As far as I know, there's not been a single trial of the original Social Policy Bonds. Social Impact Bonds, on the other hand, are being issued in about nineteen countries. The main difference is that SIBs aren't tradable and, as I argue here and here, this drastically diminishes the range and timescale of the social problems that they can set out to solve. It also, again in my view, opens them up to exactly the sort of gaming and manipulation that feature so prominently in our current policymaking environment, and from which we are trying to escape.
That said, there are genuine difficulties in getting people interested in the Social Policy Bond concept. Principal amongst these is that it's never been tried. The difficulty is that Social Policy Bonds are at their best, in that they show their most marked advantage over current policies, when the problems they target are likely to require trialling and adapting many diverse approaches to their solution.
If we already know the best approach, there's no need real need for Social Policy Bonds, and if we know the people best placed to solve the problem, then SIBs, essentially functioning as a performance-based incentive, are adequate. My contention is that, for many of our biggest and most urgent challenges, neither of these conditions apply. We don't know how, for instance, to end war, nor which combination of people and organisations are best placed to do so.
All this implies large-scale goals with a relatively long time frame. So I've found it easy to write about applying the Social Policy Bond principle to goals like world peace, universal literacy, and climate change. It's more difficult to think of less grandiose goals, immune from the possibility of gaming, that could serve as experimental examples to prove the validity of the concept. Perhaps SIBs, despite my concerns, could serve as a stepping stone toward the issuing of a Social Policy Bond, and the subsequent refinement, development and implementation of the concept, so that it can meet our large-scale challenges.