Can impact bonds improve service delivery? (Brookings)

By Tamar Manuelyan Atinc, Emily Gustafsson-Wright and Sophie Gardiner

In mid-March, the United Kingdom launched five new social impact bonds (SIB), adding to the momentum that the field has witnessed in the last few years. As the number of SIBs around the world has grown to 43, they have become more diverse both in terms of countries and social issues targeted. In the most mature SIBs market, the U.K., impacts funds have even grown to cover multiple deals instead of just individual transactions. 

Here at the Center for Universal Education, our interest is specifically in exploring the potential for SIBs to be applied toward early childhood development (ECD) interventions, but potential ECD transactions have to be informed and grounded in an understanding of both the challenges and enabling factors in taking SIBs to the developing world. The full day of discussions on impact bonds that we organized at the recent LEGO Idea conference for some 75 participants interested and involved in the field helped crystallize some of the potential benefits and the needed adaptations.

What are impact bonds?

Briefly, an impact bond is an innovative financing mechanism that harnesses private capital to fund social services. In a social impact bond the government agrees to repay the investors only if agreed-upon outcomes are achieved. In a development impact bond (DIB), a third party, such as an aid agency or a foundation, is the “outcome funder” instead of the government. This mechanism has captured the imagination of many, and the funding of a rapidly expanding few, for several reasons. These reasons include: 1) a growing interest among investors—institutional and individual—for doing good while making a return on investment (so-called blended value or double bottom line); and 2) the outcome focus of impact bonds in fields where it has not always been possible to show results or value for money. Of note, while there is some work on impacts bonds in developing countries and several transactions in the making, so far there is only one that is concluded: a DIB for girls’ education in Rajasthan, India.

What challenges lie ahead for impact bonds?

From our conference discussions, four challenges emerged as worth considering for taking impact bonds to the development arena:

  1. Achieving scale will be significantly more difficult. The current global SIB market is composed of transactions with small sums of funding (the majority being under 5 million). Despite the small size, these deals may be addressing significant parts of the social agenda on which they choose to focus. This is because of the generally small populations that they target in relatively niche areas of concern in the developed world. While problems of homelessness and unemployment may be acute for some populations, it is possible to make progress on these problems in the richer parts of the world even with small amounts of funding. However, the scale, diversity, and overlapping challenges that populations face in the developing world suggest that impact bonds are unlikely to make a dent in addressing them with similar amounts. The value of impact bonds in the developing world will more likely be in their demonstration effect rather than scale. However, a SIB fund structure, like the one being developed for ECD outcomes in South Africa, may help catalyze larger SIBs in developing countries. In SIB funds, the government issues a “rate card” of outcomes and service providers bid against those prices for outcome-based contracts. The U.K. has launched three SIB funds so far, which include a combined 21 SIB contract structures.
  2. Government engagement upfront and throughout is critical to expand and sustain impact. Government ownership of the impact bond experiment and its associated outcomes is essential in order to lay a path for expanding the delivery model to populations beyond those initially targeted. When government is the outcome funder (as it is in a SIB), it has “skin in game.” This increases the likelihood of adopting at least aspects of the model more broadly if the transaction is successful. This is more questionable when the transaction is a DIB, which does not require the involvement of government. DIBs may be necessary in particular when government capacity is weak and willingness to go outside the traditional public sector funding and delivery models is limited. DIBs can be helpful in illustrating the potential of the mechanism but in the interest of expanding and sustaining impact, stakeholders in any DIB transaction would be well advised to enlist the engagement of the government right from the start. We may also see hybrid models evolving with multiple outcome funders, including a combination of government, development agencies, and philanthropic organizations.
  3. Developing government capacity for multi-year social service contracting with contingent payments will have big payoffs.What happens when an impact bond transaction is able to deliver the expected outcomes? There may be a number of options for expanding the impact through follow-on transactions that extend coverage to other needy populations. We may also see government agencies, as service providers, adopting the now tried-and-tested delivery model. However, the most promising approach to scale and sustainability is likely to be through changes in government procurement modalities—reforms that make it possible to do more of the social service contracting on the basis of multi-year commitments to service providers, with contingent payments. The big payoff from impact bonds will come not from what happens with the small amounts of funding mobilized for an individual transaction but the potential to do much better with the government’s own resources. An outcome focus can drive funding towards cost-effective interventions, reduce the cost of government oversight, and help build data collection systems for evidence based decisions.
  4. Service provider capacity and a supply of impactful and scalable programs are likely to be in shorter supply and a more serious binding constraint. SIBs so far have largely built on existing delivery models, for which there is some evidence of achieving results, with providers that can show a track record of capacity to manage programs. The typical SIB has then taken these to greater scale for the same group of beneficiaries, served a different population group, or combined more than one intervention in an adjusted delivery model. An important challenge in the developing world is the lack social programs with results backed by hard evidence, and a scarcity of credible suppliers of social services outside the public sector. Much of the services in developing countries are provided informally by unregistered providers, with small NGOs that operate in one part of the country with rudimentary management information systems and capacity to drive performance and expand services. There are of course exceptions, but the field for many nongovernmental services is characterized by fragmentation in delivery. In these contexts, the question is whether impact bonds can help overcome fragmentation in delivery or even catalyze a new, more effective range of providers. This is one critical area where significant adaptation may be needed in impact bond practice in order to serve the developing world’s needs.

The impact bond market is growing fast but it is still in its infancy. It has a number of features that are attractive to diverse stake holders anywhere. But to become a valuable tool for development purposes, impact bonds will need market building support and will have to find ways of addressing limited supply capacity, engaging developing country governments, and demonstrating the benefits of a results focus for the efficient and accountable use of scarce resources to investors, governments, and the public alike. Even then, impact bonds may fail to achieve scale or shift government procurement processes in a meaningful way.  We are keen to observe what shape impact bonds will take in low and middle-income countries and whether they will become a useful tool in tackling some of the intractable social problems in the developing world.

This article was updated on May 1, 2015.