For the first time in three years the Golden Guru award goes to forecasters who do not work with a big and complicated computer model of the whole economy. Normally, forecasters who win the Golden Guru award - such as the National Institute for Economic and Social Research, winner for the past two years - have an advantage over economists using either smaller models or less formal methods.
In contrast, both winners this year, Michael Saunders at Salomon Brothers and Stephen Radley at the Henley Centre, describe their approach as "intuitive" or "ad hoc".
Most of the top-ranked forecasts for 1997 benefited from the adding extra optimism about the path of unemployment, although a special mention goes to Kevin Gardiner, UK economist at Morgan Stanley, for being the only one to err - just - on the side of over-optimism. Conversely, those congregated at the bottom of the league made the mistake of being far too pessimistic about how much joblessness would fall.
Mr Saunders said: "On unemployment, I had learnt the hard way by being so wrong the previous year." Mr Radley attributed his success this year to drawing on other Henley research which emphasised that the service sector, the fastest growing part of the economy, was much better than manufacturing at creating jobs.
The reason the jobless predictions sorted the sheep from the goats so effectively was that on the whole the players started 1997 with their forecasts for GDP growth and underlying inflation clustered fairly closely together. Nor were there any big surprises on either the growth or the inflation front.
The biggest surprise came last Friday, with a weaker-than-expected figure for growth in the final quarter of last year. Of course, this preliminary estimate will almost certainly be revised, but even though this could alter the exact rankings at the top of the table this year, by tradition the Guru decision is never changed.
The award is based on the unweighted sum of the absolute errors in percentage points of forecasts for full-year GDP growth, RPIX inflation in the year to the fourth quarter, and the unemployment rate based on the claimant count. The organisations included are those listed in the Treasury's monthly comparison of economic forecasts which give an explicit prediction for each of these three elements. It is, without doubt, a partial and unscientific measure of forecasting ability, but our decision is still final.
Current forecasts for 1998 start out much more widely dispersed than at the same stage last year. According to this month's Treasury summary, predictions for GDP growth this year range from 1.5 per cent (from one of this year's winners, Salomons) to 3.2 per cent. Forecasts for RPIX run from 2.2 per cent to 3.7 per cent, that is from just below target to outside the top of the permitted range - not a very optimistic picture given how weak many of the economists expect growth to be.
However, all 42 organisations reckon claimant unemployment will be no higher by the final quarter than it is now, and one is plumping for a figure just below the million mark. They have all obviously drawn the same conclusion from the very favourable mix of low inflation and bigger- than-expected falls in joblessness during the past two years.
Next year's Golden Guru result will depend therefore on whether the British economy glides in for a soft landing or nosedives into the ground.Reuse content