Can Pay for Success succeed in early childhood development?

A substantial body of evidence shows a high public return on investment to early childhood development (ECD) programs targeted to vulnerable children and families. Long-term studies demonstrate that these programs can return between 7 percent and as high as 20 percent annually for decades, with the majority of benefits accruing to the public, in forms such as government cost savings due to reductions in remedial education and crime.

Based on this research, some entities have recently begun using funding mechanisms known as Pay for Success (PFS) contracts and social impact bonds¹  to pay for ECD programs. These mechanisms—which have been used for a variety of issues around the globe, not just ECD—provide funds in advance to cover the costs of programming and services and then pay back investors with related public savings. More than 100 PFS- or social impact bond-funded projects have launched to date since 2010, when a social impact bond to reduce prison recidivism started in the U.K.

While the concept may be simple, there are complexities to navigate and trade-offs to consider in putting these deals together. There are also multiple steps involved, including developing the legal contracts among investors, service providers, evaluators, and government; providing services; evaluating outcomes; and potentially making payments to investors.

A number of ECD-related programs dovetail with the PFS approach, as many focus on prevention through nutrition, health, and early learning programs that can save government money. To date, PFS funding of ECD is quite small relative to other funding sources, such as government, philanthropy, and tuition payments by parents, and is likely not on a path toward usurping their role. Nevertheless, PFS is demonstrating that despite complexity, it can help expand proven models, test promising ones, and inform policymakers about making investments in ECD.

This article describes PFS and social impact bonds, the different components required to make a deal, and examples of ECD-related projects in the U.S. that have adopted PFS. The approach has expanded the reach of family home visiting and preschool in a state and two urban areas, respectively, while several other locations are studying the feasibility of using PFS for ECD.

An innovation in government contracting

The concept of government paying providers after they fulfill the terms of contracts is certainly nothing new or foreign; after all, government contracts for goods are often designed to make full payment after the goods are received. If the supplies, vehicles, food, etc., are not delivered, government doesn’t pay for them.

While the concept may be simple, there are complexities to navigate and trade-offs to consider in putting these deals together. There are also multiple steps involved, including developing the legal contracts among investors, service providers, evaluators, and government; providing services; evaluating outcomes; and potentially making payments to investors.

The innovative aspect of the PFS approach lies in applying outcomes-based procurement contracts to areas of government that typically pay for outputs. Examples of PFS projects to date include preschool programs, maternal and child home visiting, and interventions to reduce prison recidivism. In these types of education and social services, government would typically contract for services provided, such as the number of children educated, families visited, or former inmates treated. In PFS projects, government instead pays for specific outcomes, such as special education avoidance, positive birth outcomes, or reductions in recidivism.

Service providers involved in PFS projects can draw from different financial sources to fund their operations, such as receiving a grant from a foundation, tapping into their own revenue streams, or securing a loan. However, the latter two options pose financial risks.² 

Enter the social impact bond, or SIB. This particular type of PFS arrangement shifts financial risk to outside investors. In a SIB-funded project, investors provide operating funds for delivering services and are paid back by government based on pre-agreed metrics. In some SIBs, investors are paid back with a lower or higher return based on what benchmarks the outcomes reach. And if threshold metrics are not met, government may not make any payment. As discussed later on, depending on how the deals are structured, investors are at risk for losing part or all of their initial investment.

Necessary ingredients

In addition to a government jurisdiction, a service provider, and an investor, a PFS project typically needs an intermediary and an evaluator. (For more on how PFS deals are structured, see the figure below.) An intermediary primarily works across sectors to coordinate transactions among investors, service providers, and government. An intermediary can also recruit investors and philanthropists to fund the project and provide high-level management or oversight of the PFS process.

View full size

After services are provided, an evaluator steps in to validate results based on the pre-agreed metrics and research design. Outcome data by themselves are often not sufficient to validate results. Rather, a comparison or counterfactual is needed to answer this question: If not for the provision of these services, what would be the condition of the service population and related costs to government? One evaluation strategy is to conduct a randomized controlled trial (RCT)—that is, to randomly assign eligible people to either receive services or not receive them. This allows an evaluator to compare data for recipients and non-recipients and attribute the differences between them to the services. Aside from RCTs, projects have also used quasi-experimental evaluation designs that employ matched samples or regression discontinuity³  to evaluate outcomes.

Getting a PFS project off the ground

The first step in a PFS project is identifying a service that results in government savings. For example, a body of research on a service model may demonstrate positive effects on program recipients and reductions in government costs. This research would provide a basis for implementing a PFS project to scale the model and serve more people or to introduce the model to a new target population or service-delivery setting. Alternatively, a model with initial promising outcomes could use PFS to demonstrate its effects. To date, the majority of PFS projects are focused on scaling proven models.⁴ 

Once a service model and location are identified, a researcher conducts a feasibility study to determine the cost of the project, number of eligible people to serve, and whether service outcomes lead to measurable government savings. Government, investors, and service providers use the feasibility study to negotiate costs, timelines, and outcome metrics and related investment returns. The parties then use legal services to draft contracts.

Incentives for investors and philanthropists

SIBs are in a position to attract impact investors, who are individuals or entities seeking investments that generate both financial returns and social or environmental benefits. Impact investors currently fund projects in a variety of sectors, such as clean energy, education, and health. Based on a survey of leading impact investors, there are about $50 billion of assets under their management in the U.S. and Canada.⁵  Tapping into even a small portion of these assets could substantially increase the number and value of SIBs.

ECD PFS project example

Utah High Quality Preschool Program

In 2013, a SIB was launched in Salt Lake County, Utah, to provide preschool to children from low-income communities. Evaluation of the preschool model was based on studies conducted between 2006 and 2012. While not based on experimental data, the studies provided evidence that preschool could reduce expected special education placements. The payers include Salt Lake County and United Way of Salt Lake for the first year of the project and the State of Utah for subsequent years. Investors include Goldman Sachs as the senior investor at $4.6 million and the J.B. and M.K. Pritzker Family Foundation as the subordinate investor at $2.4 million.^*  This means Pritzker takes first losses if metrics are not met, thus reducing some of the risk to private investors. The United Way of Salt Lake serves as the intermediary.

Half-day preschool services are provided over five years to 3,500 children using the Utah High Quality Prekindergarten Program model at a variety of program types. Repayment is based on special education avoidance over a 12-year period. The senior investor is paid first, then payments are made to the subordinate investor. Full repayment of principal is expected if the program demonstrates about 90 percent avoidance of special education for students considered at-risk for needing special education. After the first year, an evaluation by Utah State University showed that of the 110 students identified as at-risk, only 1 used special education services in kindergarten. As a result, a payment was made to Goldman Sachs.^** 

Notes

* Utah High Quality Preschool Program, Pay for Success Learning Hub. Retrieved [March 1, 2018] from payforsuccess.org/project/utah-high-quality-preschool-program.

** Initial Results for Utah High Quality Preschool Program Show Success, Pay for Success Learning Hub. Retrieved [March 1, 2018] from payforsuccess.org/resource/initial-results-utah-high-quality-preschool-program-show-success.

Financial institutions regulated under the Community Reinvestment Act (CRA) may be interested in participating in SIBs because the investments could support their CRA-related activity. The CRA creates an evaluation and rating process that encourages financial institutions to meet the credit needs of their geographic service area, particularly low- and moderate-income (LMI) neighborhoods. A loan or in-kind contribution that supports an ECD program serving LMI children and families could be added to a financial institution’s CRA portfolio.⁶ 

Meanwhile, philanthropists and foundations have incentives to participate in SIBs as a means of leveraging financial resources. When a foundation awards a grant, it typically isn’t repaid; in effect, the foundation pays dollar-for-dollar for services provided. In contrast, with a SIB, there is a prospect of repayment and a return on investment if success metrics are met, and those funds can then be applied to other work that furthers the foundation’s mission.

In addition, philanthropy can play a role in recruiting private investors by taking the subordinate position on an investment⁷  or providing credit enhancement, such as a partial or full guarantee of principal repayment to investors.

By reducing risk to private investors, philanthropy can draw more private impact investing resources into SIBs. Again, compared with a typical grant, this approach can leverage more resources and serve more people. To date, impact investors and financial institutions have come forward to invest in SIBs, but have typically done so only when a foundation or philanthropist takes a position that reduces investor risk.

Finally, philanthropy can use SIBs as a method to test service models under development. The evaluation framework for determining payments can often serve to provide insights into the efficacy of a model. However, testing an in-development model via a PFS project likely carries higher risk than replicating a proven model.

Resources and technical assistance

There are a number of moving parts in PFS that work across sectors and require expertise in areas as wide-ranging as law, government and private sector finance, service provision, and evaluation. Fortunately, those interested in investigating the suitability of PFS have resources to turn to.

A number of organizations provide assistance, research, or general information on PFS. For example, three universities specialize in providing technical assistance to support PFS projects. The Harvard Kennedy School Government Performance Lab has provided technical assistance to more than half of the 20 PFS projects launched in the U.S.⁸  The Sorenson Impact Center at the University of Utah’s David Eccles School of Business provides assistance through a competitive grant process to nonprofits and government agencies that are assessing the feasibility of PFS.⁹  Meanwhile, the University of Virginia Pay for Success Lab uses teams of students to research social issues and identify locations that align with the PFS model.¹⁰ 

Outside of academia, one example of a PFS resource is the Nonprofit Finance Fund’s Pay for Success Learning Hub, which includes a map of PFS activity across the country, tool kits, and reports on projects under way.¹¹  The Federal Reserve System also features PFS and SIBs in a number of publications.¹²

ECD PFS project example

Chicago Child-Parent Centers SIB

The Chicago Child-Parent Center model, which provides half-day preschool for 3- and 4-year-old children, is the basis for a SIB that rolled out in 2014. The model was previously studied with a quasi-experimental research design that showed positive short- and long-term child outcomes associated with reductions in government costs. Government partners include the City of Chicago and the Chicago Board of Education, while investors include Goldman Sachs, Northern Trust (a financial services company headquartered in Chicago), and the J.B. and M.K. Pritzker Family Foundation. IFF, a Chicago-based community development financial institutionⁱ , serves as the intermediary.

The private investors contributed $12.66 million as the senior investor, while the Pritzker Foundation contributed $4 million in the subordinate position.ⁱⁱ 

Payment metrics include kindergarten readiness, avoided use of special education, and third grade literacy. The service period lasts for 4 years while the repayment period covers 17 years. The first evaluation by SRI International found positive effects on school readiness and special education avoidance, which led to payments to investors.ⁱⁱⁱ 

Notes

i A community development financial institution, or CDFI, is a specialized entity that provides loans, investments, training, or other services in underserved or economically distressed areas. For more on CDFIs, see our CDFI Resources page.

ii Chicago Child-Parent Center Pay for Success Initiative, Pay for Success Learning Hub. Retrieved [February 28, 2018] from payforsuccess.org/project/chicago-child-parent-center-pay-success-initiative.

iii Erika Gaylor, Traci Kutaka, Kate Ferguson, Cyndi Williamson, Xin Wei, and Donna Spiker. Evaluation of Kindergarten Readiness in Five Child-Parent Centers: Report for 2014–15. SRI International, April 2016.

In addition, a new government funding stream will serve as a resource to fund PFS projects. The recent federal budget bill passed on February 9 allocated $100 million for the Social Impact Partnerships to Pay for Results Act, which will award money to PFS projects that result in social benefit and federal, state, or local government savings.

Benefits and challenges

PFS and SIBs present a number of benefits and challenges. On the benefit side, SIBs tap into impact investment funding to help scale proven models and test promising approaches. Because PFS has an evaluation feedback loop, when results don’t hit the mark, operators have a mechanism to discontinue services or make informed adjustments to the model.

In addition, a focus on outcomes can lead to innovation in the field, as service providers look for ways to more effectively achieve outcomes. Finally, as mentioned earlier, through SIBs, government and service providers avoid financial risk, philanthropists can increase leverage of their financial resources, and investors can accrue both financial and social impact.

On the flip side are the challenges of PFS and SIBs. First, costs associated with drafting contracts, project management, and evaluation can add up. This means government savings and project size need to be large enough to absorb the transaction costs of implementing PFS. However, as more PFS projects are crafted and launched, subsequent projects may be able to reduce costs by adopting contract templates and processes from earlier projects.

Other potential challenges include implementing the process for government to set aside funds to pay investors in the future for successful outcomes, which requires a mechanism and political will. Some government jurisdictions also have challenges committing to a long-term project and allocating funds beyond a one- or two-year budget cycle. In addition, benefits might be distributed across several government jurisdictions and departments, creating what’s referred to as the “wrong pockets” problem. That is, the government agency making payments for project outcomes may not be the one receiving the primary benefits.

This mix of possibilities and limitations has led to a sector that is growing, but is far short of serving as a primary funding source for social services. The 20 projects under way in the U.S. total almost $200 million in funding from private and philanthropic investors.

Given the benefits and challenges discussed here, there are some conditions that are consistent with achieving a successful roll-out of PFS and SIBs. First, the service model delivers outcomes that are well-defined and relatively simple to measure. Second, government savings are easy to identify within one or just a few government jurisdictions or departments. Third, benefits accrue in a relatively short time horizon, ideally within a year or two after providing services. Fourth, the project has strong leadership to keep a number of moving parts on track.

Future of ECD and PFS

When the PFS framework is applied to ECD, a number of advantages and some constraints emerge. Long-term research studies of preschool PFS projects in Utah and Chicago and a maternal and child home-visiting PFS project in South Carolina have demonstrated cost savings to government. (See the “PFS ECD Project Example” sidebars at left for details.) However, a number of benefits to ECD programs accrue years after children and families receive services. In the preschool studies, benefits such as larger tax payments and crime reductions occur 15 years or more down the road, outside the typical time frame to repay private investors. Yet in the preschool PFS projects, participants have focused on relatively short-term cost savings, such as avoiding special education placements in elementary school. In the future, research that can link additional short-term outcomes to long-term cost savings could expand the number of indicators for which government would agree to make PFS payments.

In the home-visiting research, long-term studies have established that child benefits can last more than 10 to 15 years after the intervention, but the SIB in South Carolina is basing repayment on more near-term benefits, such as healthier birth outcomes, reduced child emergency room visits, and healthy spacing between births.

In addition to being long-term in nature, benefits from ECD programs tend to spread across a number of government jurisdictions and agencies, making it difficult to fully capture savings and identify which parts of government should repay investors. While focusing on short-term outcomes, current ECD PFS projects also have limited the number of government jurisdictions or agencies that participate as payers.

Government jurisdictions are investigating how PFS might apply to other ECD-related social service, health, and education investments. For example, interventions that prevent the need for child protection services or foster care have potential to save government money. In addition, effective case management, coupled with home remediation that reduces environmental risks for childhood asthma, can reduce asthma-related visits to the emergency room.

PFS activity continues to expand in the U.S. and other countries, which will help shrink the learning curve for new entrants. But the future may have as much to do with how policymakers and the public respond to PFS project outcomes as with how participants clear the logistical hurdles of putting the projects together. For example, a series of successful outcomes by a service provider may embolden the public and policymakers to simply fund the services with tax revenue instead of seeking outside investors. This would allow the public to retain all of the benefits instead of sharing government savings with banks and Wall Street firms. Some would argue such a result would be a sign of a successful PFS project.

While PFS and SIBs may not be a panacea for solving the list of challenges facing LMI children and families, the innovation and knowledge-building spawned by these mechanisms serves as a piece of the puzzle of improving outcomes for children, families, and society.

ECD PFS project example

South Carolina Nurse-Family Partnership

In 2016, a social impact bond (SIB) launched in South Carolina to provide home visiting through the Nurse Family Partnership (NFP) model to 3,200 mothers and their children over the course of six years. Nurses regularly visit families that have young, first-time, expectant mothers, starting early in pregnancy through the child’s second birthday. They provide advice on a number of topics, such as health and nutrition, safely caring for the child, and planning for a secure future.^a  The government payer is the South Carolina Department of Health and Human Services. Unlike the Utah and Chicago projects, the investors don’t include financial firms, but rather a number of foundations providing a total of $12.2 million, private funders providing a total of $4 million, and The Boeing Company providing $0.8 million. Of the foundations, the largest contribution, $8 million, came from The Duke Endowment. Social Finance US serves as the intermediary.

The NFP has been evaluated three times since the late 1970s through randomized controlled trials; results included improvements in maternal health, reductions in child emergency room visits, and increased spacing between subsequent births. Currently, under the SIB, repayment is based on some similar outcomes: reduction in preterm births, reduction in childhood hospitalization and emergency department use due to injury, increase in spacing between births, and increase in number of first-time moms served in high-poverty ZIP Codes. Instead of being repaid to investors, if outcomes are achieved, success payments will be reinvested to extend NFP service delivery in South Carolina. J-PAL North America, based at the Massachusetts Institute of Technology, is the evaluator.^b 

Notes

a Nurse-Family Partnership. Retrieved [April 4, 2018] from nursefamilypartnership.org/about.

b Nonprofit Finance Fund. Pay for Success: The First Generation. A Comparative Analysis of the First 10 Pay for Success Projects in the United States, April 2016.

Endnotes

1 For more on SIBs, see our July 2013 Community Dividend feature, “Social impact bonds offer promise of savings, profits, and positive outcomes.”

2 Luther Ragin, Jr. and Tracy Palandjian. “Social Impact Bonds: Using Impact Investment to Expand Effective Social Programs.” Community Development Investment Review Vol. 9, No. 1. Federal Reserve Bank of San Francisco, April 2013.

3 Regression discontinuity is a research method that has been used to evaluate social programs in which eligibility is determined by a threshold, such as age of child by a certain date or a specific level of household income. By comparing subjects just above or below the threshold, the research design can account for unobserved differences between the treatment and comparison groups.

4 Nonprofit Finance Fund. Pay for Success: The First Generation. A Comparative Analysis of the First 10 Pay for Success Projects in the United States, April 2016.

5 Global Impact Investing Network. Highlights from the 2017 Annual Impact Investor Survey.

6 In addition to evaluating CRA considerations, financial institutions should follow their established practices to ensure that such investments are lawful under and comply with applicable laws and regulations, fulfill their investment strategies, and are consistent with their risk-management approaches.

7 A subordinate investor sustains losses ahead of the senior investor if metrics are not met.

8 Harvard Kennedy School Government Performance Lab. Retrieved [March 2, 2018] from govlab.hks.harvard.edu/gov-lab.

9 Sorenson Impact Center, sorensonimpact.com.

10 Jennifer Giovannitti and Joshua Ogburn. “Growing the Pipeline of Pay-for-Success Projects.” Community Practice Papers. Federal Reserve Bank of Richmond, February 2018.

11 To access the hub, visit www.payforsuccess.org.

12 To access these publications, visit the Federal Reserve System’s community development resource portal at FedCommunities.org and enter “Pay for Success” in the search box.