There was a marvellous cartoon during the Hawke-Keating years showing Keating peering down into a hole with binoculars. The hole had been made by a line showing Australia’s current account deficit, which had crashed off a wall chart and through the floor.

“I swear it’s a J!” Keating is saying, referring to the once-beloved J-curve theory about fixing the current account.

Steve Keen reminded me of that on Friday. Even after losing his famed bet with Rory Robertson on house prices, Keen is still trying hard to sell gloom and doom, and reckons that, even as he treks back down Mt Kosciuszko as the price for betting Australian house prices would fall by at least 20% — “Australian house prices will also be losing altitude”. He told the ABC’s Stephen Long something slightly different last month when he predicted he’d be walking from Canberra to Kosciuszko next year but Robertson would be doing it in ten years’ time.

In Crikey on Friday Keen suggested those who doubted his long-term analysis thought “Australia is different”. The delusion that “this time is different” is one of the principle reasons why many people fail to spot asset price bubbles. He called it the “We’re different because we have kangaroos” theory.

Keen appears to be guilty of a similar delusion, I’d suggest. In his view, “Australia is the same”, regardless of circumstances. Our home prices went up more than the Americans’. The American housing bubble collapsed. Ergo, we have a bubble and it will collapse.

There’s two unfortunate aspects to that: it’s wrong, and it would be better for many Australians if it was right.

Australia is indeed different, not because the laws of supply and demand don’t work here, but because they do. About our biggest public policy problem is that we don’t have enough dwellings for people who live here, and we’re importing people at a rate of knots. Throw in the fact that Australian households are getting smaller, and we’re actually going backward in terms of housing supply.

The Housing Industry Association estimates that in 2008 we had a housing shortage of 34,000 dwellings. They of course would say that, being the industry body for the sector that benefits from housing investment. So if you want some evidence, this is a comparison of ABS data on total dwelling units with ABS data on net immigration.

The “total dwellings” line is the number of total dwellings commenced each quarter in Australia. Notice that it peaked in 2002. After another peak at the end of 2003 it has been in decline ever since, before falling off a cliff last year.

The net immigration line is the number of net arrivals from overseas. “Natural increase” – births minus deaths – is currently fairly stable at around 40,000 people a quarter, having shifted up from the mid-thirty thousands a few years ago.

For most of the last thirty years total dwelling commencements have been ahead of net immigration. The numbers aren’t important in themselves – we’re not trying to build dwellings for every single individual who arrives – but the surge in immigration from 2005 was unprecedented and our dwelling commencements not merely couldn’t keep up, they were actually declining.

That’s why there’s insufficient housing stock. And the problem isn’t going to be remedied any time soon unless Australia slashes its immigration dramatically. There won’t be a huge improvement in dwelling commencements in the next twelve months – the HIA estimates there’ll be a 13% fall in housing starts this calendar year, compared to 18% last year. Even with a return to growth in 2010, the HIA estimates we’ll still be 20-30,000 housing starts shy of what we need.

A return to even the insufficient dwelling commencement levels of recent years assumes that property developers – more of them later – can access finance. The vanishing of finance for property development is a key reason for the precipitate decline in dwelling commencements in recent months. It is by no means clear that this will be rectified any time soon, particularly given the increasingly dominant major banks have moved into mortgage lending at the expense of business lending given the collapse of the non-bank lending market.

This is why the Reserve Bank is concerned about growth in housing prices and its potential to feed inflation. I’ve quoted this statement from Glenn Stevens before and it’s worth repeating.

A very real challenge in the near term is the following: how to ensure that the ready availability and low cost of housing finance is translated into more dwellings, not just higher prices. Given the circumstances – the economy moving to a position of less than full employment, with labour shortages lessening and reduced pressure on prices for raw material inputs – this ought to be the time when we can add to the dwelling stock without a major run‑up in prices. If we fail to do that – if all we end up with is higher prices and not many more dwellings – then it will be very disappointing, indeed quite disturbing. Not only would it confirm that there are serious supply-side impediments to producing one of the things that previous generations of Australians have taken for granted, namely affordable shelter, it would also pose elevated risks of problems of over‑leverage and asset price deflation down the track.

Contrary to Harry Triguboff’s attack on the RBA today, the Bank’s concerns about housing stock reflect not a trigger-happy mentality but justified concerns about supply failing to keep up with demand.

It doesn’t amount to a lot in terms of overall housing stock, but there’s another, rather sad indicator of just how badly we’ve dropped the ball on housing. This is what has happened to government investment in public housing in recent decades, reflected in the number of dwellings funded by the public sector.

The Government’s $6.4b stimulus package investment in social housing – for 20,000 new units – was a massive step in the right direction to reverse an extended period of neglect, particularly in Victoria and NSW, but it’s a one-off investment – and now $750m lighter to pay for school halls.

While the problems of property development finance and the long-term decline in public housing investment are Australia-wide problems, there’s a fairly clear reason why our overall performance on housing has been so poor. While the rest of Australia, on average, has maintained a steady level of dwelling commencements more or less since the introduction of the GST, NSW has performed disastrously, with our biggest state entering a long decline in 2003 which is shows no sign of pulling out of.

If you want to know exactly how the performance of the NSW Government translates into a national economic problem, there it is right there.

Tomorrow: NSW and the painful political dilemmas of development.