Make them tradeable

I've explained why I don't like Social Impact Bonds and it turns out that, ten years after the issue of the first SIB - Peterborough Social Impact Bond - others don't either.

Professor James W Williams, in a report he summarises here, looks at the record of SIBs in Canada, the UK, and the US. 

[A]lthough SIBs promise a win-win-win scenario in which providers, government, and investors all benefit, in practice the circumstances in which this alignment of interests is possible (what respondents described as the “SIB sweet spot”) are much more limited than is often acknowledged. In practice, the interests of these groups have been difficult to align with many prospective projects foundering between feasibility and execution. From visions of promise to signs of struggle: social impact bonds in Canada, the US, and UK, James W Williams, 2 May

The key, I think, is that the issuers of SIBs nominate the providers, who are generally the current providers, and whose identity will not change during the lifetime of the bonds. Crucially, SIBs aren't tradeable, so the set goals are short term, and providers can profit only if they become more efficient during the short time period before the bonds' term expires. They cannot profit by realising any appreciation in the value of their SIBs before the bonds expire. Making the bonds tradeable, as I have argued here, would greatly enhance the range of social problems that we can solve, and would do so partly because it would allow new providers, who think they will be more efficient than current providers, to buy the bonds and benefit from their greater efficiency. Competition for the bonds in an open market would ensure that all providers would be kept on their toes: would-be investors in the bonds who think they can be more efficient would bid more for the bonds than they are worth to current holders, and get a chance to prove, and benefit from, their greater efficiency.

Social Policy Bonds are, essentially, tradeable SIBs. A Social Policy Bond regime would not assume that the best achievers of social goals are the current providers - a fatal assumption, in my view and one that, as well as restricting the value of the bonds, makes them liable to gaming and manipulation.

Professor Williams allows that "SIBs could be reserved for testing programs that are truly innovative and preventative in nature with an emphasis on systems-level change." They could, in other words, be a way of rewarding successful innovative approaches that can scale up. I would say, though, that the scope for innovation would still be constrained by the short time horizons built into the SIB mechanism. So I agree with Professor Williams' conclusion that "SIBs are, and are likely to remain, a relatively small, niche market."

Another recent paper for, as far as I can tell, different reasons, is even less enthusiastic. It concludes:

We do not think that [SIBs] can facilitate the maintenance and development of public services that meet society’s needs, particularly the needs of its vulnerable members. In Finance and the Good Society, Nobel laureate Robert Shiller (2012) argues that “we need more financial innovation not less.” We disagree. We have had quite enough financial innovation. Nor do we agree with [his] contention that we can “reclaim finance for the common good.” A quarter-century after first making its appearance, the Private Finance Initiative, a previous example of financial innovation—once popular among policymakers but derided by critics—has at last fallen from favor....  We do not want to wait a similar period to witness the popular discrediting of SIBs: our policy recommendation is that the experiment ends now. S Lilley et al, Using derivative logic to speculate on the future of the social investment market, Journal of Urban Affairs, 18 April

Despite my reservations, I do see pluses in the Social Impact Bond experiment. The first is that they compel us to think in terms of outcomes; the second, that they reward more efficient achievement of these outcomes. These qualities, a commonplace in the private sector, are revolutionary when applied to our social and environmental goals, where 'more efficient' applies to the improved well-being for ordinary citizens, rather than the narrow, short-term, accountancy goals of corporations.

The danger, from my point of view, is that SIBs' failings will discredit, in the public eye, all approaches that reward efficiency in the public sector. But it could go the other way, and I hope that it will: policymakers might come to see SIBs as a step towards Social Policy Bonds and, though it would involve ceding their power to choose service providers, make them tradeable.