For at least a decade, an Australian housing shortage has been a gospel truth amongst economists and policymakers, fingered as the major cause of Australia's high home prices.
So sure of its existence and malevolence was the Rudd government that it established a National Housing Supply Council to tell us each year how big the shortage was and it how it might be reduced.
But the evidence is mounting that there never was a shortage and, even if there was, that it will soon be eliminated by the stubborn east coast building boom that refuses to wither.
Last week, the Bureau of Statistics released much of the census data collected last year. On the basis of that census data, the bureau's population figures were revised, higher as it turned out, especially in Victoria.
Stronger population growth might have dented the surplus I found when I interrogated dwelling construction versus population growth figures two years ago.
But no. Australia still has a massive dwelling surplus nationally. Using an admittedly simple, but conservative, methodology, it seems that in my lifetime the nation has built at least half a million dwellings more than it needed to house its growing population.
The calculation isn't rocket science - it simply looks at how much the population has grown, divides that by 2.5 (the lowest average number of people per household thus far recorded in Australia) and multiples the result by 25 per cent to account for intentionally vacant dwellings (holiday homes, Airbnb, etc) and demolitions.
The result should be a reasonable estimate of how many new homes we need to build to match real demand.
Housing analysis an inexact science
Of course there are limitations to this analysis. Due to the availability of data, it starts in 1982 and it makes no claim to analyse whether there was a housing shortage, surplus or balance in that year. It assumes balance.
This probably turns out to be a reasonable assumption.
Historical analysis by Saul Eslake in a submission to a Senate inquiry into housing found that Australia's housing stock grew much faster than its population for the entire second half of the 20th century.
It needed to, because the average household size over that period shrank from about 4.5 to 2.6 - that means you need roughly 40 per cent more homes to house the same number of people.
That trend has stopped, with the past few censuses showing average household size relatively stable around 2.5-2.6.
It is arguable that a housing shortage and rising prices have halted the shrinking household trend.
It is equally arguable that such a trend had to end eventually anyway - an average household size of 2.5 is pretty small given that most adults still live in couples and have children at some point in their life.
Either way my estimate is conservative, because it takes the lowest average household size recorded thus far and extrapolates it all the way back to 1982, when households were bigger.
[At this point it must be noted that Mr Eslake was a member of the National Housing Supply Council and, when he wrote this submission in 2013, subscribed to the view that Australia had a significant housing shortage].
It is also arguable that dwellings are getting smaller with the rise in apartment living, but the 2016 census doesn't yet bear this out, with the average home still containing 3.1 bedrooms.
Amid lingering doubts about the sheer size of this apparent national dwelling surplus, I turned to Lindsay David and Philip Soos of LF Economics.
They use a more sophisticated methodology for their housing supply and demand analysis, but their results are very similar.
They go back to 1996 as the starting point because that is roughly the point when home prices started consistently rising much faster than inflation and incomes.
Despite excluding the period pre-1996, when Australia's housing stock was growing much more rapidly than its population, LF Economics still finds a dwelling surplus of almost a quarter of a million homes.
Housing surplus grows when prices rise
Not only that, but Philip Soos observed that, over the past two decades, the periods of greatest excess housing supply had occurred while prices were rising, while periods of undersupply coincided with weaker price growth or falling values.
It makes perfect sense. Here's why.
"Rising prices provide incentives to developers and builders to construct supply and offer it to the market," he explained.
"We saw in the 2000s with the United States, Ireland and Spain, they went on absolute building sprees and yet, all the while, their economists were saying high prices were caused in part by shortages and that obviously wasn't the case."
But, despite having co-authored an 800-page economic history of the Australian housing market, Philip Soos is well outside the mainstream of property analysts.
BIS Oxford Economics, on the other hand, is about as mainstream as you can get. The firm is an economic consultant to some of Australia's largest property and building materials firms, which rely on its estimates and forecasts for their business decisions.
Whatever its methodology, BIS Oxford doesn't see a huge national housing oversupply at the moment. Then again, nor does it see much of an undersupply, except perhaps in Sydney.
The upwardly revised population figures for Melbourne have eradicated the apartment glut BIS Oxford previously thought existed, but not for long if tower blocks keep springing up in and around the CBD at the rate they have been.
And Melbourne's gain was in large part Perth's loss, with WA's population figures revised down by well over 50,000 while Victoria picked up more than 100,000 residents.
That leaves WA with the most oversupplied market on BIS Oxford's forecasts out to 2019, with Queensland looking very risky and every other market except NSW slipping into at least mild oversupply by next year.
Sydney the only capital where housing shortage may exist
It's a conclusion Philip Soos agrees with, even if he differs on the magnitude of the oversupply.
"I would say that Sydney, given that it has the highest rent-to-income ratio and also has the highest rent growth at the moment, that's probably the most supply constrained," he observed.
"Followed by Victoria, which is Melbourne, even though it has the highest cumulative oversupply, because we've got 120,000 flooding into the state every year and most of them are heading towards Melbourne, it's still a pretty thin buffer."
If you don't believe the private sector analysts, then a group of academics from the Australian Housing and Urban Research Institute have reached a similar finding that there is no housing shortage - and they only looked at the last decade, which is when Australia's population growth has been at its highest.
"In 2015, dwelling completions increased the Australian housing stock by 2 per cent and yet again South Korea is the only country that achieved a larger proportionate increase," their paper noted.
"The net of demolition growth in housing stock exceeds population growth in Melbourne, but falls just short of stronger population growth in Perth.
"There remains a healthy margin between expansion of the housing stock and population growth in all the other state capitals and territories except Sydney."
I also called the Housing Industry Association, which once did estimates of the housing supply-demand balance.
But its senior economist said the HIA had given up, precisely because a shortage/surplus result was so heavily dependent on which base year you selected.
He explained that the HIA now watches the rental market, particularly rises and falls in rents, as a timely way to determine whether there is a housing shortage in a particular market.
Weak rental growth indicates no housing shortage
It makes sense because asking rents are generally more responsive to supply-demand imbalances and also because rents, unlike home purchase prices, are not influenced by interest rates and the availability of credit - you don't borrow money to rent a home, how much you will pay depends on your income and how well supplied the market is.
On just about every measure, except the vacancy rates compiled by real estate agents, the rental market is the weakest it's been since the last recession, which also happened to be a time when Australia was building a lot of homes for a relatively small population increase.
Pretty much whichever way you cut it, it's difficult to argue that Australia has a serious housing shortage.
For that reason, plus the positive correlation between supply and prices (there is usually surplus construction when prices are going up and insufficient building when prices are falling), it is almost impossible to maintain the argument that a housing shortage is the major reason that Australian home prices have almost doubled over the past decade having more than doubled over the decade before that.
If a housing shortage isn't the problem, a dwelling construction boom isn't going to be the solution to Australia's housing affordability woes.
The real answer probably lies in the level of speculative demand and ability to pay more for properties facilitated by record low interest rates, financial deregulation and investor subsidies such as negative gearing and the capital gains tax discount.