The latest Experian forecast is out today and it paints a broadly similar, albeit slightly more optimistic, picture to that of the recently released forecast from the Construction Products Association.

The main point of departure is on the views towards housing. Here the Experian forecasters are more bullish, if you can say that about a market that even by 2012 is expected to be almost 30% smaller than it was before the credit crunch.

Experian's expectation of a faster improvement in the housing market is largely why it sees a stronger exit from the recession.

But to reiterate the point made when I looked at the Construction Products Association forecast, these projections are made in constant prices to reflect the change in the volume of work.

So, with prices falling faster than previously thought, to get a feel for the money the industry can expect to make we have to imagine these lines showing a far deeper trough.

In that light, the overall picture we see from both forecasts is that while the worst of the falls might be over, the construction industry is still very much in the mire. Both forecasts emphasise that there are still heavy downside risks, not least the scale and impact of the cuts in public spending when they finally come.

With regard to the differing views on housing, in fairness this is to be expected, and not too much should be read into it. From where we stand, house building is a fantastically tricky market to predict, as the range of likely outcomes is pretty wide.

The impact of Government action, the financial markets and the broader economy on house building is very uncertain and even the smallest shift in the massive forces at play here can have a profound impact on how much output will eventually flow from this sector of the construction market.

A further factor queering the pitch is that the product mix of house builders is changing, outside of London, away from flats and towards more traditional family housing.

So even if the estimates for units prove correct the amount of construction output could vary quite considerably.

It is worth noting that although new homes were on average getting smaller in the run up to the credit crunch the build cost per home rose markedly. Some of this was inflation, but a large part was that there were extra costs associated with building inner-urban flats as compared with outer-urban or semi-rural houses.

So if the mix were to shift to more towards semi-rural houses and away from inner-urban flats we might expect to see a reduction in the construction associated with each new home. This would be on top of the heavy downward pressure house builders have already exerted on costs.

Here is a graph I originally put together for Housing Market Intelligence 2007. Apologies for not updating it, but it illustrates the point.

The unit cost is taken from a crude division of CLG figures for housing completions by housing output at 2000 prices taken from the ONS output figures, so it theoretically ignores the effects of inflation.

We can see the impact of the shift to flats in the early noughties in the figures for average number of bedrooms, taken from the CLG figures.

We can see that coincidental with this rise in the proportion of flats within the mix a sharp rise in average unit cost, even though the units were getting smaller.

It is not unreasonable to assume that this shift would unwind if the product mix reverted. But you have to be a bit careful here because there will be big regional differences. The more homes built in London and the South East and the more built on brown land the more the associated costs.

Furthermore, we have yet to see in the figures the full impact costs associated with sustainability.