Does “Pay for Success” Actually Pay Off? The ROI of Social Impact Bonds

Print Share on LinkedIn More

Swirling abstraction

The concept of Social Impact Bonds is so popular in many circles that politicians believe that they can score points when they make SIBs part of their campaign platforms. Winnipeg mayoral candidate Paula Havixbeck pitched SIBs as “a new tool that other countries have had tremendous success with.” The press report didn’t identify which SIBs Havixbeck might have cited as tremendous successes, given that none of the handful of SIB demonstrations has reached any point of financial or program success, though some may in the end do so.

Established politicians have latched on to the concept, too. In Chicago, Mayor Rahm Emanuel announced an ambitious program of social impact bond financing to enroll more children in pre-K classes. The program, to be launched next month, would get 2,600 children into half-day preschool classes over four years. The SIB targets would be met with payments for each pupil who, after pre-K, does not get placed in a special education program ($9,100 per pupil), is deemed ready for kindergarten after pre-K ($2,900), and scores above the national average for third-grade reading ($750).

SIBs are also the cloak of respectability sought by some wealthy investors, or so some SIB pitchers believe. In the Huffington Post’s “Impact” page of contributed articles, Sharon Foulston, who describes herself as “founding a Social Impact Bond to finance a home for single mothers,” a “Business Angel for many years mentoring and investing in entrepreneurs,” and interested in “the mystical world of Tarot,” offers a creative pitch to moneyed people. Pitching her SIB idea to “the second generation of wealth owners, the inheritors of a period of industry and sustained economic growth,” Foulston describes those inheritors of wealth as having reached “a higher level of social conscience, seeking rewards far beyond that which mere money can bring.”

In a way, it’s easy to see the attraction of Social Impact Bonds. At a time when government seems to be doing many things ineptly, from the Department of Veterans Affairs’ management of VA hospitals to the Centers for Disease Control’s mercurially unreliable protocols and actions against a stateside outbreak of Ebola, SIBs radiate the glow of a public/private panacea that might cure the dysfunction of government—even without evidence that they can do so. Preferring the nomenclature of “pay for success” over the “boring” language of Social Impact Bond, the respected McClatchy-Tribune op-ed columnist Jay Ambrose exudes that these products of “our high-energy, imaginative, entrepreneurial, can-do spirit and charitable impulses” could materially contribute to the possibility that “we will help people out of desperate circumstances, that we will become a society with less disadvantage, less crime, less unemployment.”

Why are Social Impact Bonds so popular when their track record is next to nonexistent? The combination of Ambrose, Foulston, and multiple politicians gives the answer. The American populace sees government and politicians as stymied by party divides, but the SIB/PFS concept has garnered the support of politicians across the ideological spectrum, including endorsement from think tanks of the left (the Center for American Progress) and the right (the American Enterprise Institute), with the appearance that these projects risk the moneys of investors rather than taxpayers, and if successful, will save money in virtually any area of public endeavor.

As Ambrose explained the protean potential applicability of SIBs, “A government agency identifies some social dread it wants fixed and then strikes a deal with a non-governmental organization to fix it…[which] will then find investors to fund managers who put together ways to tackle the problem.” As an example, the G8 Social Impact Investment Task Force report, released just last month, suggested that social impact bonds, renamed “development impact bonds” for their application in developing nations, could be, if properly structured and targeted, “an innovative way of tackling the social and environmental issues that constrain private sector development and economic growth: tackling issues in education such as literacy and drop-out rates; in health such as sleeping sickness, malaria and other chronic debilitating diseases; and in employment such as training unemployed youth for available jobs.”

Looking at Costs

The investors behind the highly touted but still unproven Peterborough Prison social impact bond, aimed at reducing recidivism of short-term male prisoners released from incarceration, were not profit-minded investors, but 17 foundations and charitable trusts that committed £5 million. In the U.S., the investors behind the Rikers Island project to reduce recidivism are not the “second generation of wealth owners” who had stumbled into “a higher level of social conscience” that Foulston described, but Goldman Sachs and Bank of America Merrill Lynch, banks that most Americans might not characterize in Foulston’s words, and in the case of Goldman Sachs, would receive a financial reward from the project if it worked as planned of over 20 percent, but with a guarantee from Bloomberg Philanthropies, boosted to well north of 80 percent.

The role of philanthropy is critical in the U.S. SIB experiments, and not only with Bloomberg’s unusual risk guarantee. Even in Chicago’s new pre-K SIB, philanthropy plays a important investment and financing role, with the J.B. and M.K. Pritzker Family Foundation taking a subordinate position behind the Goldman Sachs Social Impact Fund and the Northern Trust taking senior roles—meaning that if the project were not to fully pay out, Goldman and Northern get paid first, and if there were anything left, Pritzer might get it. If the project does succeed, it’s probably true that the senior loans get paid at a higher interest rate than the subordinate ones. That is similar to the financial structure that Goldman and Pritzker used for a pre-K SIB in Utah.

One gets the sense of foundations in SIBs playing roles that they have long performed through their program-related investments, providing guarantees for private investors such as banks that “risked” construction or permanent financing for various kinds of projects, though those loan guarantees typically supported capital projects such as affordable housing and facility development as opposed to the costs of social programs.

Examples of foundations that have made loan guarantees through program-related investments in recent years include the Rasmuson Foundation, which in 2010 provided credit enhancement through a first loss guarantee for new facility and housing development loans originated in Alaska by nonprofits; the John R. Oishei Foundation’s loan guarantee for Buffalo River Fest Park; and the Meyer Memorial Trust’s combined grant and loan guarantee for the Freshwater Trust’s stream and river restoration efforts in Oregon. In a more recent guarantee, the Kresge Foundation in 2013 gave the Calvert Foundation a $1.25 million loan guarantee toward Calvert’s participation in the Woodward Corridor Investment Fund in Detroit, a $30 million effort to finance mixed-income, mixed-use projects in walking distance of the city’s downtown core.

Loan guarantees like these PRI investments or Bloomberg’s for the Rikers project don’t usually involve a cash outlay, but they do put that amount of the foundation’s balance sheet at risk should the deal go sour. Somehow, the notion of Meyer Memorial risking tax-exempt foundation resources on the Freshwater Trust or even the Kresge Foundation for the Calvert Foundation’s investment in Detroit feels a little more philanthropically authentic than providing guarantees for Wall Street investment banks that have more capital than these philanthropic institutions might ever hope to see.

There may well be types of investments in which social impact bond-like structures might be significant boosts toward testing and financing high-risk solutions to social problems. If the likes of Wall Street were to invest really high-risk money in nonprofit projects like planning money and predevelopment financing, much as many community development intermediaries and some foundations do in the form of “recoverable grants,” that would significantly propel nonprofits to do things that they otherwise cannot, given limitations in governmental funding. However, at least at the outset of the SIB social experiment, the projects in the first wave of SIBs seem to be occurring in areas where the techniques for success have been well demonstrated without SIBs. In reading the information on sites such as the Federal Reserve Bank of San Francisco’s “What Works for America’s Communities,” the U.S. Department of Education’s What Works Clearinghouse, and the Urban Institute’s What Works Collaborative, as well as Lisbeth Schorr’s groundbreaking 1988 book Within Our Reach, much of what is needed in our society is known and available—the obstacle is the lack of sufficient political will to put public capital behind these strategies and projects that really can make breakthrough changes in social conditions. The evidence that paying private investors such as Goldman Sachs a premium in the form of a SIB will lead to that missing public will is not yet visible.

The prospect of private sector pump-priming capital investment plus the promise of governmental savings add up to a strategy that SIB supporters hope might break the logjam of public sector inaction. However, the pump-priming process at this stage of the nation’s understanding of social impact bonds is a bit topsy-turvy. Rather than having Goldman Sachs and other private investors take high-risk positions to demonstrate the concept, the private investors and various SIB financial packagers are looking for government to put money behind incipient SIB efforts to demonstrate “proof of concept.”

The various recent public investments geared to pay for consultants and other costs associated with structuring SIB/PFS show multiple funding sources. For example, the Salt Lake County Council just voted to give Third Sector Capital Partners $150,000 from its regional development budget to figure out its next pay-for-success project. In this instance, however, it isn’t that Salt Lake knows what it wants to do as its next SIB following its experience with its first, a pre-K project, but it is taking the money for Third Sector from its regional development budget to figure out what social problem to take on next.

Social Impact Bonds ae not actual bonds, but a financing mechanism in which private capital finances the implementation of a program, repaid if the program reaches predetermined achievement goals. In Richmond, California, to finance housing rehabilitation efforts, the city is issuing actual long-term municipal bonds paying low interest rates (in contrast to the return on investment anticipated by SIB investors such as Goldman Sachs). The bond will be underwritten by RBC Capital Markets, the Canadian investment bank, and will probably be issued at the city’s AA- credit rating. Wells Fargo and Bank of America have been identified as potentially interested investors in the bond. It isn’t quite clear how a municipality’s decision to issue housing bonds is all that new, given the long history of states and cities issuing mortgage revenue bonds and multifamily housing bonds.

Legislation in front of Congress right now, H.R. 4885, the Social Impact Bond Act co-sponsored by Indiana Republican congressman Todd Young and Maryland Democrat John Delaney, would provide $10 million for states toward the cost of Social Impact Bond feasibility studies and additional federal moneys toward the costs of third-party SIB evaluations. With an array of other federal legislators who have subsequently endorsed the bill, ranging from Jared Polis (D-CO) and Joseph Kennedy (D-MA) to Dennis Ross (R-FL) and Aaron Schock (R-IL), the Social Impact Bond Act exemplifies the bipartisan nature of support that SIBs attract, even with calls for increased funding. But when it comes to the technical assistance and evaluation costs that would typically be wrapped into the overall PFS/SIB financing structure, this legislation puts government into the position of paying for and subsidizing the costs that the SIB structure should in theory cover.

Although conservatives have been highly critical of many of President Obama’s social program initiatives, there was no noticeable Republican criticism when, on October 1st, the Social Innovation Fund in the Corporation for National and Community Service announced eight grants for a total of $12 million to launch pay for success initiatives. Each of the grantees will receive between $200,000 and $1.8 million a year over three years, with a dollar-for-dollar match. Grant recipients include entities that will structure and provide technical assistance to SIB/PFS projects, such as the Nonprofit Finance Fund ($3.6 million), the Harvard Kennedy School Social Impact Bond Lab ($1.93 million), and the aforementioned Third Sector Capital Partners ($1.9 million). It is striking that the SIF program theory is that federal front-end predevelopment financing is needed for programs with strong evidence-based track records, such as the projects often selected for SIB/PFS efforts in the U.S., and for major investment groups with substantial histories of project financing. This kind of front-end planning and development money is much less available to smaller, community-based nonprofits testing unproven program models rather than large, powerful nonprofits with major private financial partners along the lines of the typical SIF/SIB/PFS grantee.

Whatever the specific SIB structure, government is always on the hook to pay for “successful” SIBs, in that the government take-out for the project having reached predetermined goals is key—except that it involves government take-out plus a solid return on investment. It means that government has to appropriate in advance the funds that will pay the investors the costs of their working capital plus their anticipated rate of return, else investors might wonder whether a future legislative body might actually appropriate the funds promised by a previous legislature. For example, Massachusetts has had to allocate up to $27 million for its anti-recidivism SIB. Private investors (banks and foundations) have put $18 million into the project, but anticipate payments of five percent for senior loans and two percent for junior loans if the project is successful. Ostensibly, long-term costs to government are reduced by the savings from avoiding re-incarceration, but government does have to pay at some point. Programs that could and perhaps should be paid for by government without the necessity of a SIB to attract private investors, such as Utah’s pre-K SIB, defy logic. Why should government appropriate funds up front to pay investors’ costs when the concept of pre-K is proven, desirable, and broadly supported by voters and legislators?

These social programs are rarely so discrete that they can be separately financed by private investors without government playing a role at the very beginning as well as at the end. For example, the Chicago pre-K SIB requires not only upfront private investment, but $4.5 million in state grants and $10 million in city capital funds. The capital expenditure is necessary, we would guess, because there currently aren’t enough pre-K spaces in Chicago for the desired SIB pre-K population. A Goldman Sachs representative, Andrea Phillips, said, “Innovative models like social impact bonds and Pay for Success programs allow the private sector to provide the capital needed to expand successful initiatives in our cities and communities, shifting the risk of achieving targeted outcomes away from the taxpayer and enabling governments to pay only for what works,” but the reality is that the success of SIBs depends on government investment prior to the SIB project, during the SIB to pay for costs that the investors can’t or won’t, and after the SIB project once government has to ante up a payment for program costs plus a return on investment.

The public hears about private capital investment, senses governmental dysfunction as of late, and quickly finds the Social Impact Bond concept alluring even while few people know what it really is or whether it will truly save money. Voters might be surprised to learn that government has to be prepared to pay—and pay handsomely in some cases—for having allowed private investors to invest money in projects and concepts that are already well known to work. Why wouldn’t the public invest in pre-K, supportive housing, and anti-recidivism projects on their own merits? Do voters need the validation that comes from having Wall Street bankers bless the projects? We suspect that few people understand the levels of government appropriations that might be needed to support PFS/SIB projects both during their implementation and after they are completed and deemed successful—or that government might be needed to provide some up-front SIB project costs.

The queue of potential intermediaries offering their services in structuring and assisting SIBs is out there, is increasing, their motivations probably a mix of commitment to the social issues to be addressed by SIBs—at least we hope so—plus a large dollop of financial expectations for their own bottom lines. For some SIB adherents, their interest in PFS/SIBs is embellished by a strong sentiment that government is unaccountable, uninterested in results, allergic to evaluation and feedback, and largely unproductive. Their ideological predisposition against government seeps out in various settings when they declare, perhaps with unintentional honesty, that half of government programs have never succeeded with measurable results in any social program endeavor—ever. Some people may even imagine that social investors are somehow more enlightened human beings than the public servants who do the public’s work in city, state, and federal agencies. Among a segment of social entrepreneurs of late, there is a growing antipathy toward government—even as they seek out government subsidy.

There are, however, people who truly believe in the efficacy of social impact bonds, even while they acknowledge that there is much to be done to hone the concept to make it practical. Representative Joseph Kennedy III, a co-sponsor of the Young/Delaney legislation, wrote what seems a heartfelt plea for support for the legislation and for SIB models such as the Massachusetts anti-recidivism project being carried out by the Chelsea-based nonprofit Roca. But even Kennedy falls prey to the idea of barriers between the public and private sectors that SIBs can overcome. A team of Brookings Institution researchers recently concluded, “While there is still much to learn in designing and implementing social and development impact bonds, early experiences are somewhat promising in that, at the very least, they have gotten the bureaucrats, the investors and the providers of services to the same table and with aligned interests at that.” It is sort of an imaginary world view, in which these observers imagine government as somehow immune from doing the bidding of private sector interests, as opposed to the typical role of politicians feeding tax subsidy upon tax subsidy to private corporations and succumbing to the political process dominated by direct and indirect corporate campaign contributions.

There are important roles for private capital to leverage—or be leveraged by—public funding, especially if the corporate funding boosts the capacity and access to working capital of nonprofits tackling issues that don’t come with ready-made, proven solutions and truly need high-risk capital. But if the programs are known to work, as many have been demonstrated to do before the concepts of “pay for success” or “social impact bonds” ever entered the public policy lexicon, it seems so much more logical for government simply to pay for the programs to get them done and not add returns of 7, 10, 20, or 80 percent to attract Wall Street investors. And if the challenge now is to devote government money to the front-end costs of putting together investor-financed SIBs, as Salt Lake County, state governments, and the Social Innovation Fund are now doing, the process is headed in the wrong direction. The nonprofit sector needs feasibility, planning, and predevelopment financing, not big Wall Street investors or the financial intermediaries that anticipate solid paydays from the PFS/SIB concept.

  • Jan Masaoka

    Thank you for raising these points. Too much of the nonprofit sector has allowed itself to be led by the nose into supporting these efforts without adequate scrutiny. Why do we allow ourselves to be dazzled by the “allure” of one too-good-to-be-true scheme after another?

  • Brian O’Shaughnessy,

    The “pay-for-success” movement is not as sinister as Mr. Cohen sometimes portrays. Yes, there are potentially barbarians at the gate, but from a larger perspective, the conversation seeks to make measurement and financial impact real components in the world of social services. Displaying the power of effective social services in a rigorous manner is a very powerful tool that will ultimately help our country understand the amazing return on investments in human capital, versus those in war and prisons – and over-priced healthcare. Please try and understand the big picture.

  • SophieB

    No one is arguing the need for measurement and financial impact. I think the point is that it is not necessary to pay all of the additional costs that go along with SIBs (intermediaries, interest, and high levels of outside evaluation) to achieve the same thing. Looking at the big picture it seems that if governments were to actually pay nonprofits for their actual costs to provide a service, including the costs to measure it appropriately, they would likely achieve better outcomes for a lot less. Besides, as the article points out, while these are touted as preventing governments, and therefore taxpayers, from taking risks, then what is the point of using them with evidence-based programs? Perhaps they could be helpful if there use was limited to testing actual new and innovative approaches, but that is not what is happening. Partly this is because you wouldn’t test a new approach at the level of scale a SIB needs to be in order to generate the supposed savings everyone is promising, but has yet to be seen.

    The Peterborough example that is the supposed gold standard has not met its goal, which was only a 10% reduction in recidivism. To reach this outcome, there would have to be roughly 15 fewer repeat offenders per 100. At this rate, in order to result in any significant savings the number of repeat offenders must be reduced enough to allow entire sections of prisons to be shut down. There is very little cost savings until then since the majority of the costs are in staffing and facilities. And, don’t forget, the costs associated with a SIB are much higher. There is no reason that the same thing could not have been accomplished by the government investing directly in the additional services being provided for a lot less money. And, I don’t know that anyone has considered the costs of job losses, not only because of a lower prison population, but also as a result of the many small nonprofits who are closing in England due to this approach.

    In this country, there are other issues to consider as well…Often any savings that may be generated would come from a system other than the one providing the services. One government agency is not going to turn this over to another. Then there is the issue of a state getting less federal funding if they don’t spend all of their prior allocation. So if they save money one year, they get less the next, so it is not like there is going to be savings that can eventually be reinvested in the system. Then there is the reality that even at a 10% reduction in service needs per year or two, there is going to be a plateau. When this happens, there will be less money than there was originally to deal with ongoing needs.

    I don’t think that Mr. Cohen is overstating the case at all in his cautions about SIBs. He is one of the few people who is actually looking at them realistically–something that is greatly needed with the overwhelming number of enthusiastic supporters who generally fail to even portray them accurately. When all is said and done, at best SIBs appear to be nothing more than very expensive loans to the government–not in a taxpayers best interest.

  • David Lynn, San Diego Grantmakers

    While Rick raises some commonly discussed concerns and I wouldn’t in any way say PFS is the be-all-and-end-all of social solutions, I’ve got some significant issues with Rick’s positions here:

    1) A leader of one of our most widespread philanthropic publications is saying “Don’t invest if there’s not enough evidence.” That’s exactly what many of us fight against every day in philanthropy: we are the risk capital. Not all grants have to be completely evidence based first, that’s what experiments are for. And that’s exactly where philanthropy and private money can step in before risking taxpayer funds. Sure, maybe there’s some hype here, but saying we shouldn’t experiment because it’s not proven is completely contrary to the philanthropic movement.

    2) We’d all love to live in a society where government never ran deficits, was always efficient, and voters always approved tax increases when needed. Oh, wait, then we wouldn’t need social service nonprofits at all because government would always take care of those needs! And criticizing bond structures because private investors shouldn’t get paid for social solutions means we should no longer have school, water, utility, or redevelopment bonds because then Wall Street gets paid for investing in social infrastructure. Wall Street is going to get paid regardless of what we do, so we might as well try to put in structures that connect payments to results.

    There’s a big difference between focusing on a needed solution and finding a method to pay for it, and just starting with a financing method like PFS and then looking for a way to apply it. The hype may have gotten too many people going after PFS-or-bust, but that doesn’t mean PFS is not one of the viable methods to fund social solutions.

    David Lynn, San Diego Grantmakers

  • Brian O’Shaughnessy,

    SIBs are wildly inefficient as a “transaction.” I have personal experience with this problem. I have learned that they are fought from beginning to end by government at both the legislative and administrative level, and that means they have hit a nerve. The “nerve” is that we should fund services based upon a new criteria, and that is a wildly and deliciously disruptive concept. Are things really so good that you want to fight for the status quo?

    Again, SIBs have value because they put pressure on government to fund preventive services and to stop wasting money. As state after state rejects criminal justice reentry SIBs, some states have reduced recidivism 10-20%. Why?

    Seriously, you don’t know?

  • Lauren Grattan

    It’s true, the SIB model has many structural questions that need to be answered and has the added challenge of navigating yet more conflicting interests than traditional nonprofit organizations and philanthropy have.

    However, what’s also true is that as a sector we continue to beat our heads against the wall wondering how to grow philanthropic support beyond 6% of GDP. It’s not going to be JUST through the next ice bucket esque fad, better nonprofit storytelling, and data-driven decisions (though as a fundraiser, I wish that were so!)–it’s going to come from system change and motivating people who control resources in different ways than we do now.

    Reinforcing the fact that it’s not strictly proven and that foundations are tiptoeing into SIBs shouldn’t be an attack on the private sector, but rather another case in point that American organized philanthropy is cripplingly risk-averse and the vicious circle continues.

    I agree that campaigning politicians hopping on the bandwagon is a clear sign we should look more closely at SIBs. I don’t believe though that we should throw them out summarily, but instead ensure they’re structured for both programmatic success and to instill wider trust in the grander social change enterprise.

    This model very well might not work, and Rick’s are many of the questions we should be asking, but there is still great opportunity to motivate some of the other 94% of GDP to our work–permanently.

  • SophieB

    Please don’t misunderstand. I am in no way fighting for the status quo. There is absolutely no question that we need some serious systems change in the way nonprofits are funded. I will be the first one to say that the current system is wasteful and ineffective.I spend every day bringing governments and human services nonprofits together to talk with each other about how they can work together to make some of the many changes that are needed to eliminate waste and support better outcomes. But, that doesn’t mean jumping on the latest and greatest bandwagon to support something that is much more expensive without showing results is a good idea. I’ve seen these kinds of “sure fire new and improved” funding methodologies destroy too many needed nonprofits.

    If SIBs have hit a nerve that suggests the need for changing funding criteria then that’s great. But, let’s have that conversation instead of running full steam ahead with SIBs.