Social Policy Bonds: information as well as incentives

In my work in trade policy I often have to point out that zero risk of threats to our animal, plant or human life through introduced species means no imports. We accept the uncertainties and small risks for most products because we think the benefits of trade outweigh the risk-adjusted costs. Policymakers have decided that the optimal level of biosecurity risk is not zero.

Is there also an optimal level of crime? Or road deaths, or illiteracy? There are certainly levels that we tolerate or accept. These levels have evolved through the political process. They have not been calculated nor made explicit. They might be wildly inconsistent and apparently irrational. Limited resources can be allocated according to criteria that have little to do with where they will most cost-effectively improve outcomes. So we spend large sums on, for example, removing small amounts of arsenic from drinking water, when it is at least possible that we could save far more statistical lives by diverting those resources elsewhere.

What is missing is the notion of trade-offs. Our current decision-making processes neither require nor generate the costs of achieving outcomes. So we are flying blind and contentious decisions are made on the basis of speculation, media profile and the relative power of lobbyists.

A Social Policy Bond regime would be different. A government that wanted to raise longevity would auction non-interest bearing bonds that would be redeemable once the longeivty target had been met. These bonds would inject market incentives into finding the most cost-effective ways of increasing the average lifespan. The bonds would not prejudge how best to achieve that objective. Bondholders would have incentives to find and initiate the most effective projects: these could include preventive medicine, or basic health education - both of which receive derisory funding these days because they lack the public profile and media appeal of larger, more visible efforts.

As important, the market for Social Policy Bonds would continuously generate valuable information about how much the achievement of targeted objective will cost. As soon as the bonds are issued, their market value will give an estimate of the cost to the issuers of achieving the targeted objective. The prices of the bonds would be constantly changing in response to events and the anticipated effects of bondholders' actions. From these price changes it would be possible to infer the marginal cost of tweaking the objective so as to bring about further improvements. If, for instance, bonds were issued to increase the current average lifespan to (say) 80 from (say) 76, then at every moment the bonds are on the market they would generate estimates of the costs of raising the average lifespan still further.