Yes.
We have access to very granular data on household borrowing, on mortgages. We're able to model this quite closely. Where we begin to get more uncertainty in the picture is with how people actually respond. We can do the arithmetic on it, but the economics are more complex, because it concerns behaviour. For instance, as Ms. Wilkins said, we have people deciding to buy a smaller house. It still comes up as a sale, yet they managed to meet the criteria, and there are others who delay, and so on.
It's in the behavioural differences where the more statistical, historical modelling comes into play and where you have a zone of uncertainty around those kinds of predictions. It's why we can't be so definitive about how the thing unfolds. As it unfolds this year, we have three things happening at once. It's going to take more than a few data points to be able to figure out how much is due to the revised B-20 guideline, how much is just due to higher interest rates, and how much is just that pull forward and then the return to quasi-normal.
I accept your point that it is a difficult situation for people. As for the price level of housing, none of these things we've talked about are aimed at somehow controlling the price of housing. They're aimed at improving the sustainability of debt so that our financial system is less at risk. The price of housing is fundamentally driven by demand and supply, and the biggest thing there is that we have strong demand and relatively weak supply.