Letter NOT published in the Guardian. Submitted 30 April.
My two favorite commentators on the subject of risk have fallen out. Simon Jenkins (29 April) predicts that swine flu will end up with bird flu and Sars in the category of over-hyped scare that never happened. Ben Goldacre (30 April) says that if Simon turns out to be right he wont have been right, just lucky – a swine flu pandemic is a real risk.
They are discussing what risk geeks call a low-frequency high-impact event. Chances are that Simon will be proved right/lucky. If you bet against every hypothetical low-frequency high impact event that might happen you will be right/lucky most of the time. Many are hypothesized, few materialize.
The hypotheses that bird flu or Sars could have been, or swine flu might be, as catastrophic as the 1918 flu epidemic were not, and are not, implausible. But they can be assigned low statistical probabilities. What should governments do when confronted with such threats? How much should they spend in the form of money, time and economic disruption to defend against them? Building defenses against low probability threats incurs opportunity costs; chances are the money could be spent elsewhere more effectively.
Consider a much lower frequency much higher impact event a catastrophic collision between earth and an asteroid. It has happened and will do again. In the worlds present state we can but afford to shrug fatalistically. Straining every economic, scientific and technological sinew to defend against it would steal resources from, and thereby inhibit, a wide range of activities that might ultimately build an economy in which asteroid defense would be affordable.
On swine flu Im betting with Simon, but will continue to cheer Ben on every Saturday.