their clients’behavior, with direct financial consequences for failure, organizations have stern incentives to make sure their clients perform as agreed.

The particulars of how this innovation might affect service providers is unknown, but an early example might be in Minnesota, where the 2011 Minnesota legislature approved a $10 million pilot project called Pay for Performance Bonds, which share some of the goals of Social Impact Bonds.

Minnesota’s Pay for Performance legislation was championed by Steve Rothschild, founder of Twin Cities Rise!, a job training and placement organization. Twin Cities Rise! initiated a funding agreement ten years ago with the state’s Department of Employment and Economic Development, whereby it would only be paid—after the fact—if it could document successful job placements.

To underscore the seriousness of its approach to clients, Twin Cities Rise! requires clients to sign a participant agreement in which they pledge to remain actively engaged in the program or else must repay the cost of the program, not to exceed $7,500. For one young woman, who found a job on her own and left the program, that meant receiving an intimidating letter that concluded, “Consequences of non-response to that invoice could also include TCR! bringing your account to conciliation court, where an unfavorable judgment would negatively impact your ability to gain credit with lending institutions.” On April 11, 2011, Twin Cities Rise! filed suit against the client for $3,338 plus a $70 filing fee.

Jon Pratt, JD, MPA, is the executive director of the Minnesota Council of Nonprofits and a contributing editor to the Nonprofit Quarterly.