Mr. John McCarthy:

Now, the evidence base for a soft landing was ... well, it was along the following lines: there is a soft landing on the prices side, and there is the soft landing on the volume side - volumes, I mean housing output and so forth. Let me deal with the latter first. We made the assumption in various budgetary forecasts, and we modelled a soft landing on the basis of we were currently, or at the time, producing 70,000, 80,000, 90,000 units at the peak. At the time, it was estimated that, on the basis of demographic trends, and on the basis of fall in headship rates, in other words an increase in the percentage of the population who are heads of household, that there was an underlying need for about 50,000 units per annum. Now how do you go from 90,000 to 50,000?

Does it fall like that or does it like it fall like that? In the budget documentation, our central scenario was a gradual easing - a fall of about 5,000 or 6,000 units per annum. So, operating over the medium term and converging towards sustainable levels of output. That was the central scenario but we did model internally - in a document that I wrote with a colleague who is now the Secretary General - what would be the implications on the public finances of a hard landing if output was to go from 90,000 down to sustainable levels more rapidly? In fact, it didn't go to sustainable levels, it overshot on the way down, as sometimes happened. And we concluded that the impact on overall economic activity of each 10,000 reduction would be to shave about a percentage point of the growth rate, add about a percentage point to ... a half or a percentage point, I can't remember, to the unemployment rate, add a half a percentage point to the deficit, and these were simply the first round effects. Clearly, we went from 90,000 to 10,000 in the space of three or-----