16 Oct 2019, 10:18 a.m.
Despite having been on the scene for almost 10 years, impact bonds remain a niche instrument by most measures. They still represent a miniscule proportion of both public and aid spending in most countries which have used them. Using the amount of private investment they are reported to have attracted can serve as a proxy measure. The GO Lab projects database shows that US$ 441 million of private capital is committed to impact bonds, which represents a mere 0.1% of a global impact investment market estimated to value US$ 502 billion (and is but a grain of sand in the US$ 73 trillion invested in global stockmarkets). Yet it was once expected that they would grow much faster than this. Back in 2016, for example, the Minister for Civil Society in the UK set an ambition to see a market of GB£ 1 billion in the UK alone by 2020.
Why the gulf between rhetoric and reality? Some evidence (and much anecdote, including my own personal experience) points to how hard it is to get an impact bond off the ground, which might be one reason they remain niche. Public stats on how many have fallen by the wayside are hard to come by, but the LOUD report published in 2017 identified 71 groups in the UK who had received funding to develop an impact bond, and found 41 of these did not go on to launch one.
And yet, despite the challenges, many people are still excited about the concept of Impact Bonds, and pursue them vigorously. They remain official policy of the UK government (who continue to fund half of our research). The US government has legislated for their use. The EU’s institutions continue to explore them (the European Investment Fund put EU€ 10 million into a Finnish one). Even the mighty G7 have “recognised the potential of outcomes-based funding instruments”. More than 27 countries now have at least one impact bond. And they continue to generate much commentary, debate and curiosity. If so many people think impact bonds are the answer – what is the question?
One of the fundamental properties of the impact bond instrument is that it attempts to move the risk that services do not deliver the promised outcomes away from government, nearer to the organisation doing the delivery. That organisation may be more agile and better able to adapt to changing circumstances. Because the instrument links this delivery risk to financial risk, it ought to create a strong incentive to deliver success (so as not to lose money), and perhaps also to outperform (since the corollary of taking a risk is receiving a reward, as we explain here).
However, financial risk is not the only type of risk. At the start of our international outcomes conference last month, Professor Stefan Dercon questioned the utility of transferring financial risk from the public to the private sector, given a government’s treasury represents “the biggest risk pool in the entire country”. He attributed the political appetite for impact bonds to a perceived transfer of reputationalrisk – because you can play the blame game if something doesn’t work out. But in reality no-one emerges smelling of roses when things go wrong. The UK government has not emerged elegantly from the string of high profile failures of large corporate outsourcing companies like Carillion, because taxpayers had to pick up the pieces. And it is citizens who rely on those services who suffer most.
Sir Ronnie Cohen is often called “the grandfather of social investment”. He believes history has shown private capital to be the best tool for solving large problems at scale. He is hardly the only one who thinks so – the UN says that there is a US$5-7 trillion dollar funding gap in financing the Sustainable Development Goals (SDGs), and UNDP has said impact investment is key to filling it. Tom Hall made a similar point in the closing debate at the conference.
Many of the sectors and industries required to help meet the SDGs already function effectively as markets, so can be supported by private investment. Some people would argue that social outcomes can be treated in the same way – they can be “purchased” from a market. The promise of impact bonds from the private sector perspective relies on this assumption. The argument is that if the responsibility of government to provide so-called “social goods” could be linked with the capability of capital markets to respond to needs and scale up solutions, progress would be quicker. Beyond impact bonds, other approaches rely on a similar assumption – look, for example, to the experience of public-private partnerships in education. But not everyone agrees that it is right to apply market principles to social outcomes.
The first of these two reasons still rely heavily on hyperbole, without a great deal of evidence. A lack of transparency around many impact bonds – in the UK, at least – makes the first claim, about risk transfer, difficult to assess. Since contractual terms are seldom shared, we often can’t see how financial risks are really distributed between the government, the investors, and the providers. Neither do we tend to know how well the projects have performed, and therefore whether the government was spared paying for failure. For some, this defeats the point of impact bonds altogether. The second claim, that impact bonds unleash capital to solve problems at scale, will be validated if and when private capital flows in more quickly and easily than it has so far. But there is a third reason, where there is a bit more evidence. Impact bonds are one way of getting multiple players to focus on desired end outcomes, not pre-defined activities. As we showed in our 2018 report “Building the Tools”, this shift in focus has sometimes unlocked the value of collaboration, prevention and innovation in public service delivery. Conversations at the social outcomes conference suggested in some cases it has helped cultivate learning and feedback loops, and flexibility at the frontline. This all needs more study, and no one is claiming impact bonds are the only mechanism for unlocking change in practice. But change in practice is clearly needed – how many people have you heard arguing for the status quo lately? So let’s keep encouraging experimentation.