Social Impact Bonds - progress report

Social Impact Bonds (not to be confused with Social Policy Bonds) were launched in the UK in September 2010. SIBs are similar to Social Policy Bonds in the sense that investors receive higher returns if their social intervention is successful. They are being tried over four years with 3000 UK prisoners at Peterborough prison. The more their reoffending rate falls, the more the backers of the bonds will receive.

Toby Eccles, development director of Social Finance, says it hopes to launch half a dozen such bonds across the country and for a variety of social projects over the next 18 months. Other groups ... are looking to foster similar bond schemes in areas such as cutting the number of children going into care or handling people with long-term health conditions better. “Making these work is complex,” says David Hutchison, Social Finance’s chief executive. “You have to be able to measure success accurately and work out how much to pay for it and when. Peterborough is relatively easy – people either reoffend within a year or not. Bonds set to help prisoners break with past, by James Boxell and Nicholas Timmins, 'Financial Times', 10 February

I met Toby Eccles in London last June, and we discussed both Social Impact Bonds and Social Policy Bonds. SIBs are definitely a step in the right direction. The idea of rewarding desirable social outcomes (rather than paying people for simply turning up to work), while obvious in theory, is (sadly) revolutionary in practice. Social Impact Bonds have the great virtue of being easier to trial than Social Policy Bonds. But in other ways Social Policy Bonds have the advantage.

The biggest difference is that, compared with SIBs are less tradeable than Social Policy Bonds. There would be no transparent market for them. The composition and structure of the organizations trying to achieve outcomes under a SIB regime are therefore fixed and pre-determined. Under a Social Policy Bond regime, on the other hand, the type, structure and composition of organizations working to achieve the target would be subordinate to the most efficient way of reaching it. This means, amongst other things, that broad, longer-term goals could be targeted. Such goals can be more closely aligned with society's wants and needs. Rather than target a relatively narrow indicator (like the re-offending rate of a certain set of people), over a period of a few years, they could target regional or national crime rates over a period of decades.

Nevertheless, I am pleased to see Social Impact Bonds being tried and I hope that their major departure from current policy - rewards that are inextricably linked to relevant performance - is taken up more widely.