Pay for success (PFS) projects offer governments opportunities to invest in outcomes and employ new capital to meet the needs of their communities. But PFS projects also carry risks. For investors, the risks relate to the project failing to meet its outcomes or the government reneging on its commitment to pay. Investors’ perceptions of risk matter. Projects with high or unclear risk may discourage investors and prevent the project from launching. This brief helps project partners understand the risks investors perceive when entering PFS contracts and familiarize themselves with measures that have been used or proposed to manage this risk.
Pay for success (PFS) projects offer governments opportunities to invest in outcomes and employ new capital to meet the needs of their communities.1 Governments, social service providers, intermediaries, philanthropies, private investors, and other PFS stakeholders invest time and capital resources through this model to ensure people receive the services that can help them improve their lives.
But PFS projects also carry risks. For investors, the risks primarily relate to the project failing to meet its outcomes or the government reneging on its commitment to pay. Investors’ perceptions of risk matter. Projects with high or unclear risk profiles may, without commensurate financial or social returns, discourage investors and prevent the project from launching. Successfully structured projects balance investor rate of return and risk profile. Because governments are often unable (because of budget constraints) or unwilling to offer greater returns,2 alternative measures, such as project design solutions, evaluation designs, and financing structures and mechanisms, can manage or mitigate risk.
This brief is designed to help project partners, particularly governments, both understand the primary project-related risks investors are likely to perceive when entering PFS contracts and familiarize themselves with measures that have been used or proposed to manage this risk. This information will help governments understand the available tools to help manage risk, attract investors, and launch PFS projects.
Risk to Investors in Pay for Success
All investments carry risk. In traditional investments, the higher the risk, the higher the expected return. In PFS, stakeholders’ perceptions of risk can depart from this typical relationship for several reasons:
- PFS is new, and the ability to accurately estimate and price the risk of achieving social outcomes is challenging.
- Governments may be unwilling or unable to offer returns fully commensurate to the perceived risk.
- Financial incentives are not the only motivation for investors.
These characteristics have implications for governments’ abilities to assemble funding for a PFS project and for structuring repayment. Different investors have different appetites for risk and requirements for receiving a return on their investment in PFS. For example, some philanthropic investors have provided funding without any expectation of receiving a return on investment, while private investors have expected returns. Some philanthropic investors’ willingness to assume significant risk or potentially lower returns to draw other investors and launch a project they consider worthwhile and aligned with their values helps explain their participation in PFS projects (Rangan and Chase 2015).
Though other types of risk exist in PFS projects, this brief focuses on two categories that most directly affect PFS investors: performance risks and appropriations risks.
Performance Risks and Project Design Solutions
Performance risks are the chance that the project will fall short of its intended outcomes, leading to an investor failing to receive a return and losing all or part of its principal. There are several types of performance risks:
- Programmatic risk: Risk that the program does not work. This risk can decrease by selecting a program that clearly articulates how it intends to yield the desired outcomes through its activities (called a theory of change). For PFS projects that replicate an existing program, a strong theory of change should be supplemented by evidence of effectiveness (Milner and Eldridge 2016).
- Implementation or operations risk: Risk that the project fails because the program delivery isn’t executed in a manner consistent with what the design and evidence say works, often because of a service provider’s insufficient operational capacity. For example, providers could face unexpected challenges with participant recruitment and enrollment or staff hiring, or other public support that participants depend on could be reduced or withdrawn. This risk can be diminished by implementing strong performance measurement and management controls, assessing capacity and resilience of the overall public supports system, and reviewing provider capacities before project launch.
- Evaluation risk: Risk that the program’s evaluation fails to accurately measure whether outcomes have been achieved (e.g., a false positive or negative). This could be caused by poor evaluation design or implementation. Carefully selecting a rigorous evaluation design, ensuring access to necessary data, and enlisting a qualified program evaluator can reduce this risk.
- Regulatory or policy risk: Risk that new legislation or policies will change the composition of the target population or otherwise undermine service delivery. For example, raising the age at which juveniles can be tried as adults could change the eligibility criteria for a juvenile justice intervention. These risks may be unavoidable, but knowing the political landscape or anticipating potential policy changes could help stakeholders plan accordingly.
- Nonperformance or partnership risk: Risk that the project will end before its scheduled date because one or more actors in the deal do not fulfill contractual obligations. Even when employing best practices for participant recruitment, for example, a service provider may not be able to enroll a sufficient number of participants, thereby violating contracts with minimum recruitment thresholds. Understanding the capacity of fellow project partners, outlining contingency plans for certain scenarios (e.g., enrollment challenges), and tracking enrollment data to continuously improve processes can reduce this risk (Gatti 2008, 49).
In addition to the financial implications of these performance risks, investors and other project partners should anticipate the potential reputational risks of project nonperformance because the PFS field is still nascent. Conducting due diligence and implementing the project design measures described above will mitigate this risk. Other factors could have reputational implications. For example, the terms of the deal and their measures to reduce private investors’ financial risk could encourage criticism of all parties by skeptical politicians or members of the public.
Strong program and project design can help manage performance risk. Grants provided by philanthropies for PFS project development, often for feasibility studies, help advance this goal. Prospective projects have received grants from the Corporation of National and Community Service and other philanthropic investors to carry out a feasibility study assessing whether a PFS project would be possible and appropriate for a problem in that jurisdiction.4 The Urban Institute’s Pay for Success Initiative also supports the design of PFS projects through technical support including identifying evidence-based interventions, designing evaluations, and pricing outcomes (PFSI 2015).
Other philanthropic and government grants have supported general project development by funding a pilot phase to reduce implementation risk or by paying for a project cost to reduce the project budget. The Massachusetts Chronic Homelessness Pay for Success Initiative used a $1 million grant from Santander Bank and United Way to help it launch (table 2). The South Carolina Nurse-Family Partnership Pay for Success project incorporated a three-month pilot period into its project timeline. The pilot, funded by the Laura and John Arnold Foundation and the Duke Endowment, allowed project partners to identify key best practices before launching services.5
Financial Tools to Mitigate Investor Risk
Even when PFS stakeholders have minimized the performance risks above, projects still carry some risk that the outcomes will not be met. To make PFS more attractive to investors, several financial tools have been used to get deals off the ground.
Below, we describe tools that governments and other PFS stakeholders may wish to explore with investors, particularly local and national philanthropies, as they create the capital stack for a deal. Measures to reduce investor risk include junior loans, grants, and guarantees, all to lower the risk profile or partially insure the senior investor’s principal.
Up to this point, the risk associated with PFS projects has not matched the return profile directly. Credit enhancements and other measures can bring new dollars to projects by more appropriately accounting for risk (Godeke and Resner 2012; Rangan and Chase 2015).
The current techniques to reduce risk have been made possible by several philanthropic investors who have been willing to accept greater risk, lower returns, or even no financial returns (through grantmaking) to advance a project aligned with their goals.
Using financial tools to mitigate the risk to the senior investor offers benefits to several stakeholders:
- Governments: Improves the likelihood of attracting private funding for the project by making it a more attractive investment option
- Private investors: Helps offset the project’s risk-return profile by buffering investors against performance risk, depending on the type of financial structure
- Philanthropic investors: Provides opportunity to leverage their investment for maximum impact
- If the PFS project is successful, philanthropies will have funded a program consistent with their mission and can receive modest interest, or, in the case of guarantees, have catalyzed the deal and lost nothing. Table 1 shows three primary types of direct financial structures: junior loans, grants, and guarantees.
With junior loans (i.e., subordinate debt), at least two investors—one senior and one junior—invest in a project. If the project meets its targets, both lenders will be repaid according to the terms of the contract, with the senior lender receiving repayment first. In traditional junior loans, the junior lender often expects a higher potential return than the senior lender to compensate for the higher risk of being the second lender repaid. However, in PFS, the junior lenders have primarily been philanthropic actors who have accepted a lower rate of interest to help push the deal forward.6
Junior loans buffer senior loans against loss of principal if the project misses its outcome targets, but the contract permits partial repayment. In this scenario, the senior investor could be repaid first, with the junior investor taking most or all the loss.
Junior investment could also boost the effective rate of return the government can offer the senior investor. Consider a project that requires $10 million in funding with a government able to offer a 5 percent potential return. If private investors calculate that an 8 percent rate of return is the minimum threshold to make the investment attractive, a philanthropy may be willing to offer a junior loan of $5 million with 2 percent potential return, enabling the government to offer the 8 percent return to a $5 million senior loan. In the end, the government does not need to offer a higher return, private investors are attracted to the deal, and the philanthropy can support a project it values.
Grants are the most straightforward form of project support from nonprincipal investors. Grants are typically given as gifts meant to advance or achieve an outcome. In PFS, grants can be offered by philanthropy to decrease the overall project financing that needs to be repaid to investors. In addition, some grants have recoverable components. If the project meets its target outcomes, the grant is repaid to the investor (albeit without a return). In other cases, the grant is nonrecoverable and is simply used to catalyze the project.
In some projects, such as the Massachusetts Juvenile Justice Project, various grantmakers (the Laura and John Arnold Foundation, New Profit, and the Boston Foundation) provide funding to attract other investment without expecting repayment (table 2).7 In other projects, the funds have been directed to a specific end, such as the Finnegan Family Foundation’s support for the evaluation costs of the Chicago Child-Parent Center project.8 In both examples, philanthropic partners provided funding, effectively reducing the amount the government has to repay for outcomes.
In a PFS project, if the program’s outcomes fall short of the targets, the investors receive no return and lose their principal. However a “guarantee for performance” credit enhancement ensures that investors cannot lose all of their principal investment.9 Instead, in the event of underperformance, the guarantor (typically a philanthropy) secures the investor against the loss of this principal up to a specified percentage. Figure 1 displays the function of a guarantee. By reducing the amount of risk for the investor, the guarantee can make a project more appealing.
So far, guarantees have been used in two US PFS projects. The ABLE project at Rikers Island was guaranteed by Bloomberg Philanthropies for $7.2 million (table 2), which provided a 75 percent guarantee of the total investment (NYC Mayor 2012). When the project’s preliminary evaluation indicated it was unsuccessful, the guarantee buffered the senior lender’s loss. Additionally, the Rockefeller Foundation is guaranteeing about 10 percent of the up-front capital in the Recidivism and Workforce Development project in New York State.10
Other Approaches to Manage Investor Risk
Funding Specific Outcomes
One strategy for mitigating performance risk is to redistribute the risk among multiple investors. If a project expects to affect multiple outcomes (e.g., reduce preterm births and increase healthy spacing between births), investors can select the outcome in which they would like to invest. In this way, an investor who is more risk averse could invest in the outcome with a stronger history of consistent results based on previous evaluations. Then, a philanthropy or other more risk-tolerant investors could support the project by selecting the outcome with less evaluation history or less consistent results.
By segmenting the outcomes based on evidence that the program will improve each outcome and by having different investors for each outcome, investors who are more risk averse can still participate by choosing to be repaid based on the outcome with a stronger history of evidence. Because different outcomes are measured and paid at different times, investors can choose outcomes that meet their risk profile and preferred repayment schedule. Philanthropic investors often have flexible capital and are willing to invest in outcomes that may take longer to achieve.
In the Denver Social Impact Bond Program, one set of investors is being repaid for housing stability, while another is being repaid for reduction in jail bed–days (Archer-Rosenthal 2016). The housing stability outcome will be eligible for repayment earlier than the outcome for jail bed–days. This model is unique to Denver’s project, but future projects may seek to replicate it, particularly if a government or investor wants to see if a program that yielded one type of outcome could yield a different one.
Government and philanthropy have provided other project resources that reduce the risk associated with programmatic activities. This support has included service delivery or third party–provided training and technical assistance. The Commonwealth of Massachusetts, for example, provided $14 million in housing vouchers and other resources in the Massachusetts Chronic Homelessness Pay for Success Initiative. This allocation of resources reduces the capital the project needs to raise. Thus, the principal investors had to risk $14 million less private capital.
Intermediaries and service providers have supported projects by allowing fees and other payments to be deferred or subordinated and repaid based on outcome achievement. The service provider Roca Inc. and intermediary Third Sector Capital Partners deferred a portion of their fees in the Massachusetts juvenile justice project, as did Abode Services, the service provider in the Santa Clara County project (Archer-Rosenthal 2016). Other organizations such as the Corporation for Supportive Housing and Harvard’s Government Performance Lab offer technical assistance and capacity-building support.
Governments and other stakeholders should be aware that these waived fees and other supports may become real program costs after the project ends if the government decides to keep or expand the program. These hidden costs not calculated in the project’s funding should be acknowledged as part of the project’s costs, particularly if they are costs that will need to be incurred again.
Appropriations risk, also called credit or counterparty risk, is the potential that the end payer (typically a government) will not repay the investor if the project meets its outcome targets. Changing local priorities could make the government unwilling or unable to repay the investor’s principal or provide additional returns. Because PFS contracts span several years, local governments’ potential to pay might be altered because of changing leadership, economic conditions, or budget priorities over multiple budget cycles.
Governments may reserve the right (through a “nonappropriations” clause) to halt or postpone payments if money is not allocated by the responsible legislative body (Forsyth and Marciano 2013). Though this risk is not new or specific to PFS, and governments have devised ways to manage this risk, governments may face additional challenges allocating funds for future spending and the model is new and unfamiliar. While no government has failed to meet a payment obligation, appropriations risk is remains a consideration in PFS projects (Gatti 2008, 57). Below, we discuss solutions that have been used or proposed to address this risk.
Tools to Mitigate Appropriations Risk
In PFS contracts, the end payers, which to date have been local, county, and state governments, must obligate future appropriations or employ a prepaid fund to finance repayment based on the outcome of the forthcoming evaluations. If the project meets its target outcomes but the government cannot pay or pay on time, investors could lose their principal (and forgo a return), even though this is contrary to the terms of the contract. This is often a challenge for governments because they cannot promise or obligate funds from future appropriations. At the same time, without confidence that the government will pay what they owe, a private investor is less likely to invest.
PFS practitioners have come up with several solutions to this problem. Strategies for mitigating appropriations risk have been varied and specific to local legal circumstances affecting each project. Local and state governments have employed various strategies to tackle this challenge, including the following:11
- Prepaid (or sinking) fund: This vehicle establishes an account that stores partial or whole repayment amounts for the PFS contract. This type of fund is usually established at the beginning of the contract, and the government allocates money for each of the years the contract will be in effect. The fund holds monies year to year for future payment. Funds could be allocated in 2016 for services provided in 2016, even if success payments are not determined until 2018. Sinking funds have been set up in several jurisdictions such as Chicago’s Child-Parent Center project (Goldberg 2012).12
- Full faith and credit of a government entity: This is the strongest option to mitigate appropriations risk. Massachusetts passed legislation that set up a Social Innovation Financing Trust Fund backed by the full faith and credit of the commonwealth.13 Funds are allocated before the project and their use is exclusively for success payments. This fund does not have to be the same amount as the maximum obligation of the end payer. It could hold a much larger amount intended to fund outcome payments for multiple projects. These funds may be invested for capital gains or be combined with private donations.
- Multiyear appropriation: While many governments have restrictions on committing to future contracts, unless the obligations are covered by a long-term bond authority some legislatures pass multiyear allocations. New York State’s prisoner reentry project utilized this approach to make multiyear rolling allocations into a contingency reserve.14
- Rating-agency trigger: Governments’ ability to raise credit depends in part on credit-rating agency grades. If a government’s ability to repay loans is deemed risky, their cost of capital (the interest creditors expect the government to pay) increases. The failure to make agreed payments as part of a PFS contract can be explicitly linked to credit-rating agency reports to provide a clear disincentive to governments for defaulting on outcome payment commitments. Chicago’s PFS project includes a provision stating that failure to pay upon outcomes achievement will be automatically reported to rating agencies.15
So far, these government efforts have managed perceptions of counterparty risk.16 However, other solutions may need to be developed, especially if smaller county and municipal governments with greater appropriations risks begin to consider PFS. Credit enhancements, as philanthropic guarantees17 that insure against government nonpayment, may be one future solution.18
Through PFS, governments can transfer financial risk for innovation to investors and only pay for results while investors pursue a financial and social bottom line. Although investor risk in these projects is not unique, the risk-return trade-off and implications for assembling capital are. In a traditional market, without quantifying the social value of the return, investors in projects with a PFS level of risk would expect returns that have not yet been seen in a PFS project. Because the model is new, PFS differs from more mature financial products. The risks of PFS are less well understood, and there are no established markets for managing risk. Nonetheless, several measures can be and have been used to manage or reduce risks.
Performance and appropriation risks—the two primary types of risk faced by investors—can be mitigated through a combination of project design (e.g., using evidence to inform design [Milner and Eldridge 2016]), policy measures (e.g., appropriating some money to a fund every year19), and credit enhancements (e.g., performance guarantees).
Philanthropic investors have played an important role in credit enhancements and have leveraged their funds to attract significant private investment in PFS’s early years. With a clearer view of risks perceived by investors and of the techniques and mechanisms used to reduce these risks, stakeholders are more knowledgeable about how to assemble financing for these projects.
- Pay for success is an innovative financing mechanism that shifts risk for a new, evidence-based social program from a traditional funder (usually a government) to a third-party funder (usually a private organization or nonprofit). PFS projects test whether a social program can improve outcomes for a specific group. If the program works (as measured by a rigorous evaluation) the project is a success. Investors get their money back (with a potential positive return), the government realizes potential future cost savings, families and society benefit from better outcomes, and social service providers strengthen the case for funding their model.↩
- In a social impact bond, “a government entity may only be willing to pay bond-like returns, altering the risk-reward ratio in ways that can limit a [social impact bond]’s viability” (Berlin 2016, 9).↩
- Because the number of PFS deals to date is so small, any relationship between the amount of risk and the type of credit enhancement is unclear. As the market grows, a relationship between the level of risk and the type of enhancement (or the need for a credit enhancement) may become clear.↩
- “SIF Pay for Success,” Corporation for National and Community Service, accessed August 4, 2016, http://www.nationalservice.gov/programs/social-innovation-fund/our-programs/pay-success.↩
- Partners from the South Carolina Nurse-Family Partnership Pay for Success Project Pilot expect to publish a brief on the results of the pilot program in September 2016.↩
- For example, in the Massachusetts Juvenile Justice PFS project, if the project meets its targets, Goldman Sachs (the senior investor) will receive a base annual return of 5 percent whereas the Kresge Foundation and Living Cities will receive a base annual return of 2 percent.↩
- “Juvenile Justice Pay for Success Initiative,” Urban Institute, Pay for Success Initiative, accessed August 4, 2016, http://pfs.urban.org/pfs-project-fact-sheets/content/juvenile-justice-pay-success-initiative.↩
- “Child-Parent Center Pay for Success Initiative,” Urban Institute, Pay for Success Initiative, accessed August 4, 2016, http://pfs.urban.org/pfs-project-fact-sheets/content/child-parent-center-pay-success-initiative.↩
- Although an imperfect comparison, letters of credit are similar to PFS performance guarantees. In mainstream financing, banks provide a letter of credit to a seller (or lender) guaranteeing that the value of its sale (or loan) will be repaid by the buyer (or borrower). The letter of credit helps a deal, sale, or loan go forward by allaying the seller or lender’s concerns of (re)payment.↩
- “Recidivism and Workforce Development Project,” Urban Institute, Pay for Success Initiative, accessed August 4, 2016, http://pfs.urban.org/pfs-project-fact-sheets/content/recidivism-and-workforce-development-project.↩
- For more information on these strategies, see Ellen Ward, “4 Ps of Pay for Success,” Living Cities, April 8, 2015, https://www.livingcities.org/blog/809-4-ps-of-pay-for-success-policy.↩
- “Child-Parent Center Pay for Success Initiative,” Urban Institute, Pay for Success Initiative, accessed August 4, 2016, http://pfs.urban.org/pfs-project-fact-sheets/content/child-parent-center-pay-success-initiative.↩
- Mass. Gen. Laws Ann. ch. 10. 35(vv), https://malegislature.gov/Laws/GeneralLaws/PartI/TitleII/Chapter10/Section35VV.↩
- “Budget Highlights,” Ny.gov, Pay for Success Contingency Reserve, accessed August 4, 2016, https://www.budget.ny.gov/pubs/executive/eBudget1617/agencyPresentations/appropData/PayForSuccessContingencyReserve.html.↩
- See article 7.2.iii in the October 2014 Chicago ordinance “Loan agreement and contract with IFF Pay for Success I LLC to serve at-risk children to increase school readiness and reduce later public school spending,” http://www.payforsuccess.org/sites/default/files/o2014-8677.pdf.↩
- It is unclear, however, whether unresolved appropriations risk has scuttled other potential PFS projects before they developed. ↩
- In finance, a similar credit enhancement tool exists (a surety bond, or a bank guarantee, if issued by a bank). Through a surety bond, a third party promises to pay the principal (e.g., lender) if the obligated party does not live up to its obligation.↩
- This presents a potential perverse incentive risk for governments, who may calculate that defaulting is financially preferable because investors who are paid through a guarantee will not pursue legal recourse.↩
- A sinking fund.↩
Archer-Rosenthal, Dana. 2016. Pay for Success: The First Generation. New York: Nonprofit Finance Fund.
Berlin, Gordon L. 2016. Learning from Experience: A Guide to Social Impact Bond Investing. New York: MDRC.
Forsythe, Amanda, and John Marciano. 2013. “Non-Appropriation Risk in Government Contracts.” Project Finance NewsWire (February): 40–41.
Gatti, Stefano. 2008. Project Finance in Theory and Practice: Designing, Structuring, and Financing Private and Public Projects. Boston: Academic Press.
Godeke, Steven, and Lyel Resner. 2008. Building a Healthy and Sustainable Social Impact Bond Market: The Investor Landscape. New York: Godeke Consulting.
Goldberg, Steve. 2012. “SIB Legislation in the US.” The Social Impact Bond Tribune 1: 8–16.
Milner, Justin, and Matthew Eldridge. 2016. “From Evidence to Outcomes: Using Evidence to Inform Pay for Success Project Design.” Washington, DC: Urban Institute.
NYC Mayor (Office of the Mayor of New York City). 2012. “Fact Sheet: The NYC ABLE Project for Incarcerated Youth.” New York: NYC Mayor.
PFSI (Pay for Success Initiative). 2015. “Notice of Support Availability: Training and Technical Assistance Services for Pay for Success Initiatives.” Washington, DC: Urban Institute, PFSI.
Rangan, V. Kasturi, and Lisa A. Chase. 2015. “The Payoff of Pay-for-Success.” Stanford Social Innovation Review 13 (4): 28–39.
About the Authors
Rebecca TeKolste is a former project associate with the Urban Institute’s Pay for Success Initiative. Before her time at Urban, she served as a Peace Corps volunteer in rural Guatemala’s Maternal and Child Health project. She is currently pursuing an MA in global affairs at Yale University's Jackson Institute for Global Affairs.
Matt Eldridge is research products manager of the Urban Institute’s Pay for Success Initiative. He is interested in how finance and economics, specifically public-sector finance, interact with public policy. Before joining Urban, he worked at the World Bank and as a consultant on financial services policy and regulatory issues. He earned his BA from Virginia Tech and his MS from the London School of Economics, both in international development.
Rayanne Hawkins is the business operations manager for the Urban Institute’s Pay for Success Initiative. She is interested in impact investing, data driven decisionmaking and other ways that business strategies can help solve social problems. In addition to research and project management for the Initiative, she runs the only pay for success training and technical assistance online Support Center which connects questions in the field with answers from Urban subject matter experts. She holds an MBA from Georgetown University and a BA in Political Science from The College of Wooster.
This brief was funded by the Laura and John Arnold Foundation. We are grateful to them and to all our funders, who make it possible for Urban to advance its mission.
The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, or its funders. Funders do not determine research findings or the insights and recommendations of Urban experts. Further information on the Urban Institute’s funding principles is available at www.urban.org/support.
The authors thank Allison Clark of the MacArthur Foundation, Casey Littlefield of Social Finance US, Ellen Clark of Living Cities, and Sted Garber of the Annie E. Casey Foundation for commenting on an earlier version of this brief.