Australia’s long-term preference for housing over productive business investment has implications that extend beyond affordability and productivity. Capital allocation shapes the factories, equipment, skills and supply chains available when strategic conditions deteriorate. Rebuilding sovereign industrial capability will require policy settings that make productive investment more attractive and give businesses greater confidence to commit capital over time.
Every dollar of capital in an economy is allocated somewhere. Australia’s has gone, disproportionately and for a long time, into an asset that already exists. The consequence is not just that young Australians cannot buy a home. It is that when the country needed an industrial base in a hurry, it discovered it had spent forty years financing something else. To change the ship’s direction will take a lot of time and energy.
There is a version of this article that blames a tax concession, names it, and finishes in six hundred words. This is not that article, not because the tax settings are irrelevant, but because treating a single deduction as the cause of a forty-year allocation pattern is analytically lazy and, on the evidence, wrong. The pattern predates the current settings, has continually survived changes to them, and appears in economies that never had them.
I am not an economist, an accountant or a licensed financial adviser, and nothing here is financial or investment advice. What follows is an argument about where capital goes and what it does when it gets there, built from the Reserve Bank’s own research, the Productivity Commission’s own bulletins, Treasury’s own words, and a reality Australians are viscerally aware of.
Start with Treasury’s admission
In the productivity paper Treasury prepared for the August 2025 Economic Reform Roundtable, it states plainly that business investment as a share of GDP is lower now than it was in the early 2000s; and that the decline has been larger among the more productive firms, meaning capital has also been allocated less productively.
That is the Commonwealth Treasury, in a document written for the Treasurer’s own roundtable, describing a national investment problem in its own words.
The Productivity Commission reached the same place from a different direction. Its September 2025 quarterly bulletin found that Australia has experienced little capital deepening over the past decade, because we have tended to invest less of our national income in new capital. Capital deepening is the unglamorous engine of productivity: workers become more productive when they have more and better equipment to work with. When it stalls, output per hour stalls with it. In the same bulletin, labour productivity rose 0.3 per cent in the June quarter; but only 0.2 per cent across the whole year.
Reserve Bank research had already documented the mechanism. Its 2023 discussion paper Doing Less, with Less found that slower capital deepening contributed to the productivity slowdown, and that non-mining investment had fallen to a historically low share of non-mining output.
Three institutions. Three methods. One finding.
The divide is a cohort fact, not a feeling
The Australian Bureau of Statistics compared 25- to 39-year-olds across three censuses. In 1991, two-thirds of Baby Boomers in that age band were homeowners. By 2006, 62 per cent of Generation X were homeowners. By 2021, 55 per cent of Millennials were homeowners. Boomers at that age were three times more likely to own their home outright than Millennials were at the same age.
The Australian Institute of Health and Welfare (AIHW) puts a longer lens on it. Among 25- to 29-year-olds, home ownership fell from 50 per cent in 1971 to 36 per cent in 2021. Among 30- to 34-year-olds, it fell from 64 per cent to 50 per cent.
Here is the part that gets left out of the generational-warfare version of this story, and it matters. Ownership also fell for people approaching retirement: among 50- to 54-year-olds, from 80 per cent in 1996 to 72 per cent in 2021. This is not simply a transfer from the young to the old. It is a narrowing of who owns anything at all, and it has been running for half a century under governments of both persuasions.
The objection: housing is investment
It is, and this is the strongest argument against everything above.
Construction employs hundreds of thousands of Australians. New dwellings are real capital formation; they are counted as investment in the national accounts, and a country with a growing population that fails to build them creates a different and more immediate crisis.
Nobody serious argues Australia should build fewer homes. I am not arguing it either.
But the aggregate hides a distinction. Capital directed at building a new dwelling adds to the stock. Capital directed at bidding for an existing dwelling mostly changes who owns it and at what price. The first is investment in the economic sense. The second is a transfer with a mortgage attached. Both show up as credit growth. Only one of them ends with a machine on a factory floor.
The headline ownership rate has held up: around two-thirds of Australian households owned their home at the 2021 Census. What sits underneath it has changed. The International Monetary Fund’s Financial Soundness Indicators put Australian household debt at 112.1 per cent of GDP at December 2024, easing to 111.6 per cent by September 2025; historically high, though below the peak of roughly 124 per cent reached in 2016. A steady ownership rate carried on a heavier mortgage is not the same asset position, and it is not the same balance sheet available to fund anything else.
The objection that actually stings: business credit is growing faster right now
It is, and the honest thing is to say so.
The Reserve Bank’s financial aggregates for June 2026 show business credit growing 10.8 per cent over the year, against housing credit at 7.5 per cent. Business credit has been the main driver of total credit growth for months. Anyone arguing that Australian banks will not lend to businesses is arguing against the current data, and they are wrong.
Two things can be true. Flow is not stock, and a strong recent year does not reverse a forty-year composition. And credit growth is not the same as capital deepening; a business can borrow to refinance, to acquire a competitor, or to buy commercial property, none of which puts better equipment in front of a worker. The Productivity Commission’s finding is about what the money bought, not how much of it moved. Let me be clear: many will “fight and die” on that hill, talking about the mobility of cash flow as if it is the whole story. It is not.
If the current lending pattern holds for a decade, the argument in this article weakens considerably. That is a testable claim, and I would rather state it than hedge it.
The regulator has now said the quiet part
In November 2025, the Australian Prudential Regulation Authority (APRA) did something it had never done before: it activated a debt-to-income limit on residential mortgage lending, effective 1 February 2026, capping loans at six times income or more to 20 per cent of each lender’s new mortgage lending, applied separately to owner-occupier and investor portfolios.
Read APRA’s own framing carefully, because it is more modest than the headlines. The regulator said the limit is not currently binding at an aggregate level and is not expected to affect near-term access to credit. It is a guardrail, put in place before it is needed, alongside a serviceability buffer of 3 per cent and a countercyclical capital buffer of 1 per cent.
The significance is not the size of the intervention. It is that the prudential regulator has now formally identified the concentration of household leverage in housing as a system-level risk worth pre-empting; and that it took until 2026 to reach for the tool.
Why this belongs in a defence publication
Because the argument in the companion piece to this one has a balance sheet behind it. An industrial base is not summoned by a strategy document. It is the accumulated residue of thousands of private decisions to put capital into plant, tooling, testing equipment and the people who operate them: decisions taken years before anyone needs the output. A country that has spent four decades directing its savings toward existing dwellings has, by definition, spent four decades not making those other decisions.
When the strategic environment deteriorates, and the government announces that sovereign manufacturing is now a national priority, it is not creating an industrial base. It is discovering how much of one it has. Capital allocation decisions made long before the announcement set the answer.
What would actually shift it
Not a single lever, and not the one everybody argues about.
The Productivity Commission’s own proposal, in its final report released in December 2025, is a package: a company tax rate of 20 per cent for companies with revenue below $1 billion, held at 30 per cent above, coupled with a 5 per cent net cash flow tax on all companies that allows the immediate deduction of capital expenditure. It is the second half that makes the first half revenue-neutral, and it is the half most commentary leaves out. It is a serious idea from a serious institution, and I note it without endorsing it, because I am not qualified to price it and neither is most of the commentary about it.
What I can say is what the evidence points to. Capital follows risk-adjusted return. Australian capital has behaved rationally, given the returns on offer. If the country wants more of it in productive assets, the work is on the return side: planning and approval timelines, energy costs, the depth of the domestic market, the speed of government procurement decisions, and whether a small manufacturer can get a contract decision inside a business planning horizon rather than outside it.
That last one is not an abstraction. It is the same finding the Auditor-General has been publishing about defence acquisition for eighteen years, arriving from a different direction.
Australia did not accidentally under-invest in its productive capacity. It made a very large number of individually sensible decisions to put money somewhere safer, and it is now discovering what those decisions cost when the safe thing turns out not to be.
