HM the Queen famously asked academics at the London School of Economics why nobody saw the credit crunch coming. In fact, if the Queen took an interest in these things, she would have realised that professional economic forecasters are often spectacularly bad at forecasting big changes in GDP growth, exchange rates and practically any other macroeconomic parameter you can think of.
Medical doctors, abdominal specialists, are worse at their job of diagnosing illness than a simple algorithm based on Bayes's Theorem. Professional clinical psychologists are worse at predicting patients' behaviour than their secretaries. Political pundits are no better at forecasting political developments than the average lay-reader of the Economist, and both are much worse than a simple linear model [see Expert Political Judgement]. The voting patterns of the Supreme Court of the United States are better predicted by a simple statical program than by legal experts. Monkeys throwing darts at the FT pick better-performing stock portfolios than professional fund managers.
We know all these things, but we still look up to experts to tell us what to do, whether we are an individual investor, or Gordon Brown asking what to do about those troublesome banks. Given the problems experts have predicting the future, how can we possibly be confident that their recommendations for influencing the future will remotely reliable?
