The latest forecast from the Construction Products Association puts the annual peak to trough fall in construction at a shade above 20%.

That probably translates to a 22% to 23% fall peak to trough on a quarterly basis, which compares with the 15% seen in the 1990s recession.

This is a sharp downgrade from the forecast made just three months ago and is evidence of how the industry's prospects have slipped deeper into the mire, despite the much talked of green shoots.

Graph 1 illustrates just how much worse prospects have become in the eyes of the forecasters.

CPA fcst comparison summer 2009.gif

The detail of the revisions will be carried elsewhere, but it is worth noting that the drastic drop in the mighty commercial work coming so close on the back of a weakened housing sector plays a large part in pulling down the figures.

But what concern me more than the level of workload are the implications for the structure of the industry and more particularly the workforce.

If the pattern of job shedding follows that of the 1990s we could see more than 800,000 jobs shed. That is more than one in three.

How do I arrive at this shocking figure?

Well between 1989 Q2 and 1993 Q3 workload dropped 15%. That cost the industry 543,000 jobs or 23%. But the job losses didn't stop there. During the sluggish recovery roughly 80,000 more jobs were shed.

This saw the construction workforce (measured by the Labour Force Survey) drop from 2,392,000 to 1,771,000 - a 26% fall.

As Graph 2 shows jobs dropped in the previous recession far faster than output and they continued to fall well after the upturn started.

Jobs and output 06 07 09.gif

If a 15% fall in workload leads to such devastation in the workforce, what would a 22% fall do, given that the upturn looks set to be very slow indeed?

Well given that the projected scale of the decline expected by Construction Products Association is in the order of half as much again, it is not unreasonable to assume that more than a third of jobs could be at risk.

Put another way we could see the industry shed 800,000 over the next three to five years.