So the latest stab by the Construction Products Association to make sense of the economic chaos suggests that construction output will fall 7% over the next three years. That is a loss in cash terms of about £8 billion in work by 2010.

The pace of decline in sentiment within construction is powerfully illustrated in the latest forecast from the association which is decidedly gloomier than the forecasters expected when they put numbers to expectations in the summer.

The forecasters now fear that the construction recession will be deeper and longer lasting, stretching out a full three years.

Not even by 2012 will the industry have recovered the level of workload it saw last year, say the forecasters. And the risks look heavily piled on the downside.

The association provides two alternative scenarios to its central forecast and even the upbeat outlook paints a grimmer picture than its forecast produced only three months ago.

The grimmest news in the forecast is the change in sentiment towards the commercial sector. Peaking this year, the expectation now is that workload in the sector will collapse by almost a quarter.

To put the expected recession in construction overall into context, it is a year or so shorter and about half as deep as that experienced in the early 1990s. That recession cost 600,000 jobs from a workforce of almost 2.4 million compared with the almost equivalent workforce of just over 2.2 million.

While I do not wish to be overly pessimistic, this forecast takes a rather gentle view of the repair and maintenance sector. In the 1990s recession, while (broadly speaking) the whole industry output fell 13% to 14%, the RMI sector took a real beating falling about 19%.

This kind of makes sense - money is short so put off repairs etc, particularly in the private residential sector.

However the association's forecasters foresee a much shallower drop this time around in the RMI sector than in new work - they could be right, but I will remain sceptical.

Mind you I feel for the forecasters as they have a major problem - they have to forecast against the official figures. And frankly I am not the only one worried that some official statistics are becoming disconnected from reality.

The stats for construction output suggest what amounts to a mini boom in repair and maintenance so far this year, and I for one have not seen any noticeable increase in kerbside skips - quite the reverse.