Economics: A prophet of gloom takes Golden Guru

IT IS the time of year for my small service to the users of economic forecasts, the Golden Guru competition. Now that the Central Statistical Office has given us its first guess at what happened to last year's growth, it is possible to compare forecasters' performance with what actually happened to the economy. The winner, in what is now the eighth annual survey, is one of the Chancellor's seven wise men, Patrick Minford of Liverpool University.

Professor Minford is a worthy recipient of the award because he has never been afraid to stand outside the pack. He reminds me that he did, for example, win my wooden spoon for the worst forecast of 1989, when he exceeded even his usual ebullience with a super-optimistic forecast of the first year of the slowdown.

But on this occasion he was one of only two forecasters to predict that there would be a further decline in national output in 1992. (The second was another wise man, Wynne Godley.) Every other forecaster predicted a rise, just as more than half had predicted a rise the year before. The proliferation of forecasters - now 43, compared with just 12 in the 1986 survey - is not improving the collective product.

Nor is the survey a respecter of reputations. Tim Congdon of Lombard Street Research, the most vituperative of the Chancellor's advisers, recently claimed that his views should be taken more seriously than those of the others because of his forecasting record.

How does he square up? Professor Congdon redeemed his poor forecast of 1991 (when he expected 1.2 per cent growth compared with a 2.4 per cent decline in output) with a better than average forecast for 1992. But he still expected growth, when the economy shrank by 0.5 per cent. In common with most other forecasters, he has got the overall direction of the economy wrong for two years. More modesty might be in order.

If overall output performance is the key measure, Professor Godley has been more consistently right than anyone else. Throughout the recession, he has been one of the handful of people to appreciate the strength of the deflationary forces in the economy, notably the high indebtedness of consumers and companies and their consequent reluctance to spend. In May 1991, he was the only forecaster to project decline in 1992.

To my knowledge, the only other economist who has read this recession correctly is Christopher Dow, the former Bank of England economics director, who writes a circular for the brokers Leopold Joseph and Sons. In January 1991, for example, Mr Dow told us that 'output will fall this year, and again next. Signs of a recovery should appear only by the end of 1992.' Mr Dow, though, does not give detailed forecast numbers, and so cannot be included in this survey. Nevertheless, investors will learn more from his counsel than from any amount of more precisely expressed error.

The survey's basis of comparison is the last forecast made by each group at the end of 1991 for growth, inflation and unemployment. The difference between the forecast and the out-turn for each variable is then arbitrarily added up to give a guru index: the higher the index, the worse the forecast. The figures are taken from the Treasury's round-up of forecasts, Elsevier's 'Economic Forecasts', 'Consensus Forecasts', and the Financial Times round-up.

In each case, the group's forecast measure is compared with the out- turn for that particular measure. Forecasters who appear in the Treasury's round-up predict the fourth- quarter rise in the retail price index (which was 3.1 per cent), while those in 'Consensus Economics' predict the easier year-on-year rise (which was 3.7 per cent). Prof Godley predicted the rise in the GDP deflator, which was 4.4 per cent.

I am the first to concede that this is all rough and ready. However, more sophisticated methods that try to give special weight to forecasts of things (like GDP) that are more difficult to get right do not produce radically different results. So I have stuck to my eight-year formula.

This consistency allows us to make some comparisons over time. After two particularly grisly years in 1990 and 1991, the average Guru Index appears to be improving a little, although still a far cry from the relatively stable years of 1985 and 1986 (when even the Treasury won the prize). The Treasury was again worse than its own average of independent forecasts. Despite the resources the Treasury pours into forecasting, it has managed to beat the average on only five out of eight occasions.

It is the mainstream forecasters who once again come badly out of the survey. Usually bunched close to the Treasury's forecast, they confirm the adage that economists are bad at spotting turning-points in the economy. Forecasting is still essentially a souped-up extrapolation of the recent past, and is not up to spotting discontinuities.

Partly as a result, the consensus forecast was also spectacularly wrong on some of the detail. The independent forecasters, for example, expected the Public Sector Borrowing Requirement in 1992-3 to be pounds 18.5bn. The out-turn was pounds 36.5bn. The average prediction for the balance of payments foresaw a deficit of pounds 8.6bn, against an out-turn of pounds 11.8bn.

Nor is it confidence-inspiring to find that even good forecasts are often right for the wrong reasons. For example, Prof Godley's forecasts of high unemployment in the early Eighties were right not because output fell by as much as he predicted, but because productivity rose by more. In the winning forecast, Prof Minford predicted that output would drop despite a rise in consumer spending of 1.5 per cent in real terms, because investment would drop. In fact, consumer spending was flat and investment fell.

Given the need for forecasting in policy and planning, there is another worrying factor. Someone always produces a good forecast, but it is rarely the same person as before. Even those who are above average are prone to serious lapses. It seems there is still nothing so uncertain as the future.

(Table omitted)