I've just received an email from James Hastings, head of construction futures at Experian Business Strategies with the latest forecast update attached.

And, not a surprise, it was gloomier than the summer forecast, which then saw the industry overall stumbling rather than plunging into recession this year and next. It is the most optimistic of the main surveys in the sector released over the past month.

The figures from Experian suggest that if you take housing work out of the equation then output in 2008, 2009 and 2010 will be above the levels in 2007 in real terms, with infrastructure and public non-housing work very buoyant over the coming year or two.

The latest numbers suggestion a peak to trough drop in total construction output of about 4%, compared with a slide of a bit less than 3% that was the view in the previous forecast. Also the recovery forecast for 2010 is now expected to be rather more muted.

Experian doesn't foresee quite such a drop in the massive commercial sector as do the Construction Products Association or Hewes & Associates.

Also, a big factor underpinning the forecast's more optimistic take on the future is that it takes a more positive view of the shape to future work in the RMI sector (repair, maintenance and improvement).

I spoke to James about my unease with the official figures for RMI work and the consequences for any forecast. The official figures provide the base numbers for the forecasters against which they have to forecast.

Frankly I don't believe the direction of the official numbers on RMI work of late. I have mentioned a few times here that they don't seem to fit with my experience of the world and the anecdotal evidence I pick up.

Quite rightly James suggested, politely as he does, that there are serious problems with relying on anecdotal evidence.

He's right in that, but I remain unconvinced by the numbers.

This is why I will repeat again what I say of all forecasts. It is not the final numbers that hold the real worth, but the thought that goes into getting those numbers and the assumptions made. Each and every one is up for challenge and interpretation, but it is always good to know what the latest expert thinking is and how they see the world.