Category: social impact bonds

Last week Instiglio updated its world map of social impact bonds and development impact bonds. Now the map contains all the latest programs in England, Australia, the United States, and elsewhere, and SIBs/DIBs that were already on the map have now been updated with the latest information. Take a look at the map here. The map uses a Google API that links to a spreadsheet with detailed information about each SIB and DIB. The underlying spreadsheet is available here.

Social impact bonds have caught the attention of at least a dozen governments in high-income countries such as the UK, the US, Australia, and Canada. Yet discussions around the use of this innovative performance contract for international development in emerging economies have started relatively recently. My own social enterprise, Instiglio, which I founded with my partners Avnish and Michael, has been making some progress in this space. There have recently been three major developments in this space that are worth reviewing:

1. The creation of the Development Impact Bond Working Group. Social Finance UK, which structured the world’s first social impact bond over in England, and the Center for Global Development, which had been the intellectual champion behind the Cash on Delivery model, created a working group that is comprised of top officials from leading development agencies and is tasked with exploring the application of social impact bonds in emerging economies. This group has released information about two meetings (here and here) and created web pages to track its progress (here and here).

2. International development agencies have formally and informally expressed interest in social impact bonds. First, the Canadian International Development Agency has released a proposal to study the application of SIBs to access to finance. Second, DFID and Social Finance UK conducted a scoping study of a SIB for family planning. Third, the Inter-American Development Bank, the Asian Development Bank, several other development finance agencies have started internal discussions around the social impact bond model.

3. DFID, UK’s development agency, has included a request for social impact bond approaches in its requests for proposals through the Girls’ Education Challenge Innovation Window. This is the first public request for social impact bond programs in international development of which I am aware. Here is language from the The Girls’ Education Challenge: Guidance to applicants submitting a Full Application for the Innovation Window:

DFID is therefore committed to funding a small number of projects that use PbR (which implies grant payments in arrears following independent verification of pre-agreed results, or may also involve the ‘development impact bond’ approach) and will be looking out for suitable projects during the assessment of Full Applications. These projects will undertake additional evaluation and research on the efficacy of the PbR mechanism, as well as that of the intervention.

And here is language from the Frequently Asked Questions for the Girls’ Education Challenge Innovation Window:

We are interested in funding some projects through the Girls Education Challenge Innovation Window which will be wholly delivered through Payment by Results (PbR). This is a relatively new approach to delivering overseas aid, whereby payment is only made when pre-agreed results have been achieved. The Innovation Window of the GEC is an opportunity to demonstrate and evaluate the use of PbR, within a sample of pilot projects, to increase Value for Money (VfM) and achieve better learning outcomes for girls. Proposals looking to use or trial a PbR mechanism are encouraged, though we appreciate that PbR will not be appropriate for all projects.

Reuters recently published an article that describes the ongoing creation of a social impact bond pilot for a preventative healthcare service in Fresno, California. In this pilot, the nonprofit organization Collective Health is testing whether services to reduce environmental triggers for individuals with chronic asthma will generate savings to the insurance companies that pay for those individuals’ resulting hospitals visits. This blog post reviews the program and offers some analysis.

The problem: Fresno County has 200,000 individuals living with asthma. These individuals account for over 6,000 emergency room visits and over 1,100 hospitalizations annually. The total annual cost, including foregone worker productivity, is $87 million. (Source)

The solution: Several services can reduce home-based asthma triggers by improving indoor air quality. These services can reduce asthma triggers and prevent costly ER visits and hospitalization.

The barrier: Individuals switch health insurers too often for any single insurer to realize financial gains from asthma preventative services. Therefore, insurers under-invest in this preventative programs for asthma.

The innovation: If insurers are convinced that these preventative services reduce their costs, they will have the incentive to create a collective pool of funds that will invest in these services and generate cashable savings to all the insurers. (This solution has the accompanying collective action problem that each insurer will want to reduce their financial contribution into the pool and increase their financial benefit from it.)

The caveat is that neither the current Fresno pilot nor the envisioned future program resembles the social impact bond model that is being designed in Massachusetts and New York. In the Fresno pilot, the California Endowment is paying service providers up-front to reduce chronic asthma triggers, while actuaries work to determine the financial outcome of the program months later. This is not a social impact bond; it’s just good investment into a pilot to test whether savings materialize.

In the program envisioned after the pilot, insurers will collectively purchase services to reduce asthma-related hospital visits and benefit from the resulting financial gains. The social impact bond component of this model is that private investors will give service providers money, and insurers (or a fund they jointly create) will repay these investors. However, if the service is proven to be effect, the insurers can simply pay for the service and include a basic performance bonus to incentivize good outcomes, rather than creating a social impact bond with private investors. In order to determine if the social impact bond is the most appropriate financial mechanism for preventative asthma services, I think the designers of this program should consider two questions:

1) If insurers are convinced that the preventative service saves them money, will they be able to come together and purchase this service?

2) If insurers are able to purchase this service, can they disburse money in expectation of the savings, or are they constrained (in some legal, technical or budgetary way) in paying only out of the resulting savings?

3) If insurers can make an up-front purchase of this preventative service, is a regular performance bonus to service providers sufficient to generate the same level of savings that would exist in a social impact bond? In other words, are the benefits of the SIB worth the costs of creating it?

Data for this blog post were taken from secondary sources, including the Collective Health website and the Reuters article on the Fresno program.

As part of its effort to create a pay-for-performance pilot, Minnesota recently released a request for information that asks the public to submit information in response to multiple questions about the potential design of a pilot program. The state has expressed willingness to issue bonds up to $10 million to finance the pay-for-performance pilot. The RFI asks for feedback on programs in workforce development and supportive housing. There are several similarities to, and differences from, the pay-for-success model that is currently being designed by Massachusetts.

– Minnesota has issued two RFIs, one for service providers and another for third parties. – Massachusetts issued one RFI, but two RFPs, asking service providers and intermediaries to apply separately for the invitation to negotiate with the state. One purpose of this separation was to avoid a situation where the intermediary entered into an exclusive relationship with a suboptimal service provider, or vice-versa. A downside of this separation was that it was much harder for the intermediary to create, plan, and describe a potential program delivery model without knowing for sure which service provider’s model it would ultimately use. Minnesota seems further removed from this decision. It currently seeks responses that would inform its program design, rather than RFP responses from which it would select its preferred service providers and third parties.

– Minnesota’s RFI mentions that the state might play the role of an evaluator. The Massachusetts RFP did not discuss that option. If the state commissions the social outcome, agrees to pay for it, and evaluates the service providers’ success in achieving that outcome, then it will have to manage a perceived – and perhaps a real – conflict of interest.

 – The bond. Minnesota is interested in issuing a bond to pay service providers for successful achievement of social outcomes. Here is the logic of issuing a bond, as I understand it. The service providers’ program would reduce costs or increase revenues to the state. These financial benefits would appear in the state’s budget during the service providers’ program and after the program; service provision will create other benefits, as well, but they may not appear on the budget at all. Massachusetts and Minnesota have chose two different options for paying for all these benefits. Massachusetts has requested budgetary authorization (for $50 million) to pay service providers if they achieve predetermined outcomes. Minnesota has chosen to issue a bond (for $10 million) to pay service providers if they achieve their outcomes. Minnesota will presumably repay the bond at least in part from the financial benefits that the service providers’ programs created.

 – Service providers’ funding gap. In the proposed designs of both states, service providers will need to self-finance or raise external capital to fund the implementation of the program. The state pays only when – and if – predetermined outcomes have been achieved. In both cases, there is an opportunity for private capital to fund social outcomes.

Sources
– Minnesota Third Party RFI
– Minnesota Service Provider RFI

McKinsey’s Social Sector Office has continued their work with social impact bonds. Yesterday, they issued a 68-page report that analyzes the nascent market for social impact bonds in the United States. This report is part of their larger work on SIBs, which includes:

• “Will social impact bonds work in the United States?” –  a brief overview of SIBs issued in March 2012.
• “From Potential to Action: Bringing Social Impact Bonds to the U.S.” – This report, issued on May 15, 2012. 
• Rapid Sustainability Assessment – A toolkit aimed to help potential funders, providers, and intermediaries determine their organization’s suitability for participation in the SIB ecosystem. Publication date is TBD.
• Capabilities Due Diligence – A more thorough evaluation for each of the stakeholders. As I understand, this will focus on due diligence analysis of potential service providers. Publication date on this is also TBD.

A great article by Katie Gilbert about Steve Rothchild’s human capital performance bonds came out yesterday in the Institutional Investor.  Some interesting quotes:

On states looking into SIBs:

The state of Massachusetts formally sought out information about pay-for-success contracting over the summer, and this winter will issue a request for proposals from pubic social programs interested in the model. In New York City, a pay-for-success system will soon be applied to a program that works with the adolescent portion of the adult justice system and seeks to lower the chance that they’ll recidivate. New York State is also exploring the concept, and hopes to present a concrete pilot to test pay-for-success in six months to a year. Similar signs of interest are springing up in Virginia, Michigan, Indiana, and parts of California.