On a recent trip to the US, I noticed that the discussions around ‘Pay for Success’ were a little different to those I’d been having on ‘Social Impact Bonds (SIBs)’ with other countries. Particularly in the measurement community, there was an idea that Pay for Success took measurement of social programs to a new level: that ‘Pay for Success’ meant paying for an effect size (by comparison to a control group), rather than ‘Pay for Performance’ which paid for the number of times something occurred. Continue reading
(Social Finance, Towards a New Social Economy, 2010)
During a recent lecture on social impact bonds for the Saïd Business School’s Executive Education Impact Investing course, one of the students asked me, “Am I already part of a social impact bond?”
Social impact bonds (SIBs) are a new construct, with the term coined in 2008 and the concept piloted by Social Finance in Peterborough in 2010. But arrangements that fit our current understanding of SIBs already existed and continue to operate. For organisations involved in these arrangements, the benefits and challenges of identifying with the SIB phenomenon will determine whether or not they associate themselves with the term.
The lady who asked the question was running a non-profit organisation that had an outcomes-based contract with the Louisiana State Government to provide a service. A loan from a local finance institution helped to cover their working capital.
In order to examine whether she was indeed part of a SIB, we turned to the Cabinet Office definition.
In this arrangement, the first two conditions were certainly met.
To determine whether the third was, we examined whether her creditor could be considered an investor. Her organisation existed for the purpose of providing the service and the contract accounted for the entirety of their service provision. The loan was made on the basis of this single contract with government. The lender was a legally separate entity from both the non-profit and the government. So it looked like her creditor could be considered an investor.
The fourth condition relates to the financial risk of the investor. Was the finance institution truly taking on any financial risk? Indeed, there were two distinct types of financial risk taken on in this situation. Firstly, there was the risk of non-performance; that the non-profit would fail to deliver the results that would trigger payment from government. Secondly, there was ‘appropriations risk’, the risk that government doesn’t approve the payment of a previously agreed contract . If the non-profit, for either of these reasons, did not receive sufficient funds from the government to pay the loan, they would become insolvent and the finance institution would not have the loan repaid. So the fourth condition was met.
This non-profit was certainly part of a social impact bond as defined by the UK Cabinet Office, although none of the parties involved were using the term.
Other examples that may meet the definition of a SIB but not identify with the term include charter schools, which are independently run with three year contracts based on outcomes. If they establish a bridge loan fund to cover the cash flow gap, they may meet the four necessary criteria. These bridge loans are often used for capital infrastructure spend and the financial risk for the lender is sometimes rated, so certainly exists.
- When we assess SIBs and compile data or ‘track record’, it may be worth looking at these existing arrangements that fit the SIB definition.
- Organisations that are part of arrangements that fit the SIB definition might consider the benefits and challenges associated with the SIB label (e.g. will appropriation and other payment risks be reduced if the contract is ‘announced’ as a SIB).
- The Cabinet Office definition came after several SIBs had already been developed and was not written by the Social Finance visionaries responsible for articulating and implementing the original concept. Does it misrepresent the concept? If so, do we need to change the definition?
 Appropriations risk is a feature of the US governmental systems and is described by Steve Goldberg in his fourth SIB Trib: “Even if a state signs a valid contract calling for the payment of money, it doesn’t (indeed, can’t) pay what the contract says it owes unless and until the state adopts an appropriation law specifically and affirmatively authorizing the expenditure. What’s more, legislative authority to appropriate funds only lasts as long as the state’s constitution allows budgets to remain in force. In most states that’s one year, in some, two. So not only does a state have to enact budget legislation and then pass separate appropriations bills to legally spend any money, they can only do the latter within the one- to two-year time horizon of the former” (Goldberg, 2013).
- There was no hype before the Social Impact Bond (SIB) was announced – this is in contrast to other Governments who have announced their intention well before finalising details. Perhaps the deal was developed quickly, which may well be the case with Bloomberg controlling both the philanthropic guarantee and political appetite. Or perhaps the parties didn’t want to announce before they had signed the contracts, avoiding the pressure of having to live up to an announcement of intention.
- In addition to a press release, the City of New York released a briefing pack outlining the justification, partners, program and payment terms of the SIB. The terms of Government payment are clearly set out, unlike in the Peterborough SIB. For a new funding mechanism, this certainly promotes understanding.
- It’s called a Social Impact Bond- Obama’s administration announced funding for “pay for success” bonds and the NSW Government has branded them Social Benefit Bonds. Using the same term as Peterborough makes it clear that they follow the same model and helps keep related literature together.
- Bloomberg Philanthropies guarantees it – they’ve only guaranteed $7.2m of the $9.6m Goldman Sachs investment, but having the guarantee reduces investor risk and may be one of the key factors in getting this deal finalised. It’s not clear under what terms the guarantee is paid or what happens to it if it’s not required. The fact that it’s a grant seems to suggest that MDRC would repurpose it. This also means that like Peterborough, if outcomes are not achieved, there is no payment of public monies.
- The Goldman Sachs payment to MDRC is described as a loan – people often ask of SIBs “Are they debt or equity?”. In fact, they are neither, they are multi-party contracts. Payments from Government are due to an outcomes-based contract. The private investor is really buying a futures option. Classification as debt sheds light on how Goldman Sachs will account for this investment. It also may suggest that Goldman Sachs will take a hands-off approach. If this is so, the benefit of investor skills and incentive are not transferred through to the service providers. This transfer has historically been referred to as one of the reasons for involving external investors, but we’ve yet to see a SIB where investors have a relationship with service providers.
- Similar to Peterborough, there is a third-party, independent evaluator. This may prove to be an essential feature of all SIBs as they emerge.
- The intermediary is MDRC, a research centre – the strengths of MDRC are in program evaluation and design. Intermediaries in other SIBs bring financial investment or service delivery experience to the SIB. In NSW, the intermediary role will be played in part by banks and not-for-profits, as well as by the intermediary Social Finance (no relation to the UK or US Social Finance).
- The payment terms state that a 10% reduction in recidivism is the threshold for investors to break even, although when cost of capital is factored in, this will represent a loss. It is not clear what the 10% reduction is in relation to i.e. what the comparison group is. The SIB also uses stepped payment terms, rather than a sliding scale, which is simple and unambiguous.
- There is value in being a first mover – the press releases announces the “Nation’s First Social Impact Bond Program”. There is a conflict between the kudos received for being a first-mover and the confidence gained by waiting to see how other SIBs turn out. It will be interesting to see whether SIBs become more popular and mainstream, or whether development of this model will slow once its novelty wears off.
- Benefits to nonprofit providers are described in the briefing pack as “a committed funding stream not subject to budget cuts”. The literature that exists on SIBs describes the major benefit to be providers being the flexibility afforded by focusing on outcome, rather than current prescriptive Government contracts. This could be a sign that MDRC will be prescriptive in what it asks of the two providers, or that prescriptive contracts are less of an issue in this jurisdiction.
- Is there conflict between innovation and evidence? The briefing pack justifies the SIB as “encouraging innovation in a time of fiscal constraints”. The idea that SIBs encourage innovation by allowing Government to pay only if outcomes are achieved might be one for the long-term market only. At the moment, the reputational risk of failure to all parties and intense media scrutiny has resulted in SIBs providing services with a strong history and evidence-base that may require strict adherence to service models. These services are so safe they would be ideal candidates for a direct outcomes-based contract with a provider. The real innovation occurring here might be in how the delivery partners have to work together towards agreed outcomes.
- They’re creating a pipeline – the press release refers to an August 2 request for expression of interest for additional social impact investment projects by the City of New York. This would suggest that a pipeline of proposals will be established, but I haven’t been able to find the request anywhere. It would be interesting to see whether outcomes are suggested and priced, in the manner of the NSW Government, or whether this will be up to the market.