Where are we at with Social Impact Bonds?
9 August 2016
A good time to reflect
Social Impact Bonds (SIBs)—known as Pay for Success in the USA—have now been part of the policy agenda for some years.
The model is essentially a payment by results contract funded by investors who are paid only if the outcomes are achieved. Established in the last days of the last Labour government (yes, there was one once) they were taken up strongly by the Coalition, pushed by David Cameron as Prime Minister, and at least not blocked by the Treasury under George Osborne. Ministers like Iain Duncan Smith liked them and diverted money towards them. Similar initiatives have started to appear across the world.
But the British politicians who acted as cheerleaders for SIBs have now left the stage, somewhat suddenly. And we do not know if the new Theresa May/ Philip Hammond axis will embrace them anywhere near as vigorously.
We are also seeing various experts and commentators weighing in on the topic as well as on the more general issue of social investment itself. So it’s a good time to reflect on where the whole SIBs debate has got to. Are they working? Or do we need to find a new policy toy to play with?
SIBs may have been talked up a little too much
A major contribution in the crusade to talk up SIBs is Social Impact Bonds: the Early Years, a review of their progress published last month by the organisation Social Finance and its various overseas offshoots. Naturally—given that these organisations make quite a lot of their revenues by creating, managing and lobbying for more SIBs at all levels of government—they are bound to look for the good rather than the bad (or even the mundane). Nonetheless, it is an interesting survey.
The paper gives some (limited) details of the first 22 SIB projects, and argues that they are mostly achieving positive outcomes (although the comparator is not always clear). And even where they don’t work—as most famously with the US SIB at Rikers Island Prison, where recidivism and reoffending did not reduce—the authors count this as a benefit because the public sector does not pick up the tab for that failure. The more pertinent point may be that the beneficiaries didn’t receive a decent service.
There is a lot in the paper about what SIBs bring in learning and in helping to establish evidence where once there was none. This may indeed be one of the major potential benefits of SIBs, but so far there has not been that much serious openness and scrutiny surrounding SIBs, so the learning for the sector and society as a whole feels pretty limited so far.
But it looks like they could be here to stay
The first thing to acknowledge is that the payment by results way of procuring services—by being paid partially or even primarily on the outcomes achieved—is a trend that is here to stay. And it will get more sophisticated. The fundamental question, then, is what SIBs add to the mix.
Before getting into more detailed pros and cons of SIBs, which will be the focus of my next blog, we should be clear that this approach just won’t be right for many social policy interventions, let alone all social organisations. It isn’t hard to see why it looks tempting: the constant search for the perfect intervention lurks not far beneath the surface of the SIB model, but it is often the wrong road to take where we are faced by complex, multi-faceted problems. For most charitable aims, especially when carried out on a small scale, SIBs won’t be the right way to try and solve problems, nor the relevant option, nor the most cost-effective.
We also need to consider whether the public would warm to it if SIBs were suddenly adopted much more widely: it feels pretty unlikely that they would swallow the idea of people making rather large rates of return on social good if this started happening at any serious scale.
So, that’s the lie of the land. In my next blog I’ll drill down a bit more into whether SIBs have met the hype some have built around them. Where they’ve been tried, have they worked? And are they a long-term answer to some of our social problems?