On Monday, Bernard Keane took me to task over my call that property prices will fall next year on the basis of the “we’re not building enough dwellings” argument, and uses ABS data comparing net immigration to total dwellings to support his case.

If immigration was the sole cause of the need for new houses, and if we housed one immigrant per house, his chart might have some significance. But we have a bit of local population growth to account for as well, and of course the ratio of people per house is a bit lower than one. So a better comparison can be made by comparing the flow of new Australians to the flow of new dwellings, and seeing how that ratio rates against the ABS data that tells us that the average home has 2.54 occupants (the relevant ABS files are ABS8752 for dwellings and ABS310101).

That comparison shows that we between 1985 and 2009, we constructed an average of 1 dwelling per 1.79 new Australians — well below the current average occupancy rate.

On that data, we weren’t building too few dwellings for the past 25 years — we were building too many. Only since the financial crisis began has the rate of construction fallen below the level needed to maintain our current population to housing ratio.

If the “house prices are all driven by population” argument were correct, then we would have had falling house prices for the last 25 years — but we haven’t of course. Instead we’ve had rising house prices, and to explain it, we need to find a related factor that has been rising.

That factor is gearing: the level of debt taken on by Australians to finance house purchases, and the ratio of borrowed money to house purchase prices. Rising leverage drove prices up in a Ponzi bubble, and falling leverage will drive them down again too.

And that bet: I have probably lost the second half of it, when my interest was always in the first. I wasn’t trying to call the peak in housing — if Rory had suggested a bet on that front, that 131 on the ABS House Price Index was the peak, I would have refused it.

Instead the bet was about how far house prices would fall from their peak, whatever that might be — in which case I said I expected a fall like Japan’s over a similar time frame: 40% over 10-15 years.

For the sake of closure however, in late last year (2008) I said that I expected prices to be falling by late 2009, and if they weren’t — if in particular they broke the September 2008 peak of 131 — then I would walk, on condition that Rory also had to don his joggers if over the long term, my primary call was proved right. The Statute of Limitations on that is 2025 — 15 years from now.

If we hadn’t had the First Home Vendors Boost (let’s call it what it is), then I very much doubt that I’d be facing a jog in the Snowy this year. But the Boost happened, it renewed the dying bubble, and I’ll probably have to walk next year. But before 2025 I expect a somewhat older Rory to also have to burn some foot leather on the Snowy track.