The latest forecast from the Construction Skills Network (CSN) suggests that the current recession in construction will have led to a drop of about 400,000 in the number employed by the industry once job shedding ends in early 2011.

This would mean a drop of about 15% in the workforce. That appears at first sight pretty savage. But set against the previous recession this would be getting off relatively lightly, with a shorter and less severe period of job cutting.

It’s a bit fiddly to get back data to match the numbers used by the CSN model, which puts valuable time in painting a picture of the construction industry as we know it rather than how the official stats see it.

But we can get a feel for the savagery of the previous recession from the workforce jobs figures. These show that between September 1989 and December 1996, when the bottom was reached, the number of workforce jobs in construction fell by 600,000. That was a fall of about a quarter.

This measure is not the same as that used by CSN. For a start it is a measure of workforce jobs not the number of people engaged in then work, so we might expect it to be a bit more squidgy than a measure of people actually employed.

On the other hand it underplays the numbers lost to the industry as a whole as it doesn’t include the professions such as engineers, architects and surveyors, many of whom will have lost jobs in that recession.

Taking all that into account, the current forecast does raise for me the questions of whether this recession is expected to be less severe than in the early 1990s and whether there might be reasons to suggest the relative impact on jobs will be less.

Firstly, there is of course good reason for the CSN to forecast a shallower fall in jobs, certainly given the current data. The official figures as they stand suggest a shift in the balance of work from new build to refurbishment.

With repair and maintenance employing roughly twice the numbers for a given level of output as new work does, we should expect a more muted decline in jobs than the overall fall in output might suggest.

The mix of work does have a significant bearing on the overall numbers employed by the industry. The question is how might this change going forward?

Secondly, when we look at the forecast we have to examine where the risks lie. And it would seem that they lie very much on the downside.

So, while the forecast paints a picture of what CSN expects is most likely - job losses of 400,000 - there is a much bigger chance of things being worse than that compared with the chance of things being better.

In that sense we should be prepared for worse.

Leaving aside the questions thrown up by the output from CSN. There is a problem regarding the accuracy of the base data against which the CSN is forecasting and which it has to use to run its model.

There are, as have been mentioned many times on this blog, concerns over the data. The paucity of good information on the flow of migrant labour into and out of the construction industry is of particular concern.

Thankfully what makes the CSN work particularly useful is that its research and forecasting model is backed up by huge amounts of input from people active in construction up and down the country. This should help to smooth out some of the wrinkles caused by any iffy data.

But while we can question the detail of the forecast, what is not in question is the central issue of what job losses on this scale pose for the industry, which once again looks set to be confronted with a skills crisis once it finally emerges from recession.

The get out last time of importing skilled labour from abroad may not be as easily available in the later years of this decade.

Worse still the demographics are against the industry, as the average age of its workforce has been for years steadily rising. This inevitably leads to rising rates of natural wastage from the workforce.

It is the curse of this industry that the time when it most needs to recruit young people, who in time will develop and provide the core skills for the future, is the time when it is toughest to do so – during a recession.