It was a remarkable admission from Bill English last week: that no housing is being built for the lowest-income fourth of the population.
"Like, none," he said. "It's unsustainable."
But what is the Government's response?
So far there is none, beyond some rather sinister observations about the mounting cost to the taxpayer of housing subsidies, chiefly the accommodation supplement, and how they can probably be "targeted better".
It is not a problem which has suddenly just arisen.
The Productivity Commission's report on housing affordability notes that a distinctive feature of residential investment (economists' speak for building homes) in New Zealand is that new supply has tended to be large and relatively expensive houses, or to a lesser degree apartments, targeted at the top end of the market.
Tuesday's building consents data suggest that remains the case.
Over the June quarter, the average value of the nearly 1000 consents issued in Auckland was $346,000. Over the past five June quarters it has been just under $300,000.
And that is the value of the building work, not the section, which in Auckland typically costs 50 per cent more than the house built on it.
The commission notes that the trend of building skewed to the top end of the market has been in place since the 1960s.
Since then - in Auckland and the rest of the country - the share of new dwellings whose price would put them in the bottom quartile of the market has fallen from around a third to just 5 per cent, while the proportion of new builds in the top quartile has climbed from 10 per cent to around half.
When the supply side falls well short of meeting demand prices will rise. And boy have they risen.
Between 1970 and 2000 real house prices had cycled around a slightly increasing trend, taking 40 years to double. But in the last decade house price inflation accelerated, prices doubling again in real terms between 2000 and 2007.
While other countries also experienced housing booms (the polite term for outbreaks of virulent house price inflation) during the 2000s, New Zealand's was one of the largest.
By the peak in 2007 it had outstripped Australia, Ireland, France, the United Kingdom, the United States, Canada, the Netherlands and Belgium - though Spain's was larger. By the end of the decade, of that group only Australia had experienced a larger increase in real house prices, underpinned, of course, by much stronger growth in incomes.
They have come off their peak, going essentially sideways over the past five years in nominal terms and therefore falling when adjusted for CPI inflation.
But real house prices in New Zealand are still about 25 per cent above their long-run trend, Bank of New Zealand economist Craig Ebert calculates.
Other metrics like house prices relative to average earnings and relative to rents are likewise off their 2007 peaks but still very stretched by historical standards.
The bubble, in short, never burst.
We escaped the misery of widespread negative equity, with people trapped in houses suddenly worth less than the value of their loans, and the need for taxpayer-funded bank bailouts.
But for how long?
Because now, with ultra-low and still-falling mortgage rates, the bubble is inflating again.
Not to worry, we are told, rising house prices are just a signal for the market to build more houses and relieve the pressure.
Will it, though?
The commission's analysis highlights two major reasons to doubt it: the prohibitive cost of land, especially in Auckland, and the cottage industry structure of the construction sector.
It remains to be seen whether local bodies, including the Super City, are ready to ensure enough land is available to bring the cost of sections down to levels those on lower incomes, including young couples, can afford. And whether the effect of zoning changes would just be vitiated by speculative land banking.
The cottage industry problem is that the building sector, with a few notable exceptions, is made up of small businesses geared to building houses on demand one at a time, but not to larger projects with economies of scale.
Two factors have mitigated - at a stiff price - the social effects of all this dysfunction.
One is emigration, the loss of young people despairing of ever being able to afford to buy a home here.
The other is that during the 2000s rents decoupled from rising house prices, breaking a link which had held steady over the previous 40 years.
Rents rose, of course, but at roughly the same rate as consumer prices generally, and therefore much more slowly than house prices, which increased at around 12 per cent per annum in real terms between 2000 and 2007.
To the Productivity Commission that indicates growing demand for rental accommodation was met by increased supply as the number of landlords surged - to more than 200,000 by 2007, nearly three times the number in the early 1990s.
"Landlords have been prepared to accept low yields [to well below 4 per cent on one official estimate] on their rental properties, given in part expectations of capital gains," it said.
Investors have responded to signals from the tax system: the prospect of high leverage and untaxed income, in the form of capital gains, from rental properties compared with the disheartening front-loaded taxation of superannuation schemes.
In addition there is a subsidy, in the form of the accommodation supplement, flowing from taxpayers to landlords. The Government and Labour are now questioning whether that represents value for money.
So what is to be done?
Increasing the supply of land in ways that substantially lower the cost of sections is clearly necessary, but not sufficient.
Maybe we also need to acknowledge the possibility that our parents and grandparents were smarter than we are.
They were not wedded to the idea, dominant since the mid-1980s, that whatever the problem, the solution is a market.
Faced with an inadequate supply of decent housing, they built state houses - and would be astonished at the prices some of them now change hands for.
And they made State Advances loans available, at concessionary interest rates, for first-home buyers who could only use them to buy new homes subject to size limits.
That policy targeted what are again the missing lowest rungs of the housing ladder.
Right now the Government can borrow 10-year money for 3.4 per cent.
Housing Minister Phil Heatley, on TV3's The Nation on Saturday, said the Government already had $15 billion invested in 70,000 state houses.
"We're in no position to build more and more state houses, you know, we're just not in that position."
Well, why not? The need to rein in public debt?
When it suits it, as with state asset sales, the Government likes to pretend there is only one side to its balance sheet, the debt side.
But there is a world of difference between increasing public debt in order to fund an operating deficit, and increasing it to fund the acquisition of long-lived assets.
Surely the financial markets can tell the difference.By Brian Fallow Email Brian