Social impact bonds: a letter from the frontline
After seven years ‘in the trenches’ as part of Australia’s foray into social impact bonds, Elyse Sainty shares her insights around SIB myths and legends (part 1). In part 2, she outlines her ‘thoughts and prayers’ for the future.
For over seven years Elyse Sainty has been ‘in the trenches’ leading SVA’s social impact bond (SIB) and outcomes-based contracting work. She has helped service providers develop more than a dozen proposals, been at the table during eight contract negotiations, managed five active SIBs, worked with eight different line agencies across four state governments, and secured SIB capital from 160 investors. In this two-part article she shares her reflections and insights.
It has been quite a rollercoaster ride being part of the obscure world of social impact bonds while they were largely unexplored territory, and everyone involved had to do a lot of making it up as we went along. There have been times of frustration, exhilaration, and the satisfaction of complex problems solved. Strong friendships and strained relationships. Lessons learnt, sometimes the hard way, and lots to celebrate, particularly when things work out as hoped and meaningful positive change happens in the lives of those that it is all about.
Is it all worth it? In short, yes, but with a few caveats. In this article I’ll expand upon this, challenging a few myths (in part 1) and sharing some of my personal ‘thoughts and prayers’ (in part 2) about the role that SIBs could play in the evolution of a system that focuses on what is important, uses evidence to shape responses, and delivers for the community.
It is worth saying up front that the musings are purely my own, and I reserve the right to change my mind completely as events unfold.
In writing this article I’ve assumed the reader has some familiarity with how SIBs work and are developed. For background information, you can read more here.
Part 1: Myths and legends
In years gone by I have sometimes commented that there have been more words written about SIBs than dollars raised by them. The balance has probably tipped now, but it is fair to say the topic has produced a somewhat surprising amount of commentary, with opinions covering the spectrum from wild enthusiasm to downright hostility. Some of those perspectives on SIBs I regard as ‘myths and legends’, and I’ve tackled seven of them below.
1. SIBs bring new money to the social sector
Additionality: Extent to which a new input (action or item) adds to the existing inputs (instead of replacing any of them) and results in a greater aggregate.
SIBs do bring in new money, but then it has to be given back, so it is not additional funding.
This misconception is fading, but there has been a tendency for people to confuse capital with revenue (“SIBs enable organisations to diversify their revenue streams”). The only time investors leave their money behind is when the program hasn’t delivered as expected – which is core to how SIBs work but not exactly an objective.
Based on conversations over the years, I think it has also taken some in government a while to understand that SIBs aren’t a magic source of extra funding, and further that if you ask someone to take a risk, it is reasonable to expect that there is a price to be paid to them for so doing This means that the expectedcost of a program is higher under an outcomes-based contract than if you just fund it on a fee-for-service basis.
So why bother? The flip side is that payments are deferred until outcomes are measured and (at least some) savings captured, which has budgetary benefits, and of course if things go poorly taxpayers are protected. Outcomes-based payment provides an option for governments to take a punt, to try new things, without bearing all the associated risks.
I also believe that there are non-monetary aspects of SIBs that do deliver ‘additionality’: new perspectives, new data, new discipline.
2. SIBs = investors profiting from misfortune
Schadenfreude: pleasure derived by someone from another person’s misfortune.
This criticism surfaces from time to time, and I think misses a couple of fundamental points.
Firstly, investors only ‘profit’ if fortunes improve – interests are aligned in a positive way.
Secondly, it implies that social sector organisations should not use some of the financial instruments that are regularly accessed by commercial organisations; that they should stick to philanthropic support. Whereas I think they have every right to do so (and to generate a financial surplus if they do a good job).
Teasing this out further, the financing component of a SIB can essentially be broken down into two financial services:
- Bridging finance (needed because outcome payments are received after the costs of delivering a program are incurred); and
- Insurance (protecting against the risk of an adverse event, in this case a program not delivering the level of performance expected).
Each of these components are valid services that any organisation entering into an outcomes-based contract may require, and it is reasonable that the provider of the financial service (in this case investors) is paid appropriately for doing so – the ‘cost of capital’. In the Australian context, the largest pool of potential investor capital is superannuation funds, and fund trustees have a fiduciary duty to ensure that returns are commensurate with the risks being undertaken.
In some cases, philanthropists may explicitly (by providing a guarantee) or implicitly (by ‘mispricing’ their required return) provide these services on a non-commercial basis. Philanthropic support can be a wonderful catalyst for a particular project, especially if the risks involved are high, and thus theoretically more costly to offload. I do worry, however, about the long-term viability of outcomes-based contracting if it becomes expected that philanthropists will step into the breach.
3. SIBs are complicated; outcomes-based contracts aren’t
False dichotomy: a type of fallacy that presents two options as being opposed that aren’t necessarily opposites.
The first part of this statement is no myth, and I’ll have more to say on that front in the next instalment of this article (sneak peek: it doesn’t have to be that way).
But I do get frustrated when I hear the view expressed that SIBs are complex and expensive, and thus a ‘pure’ outcomes-based payment model (or ‘pay for success’) is a better option. This glosses over three truths:
- SIBs encompass the outcomes-based payment part of the equation anyway, and that is where the development demons have lain as the learning curve has been climbed. (Getting access to useful data, choosing sensible metrics, determining the counterfactual methodology and figuring out a payment structure that is fair to all parties under a range of performance scenarios.)
- Governments should be indifferent to how service providers choose to structure their financial affairs around an outcomes contract. Some will have large enough financial reserves that they don’t need the ‘bridging finance’; some will have large enough risk appetite that they don’t need to ‘insure’ performance risk; others will have neither. Unless this last group can access at-risk capital through a SIB style arrangement – or philanthropic support to cover the downside – they will be precluded from participating in outcomes-based contracts.
- If a service provider chooses to internally fund program delivery until an outcome payment is received, and wears the risk that they may not get paid at all, then they should be able to put a reasonable price on doing that. The cost of capital doesn’t evaporate just because the capital at risk belongs to a non-profit.
It is true that under a SIB structure there is an incremental cost involved in marketing to investors and appropriately managing and reporting on their investment, but this should be able to be kept to a small proportion of total program costs (and underlies SVA’s view that small SIBs are not viable).
4. SIBs direct more money to preventative programs
Extrapolate: extend the application of (a method or conclusion) to an unknown situation by assuming that existing trends will continue or similar methods will be applicable.
This is not a complete myth, but the claim is often overstated. While most SIBs have a focus on prevention, it has generally been on preventing more of something that has unfortunately already happened quite a lot for the individuals being supported. Preventing more days in hospital, more years in out of home care, more stints in prison. These are perfectly sensible objectives, but not prevention in a purist sense.
The further you move into truly preventative territory, for example working with vulnerable children who have the potential to end up as the prisoners or welfare recipients of the future, the harder it is to get a business case predicated on government savings to stack up.
The core problem is that without brilliant predictive modelling you will need to work with quite a few individuals who wouldn’t have gone on to have poor outcomes. The program might still be very beneficial to them, their families and the community, however the pool of future government savings is diluted while the program cost isn’t. A compounding problem (literally) is that the savings generated are many years in the future, and a lot of extrapolation is required.
Finding the right outcome metric is also much more complex in the truly preventative space. Metrics directly linked to government expenditure (days in hospital, convictions and the like) are often not applicable. Not only do you need (as usual) to identify a metric that has decent baseline data (on the current level of outcomes), it needs to be measurable within a reasonable time period (for example, ‘school completion’ is a long way off if you are working with five-year-olds), and you need to be satisfied that it is well correlated with savings even further down the track.
Which brings me to the next point…
5. SIBs are all about government savings
Miss a trick: to not take advantage of an opportunity or situation to gain some benefit.
I believe we are missing the bigger picture in framing SIBs as a tool primarily designed to reduce government expenditure.
Governments spend money all the time on initiatives that are rightly aimed at improving wellbeing or community amenity. We do not ask the question “will this family violence support program produce government savings that exceed the cost of the program?” if the funding is provided on an activity or block-funding basis. It is odd that if the funding shifts to an outcomes basis suddenly that question is all-important.
I am certainly not suggesting that quantifying future reductions in the consumption of government services (hospitals, income support, policing and so on) is unimportant. I am suggesting that value should also explicitly be placed on benefits that don’t accrue to departmental budgets. Money doesn’t have to be saved to make something worthwhile. This would certainly ease some of the business case contortions that are regularly conducted at the moment, and would put the focus on social impact, rather than budget impact.
As a corollary to the above, I also believe we should be better at quantifying the impact on service utilisation (and hence government expenditure) for programs that aren’t funded on an outcomes basis. (More on this in the next instalment.)
6. SIBs are privatisation by another name
Nefarious: wicked and immoral.
I cannot speak for the motivations of governments elsewhere, but I honestly have not detected any signs of a neo-liberal plot to destroy the public service within Australia.
Instead, I see a genuine desire to try something new that might lead to greater insight and better results.
To some extent, there is also a desire to respond to social sector pleas to improve the way services that are already delivered by non-government organisations are procured. One of the key fringe benefits of outcomes-based contracts is that they tend to be longer-term and larger-scale because you need enough ‘data points’ to fairly judge outcomes (reducing statistical error). This enables the service provider to employ and train staff with confidence and provides greater revenue stability. There are other benefits that can arise, but are very much dependent upon the contract management approach taken by government: less micro-managing of inputs/activities; more room to innovate and deliver a service focused on the needs of those being supported.
7. SIBs lock out the little guys
David and Goliath: a contest where a smaller, weaker opponent faces a much bigger, stronger adversary.
This myth has a little truth to it, but does not tell the whole story. The organisations we have worked with to successfully develop a SIB have ranged from the very small ($3 million annual revenue) to the extremely large ($1.5 billion annual revenue).
It is clearly much more challenging for a small organisation to dedicate the necessary resources to what has to date been a lengthy process, but doing so can be a strategic game changer, as Hutt St Centre in South Australia will attest. I think in some ways it can actually be harder for a large organisation to fit the development and management of a SIB into their existing structure and processes, and to maintain the patience for what may after all be a very small component of their activities.
To make a SIB feasible for a small organisation, a number of planets need to align:
- They need strong board support and funding for the development process, and an appetite for the spotlight.
- They need a government partner that takes a collaborative, non-adversarial approach to the development process. (The power imbalance can otherwise be overwhelming.)
- They need an intermediary partner they can lean on heavily, particularly in relation to the technical aspects of SIB design.
- They need to have or invest in strong core business capability, for example robust client data management systems.
That brings my examination of some of the ‘myths and legends’ of SIBs to a close. In summary, I believe that SIBs are not the silver bullet they are sometimes made out to be, but nor should they be viewed with deep scepticism. And there has been a tendency toward ‘typecasting’ that could prevent a clear-eyed view of how the principles and processes inherent in SIB ‘DNA’ can have broader application. In part 2 of this article I’ll explore how some of the challenges of SIB development can be mitigated, and the significant opportunity that lies in spreading that DNA.
[Part 2 will be published next month.]
For more information, contact Elyse Sainty on firstname.lastname@example.org