Avoiding disaster

Discussing the origins of the current credit crisis, Jeff Madrick writes:

The mortgages traveled such a long distance from institution to investor that no one was in personal touch with the actual mortgage holder any longer. Now, the likelihood of defaults was assessed not by someone who tracked a specific mortgage holder but by complex, computer-generated statistical models of the entire portfolio of mortgages. Like all such models, no matter how mathematically intricate, they required an estimate about the future based on the past - an estimate that was inherently incapable of adequately taking into account the consequences of a historically rare and therefore seemingly unlikely crash in housing prices. How we were ruined and what we can do, 'New York Review of Books', 12 February

Society's burgeoning complexity means that the chances of similar disasters occurring in all areas loom ever larger. By their nature, they cannot be anticipated. Even if we can see looming calamity in, say, the way the climate is changing, or the proliferation of nuclear weapons, we cannot know in advance the precise pattern that the calamity will take, still less do anything to avoid it. Our interconnectedness, aided (as in finance) by something approaching a policy monoculture, can amplify the consequences of unanticipated events. Remedies may be too little or too much - but they will always be too late.

One solution could be to issue Disaster Prevention Bonds, which could act so as to moderate or countervail the incentives currently on offer to the people and corporations who are quite happy to militarize our planet. These bonds could function as an insurance policy, rewarding people who work to avoid human catastrophes of any sort, specified or not. Disaster Prevention Bonds would not prejudge how human calamities shall be avoided, but would simply reward the sustained non-occurrence of such calamities. Under a bond regime, diverse, adaptive approaches that are efficient would be rewarded. Failing policies would be swiftly terminated.

Different bond issues could be used to insure against different broad categories of disaster, but always with a focus on outcomes rather than abstract variables; ends rather than means. In the area of finance, for instance, rather than target economic indicators (such as growth rates or house prices), bonds could target numbers of people unemployed or homeless.

The bond principle could be applied in other policy areas. Though we might fear a nuclear exchange or climate change in particular, Disaster Prevention Bonds targeting disasters need not be so specific. They could, for instance, target separately the potentially calamitous consequences of man-made or environmental disasters. In such bond issues the desired outcome - the avoidance of millions of avoidable deaths - would be specified, but not the means of achieving it.