Nobody knows, even assuming they care

[E]conomic historians will probably spend the next 75 years debating whether monetary or fiscal policy dragged the developed world out of [the current] recession, just as they still discuss whether the actions of the Roosevelt administration really shortened the Depression. Heat and Dust, the 'Economist' (subscription), 24 April

Cause and effect in economics, as in society and ecology, are nearly impossible to disentangle. There are too many variables and time lags. Conditions are never replicable; experiments are close-to-impossible. As well, the interests of politicians, bureaucrats, and technocrats are rarely exactly aligned with those of society. Yet the policymakers' influence is huge, as is the scale of the problems they attempt to solve. In short, we don't know how best to solve society's complex and urgent problems, and we can't be certain that policymakers really want to.

That's where Social Policy Bonds might offer an improvement over today's approach to policymaking which, as the excerpt above points out, is somewhat haphazard. A bond regime would mean that policymakers don't have to have advance knowledge of the best way of solving social problems: all they would have to do is target each problem, and supply a rough estimate of the maximum value to society of its solution. Investors and potential investors in the bonds would work out the best combination of ways in which the problem could be solved. They would be highly motivated to do so throughout the time it takes to solve the problem which, for broad, ambitious, social goals, could be years or decades - much longer than the usual planning horizon of politicians. And they would do so adaptively; continually monitoring the impacts of their approaches and projects so as to maximise their benefit per dollar spent. They would choose from a diverse array of possible approaches, because that would maximise their chances of above-normal success. And because bondholders' rewards would be inextricably linked to society's benefit, everyone would gain. Rather than the random approach to policy that generates the bewilderment described by the Economist, a Social Policy Bond regime would generate valuable information about those approaches that work and those that don't.

The contrast with today's top-down, one-time-only, one-size-fits-all approach is stark.