The Betfair Prof / Leighton Vaughan Williams / 20 April 2009 / 1 Comments

Following the latest devastating earthquake in Italy, the Betfair Prof, Leighton Vaughan Williams, asks if it's time for prediction markets to be tested to give early warning in the future...

In a webcast released by the US Geological Survey (USGS) on 20 May, 2008, geophysicist Mike Blanpied noted that "... there's currently no organization or government or scientist capable of successfully predicting the time and occurrence of earthquakes." Yet historians writing as far back as 373 BC thought they had it figured out and that it was to do with the behaviour of animals. The case in point was a quake that devastated the Greek city of Helice. Just days before, the reports say, a veritable Noah's ark of animals, including rats, snakes and weasels simply upped and left.

Move forward now to February 4, 1975, and the earthquake in Haicheng, China, which measured 7.3 on the Richter Scale. This time seismologists issued warnings about the earthquake a day before it took place and ordered evacuations, the first successful earthquake prediction in history. Once again there were reports, some documented in the academic literature, of unusual behaviour by animals ranging from rats and snakes to cows and horses in the days before the quake struck.

Professional seismologists are sceptical, some explaining the successful prediction of the Haicheng quake in terms of a number of small tremors ('foreshocks') which preceded the big one. Even so, it is common sense that a lot of animals do have their noses close to the ground, and might well detect the first signs of an impending earthquake a little before the rest of us.

The USGS's Mike Blanpied agrees that on many occasions "... an animal will detect the early arrival of the faint waves that come out first from an earthquake [known as P waves] and become aware that the ground is shaking before the humans around them become aware of the heavier shaking that follows. And so an earthquake may cause animals to react a few seconds or even a minute before the humans nearby." Not long, but maybe enough time to take elementary precautions before the heavier shocks to come. The problem, of course, is that animals often behave strangely, for any number of reasons, and a quick dive under the stairs every time the cat takes fright does not really constitute an earthquake strategy.

Biologist Rupert Sheldrake, interviewed in 2003 for National Geographic News, is more optimistic about what can be learned from animals in making useful forecasts of impending earthquake activity, and proposes a hotline or web-site where people could post reports of unusual patterns of animal behaviour. This would be analyzed by software designed to pick up abnormal activity from a particular region. The information would be checked, argues Sheldrake, to discount other obvious explanations, and would be used in conjunction with conventional seismological analysis.

So why not go one step further, and construct a well-defined, sophisticated earthquake prediction market? If the information is out there, in whatever form, such a market might be an ideal way of aggregating it so as to produce the best possible forecast. Will it work? Of that we can't be sure. But with so much at stake, don't we have some sort of duty to try?

Professor Leighton Vaughan Williams is the Director of the Political Forecasting Unit and Betting Research Unit of Nottingham Business School, Nottingham Trent University

To donate to the Red Cross' appeal to help the survivors of the earthquake in Italy click here.