Chapter 2 · Housing × immigration × construction

The Ladder

Australia does not build enough homes for the people it invites. Everything else in this chapter is arithmetic, and you can do it yourself with the sliders below.

Start with the only numbers that matter. Australia's population grows by net overseas migration plus natural increase. Divide by average household size – about 2.4 people per dwelling – and you get the number of homes the country needs to finish every year just to stand still. Add the homes demolished or falling out of the stock. Compare with the number actually completed. If the second number is smaller than the first, prices rise until enough people give up. That's it. That's the whole housing market.

Recent form: net migration ran at roughly half a million people in the peak year, against completions of under 180,000 dwellings. At 2.4 people per home, 538,000 migrants need over 220,000 homes on their own – before a single Australian-born couple forms a household. The National Housing Accord's target of 240,000 homes a year has never once been hit. The gap compounds, the price of admission to family formation rises, and the birth rate – already at a record low – follows it down. Casey's back-of-envelope claim is that returning prices to a level consistent with the ongoing social reproduction of Australian culture requires roughly doubling supply. For calibration: the government's own housing council puts the Accord-period shortfall at about 220,000 homes – a 25% lift on current output, not a doubling. The difference is the target. The council is pricing "meet new demand"; Casey is pricing "make the existing stock cheap enough to start a family in", which means building past demand for years. You can test both claims below, rather than taking either on faith.

The gap, as built

Dwellings completed vs dwellings required by population growth · Australia, annual

Required = (net overseas migration + natural increase) ÷ persons per household, plus replacement of demolished stock. Sources: ABS Building Activity, ABS National, state and territory population. Full sources & formula.

The simulator

The model is deliberately simple enough to hold in your head: population growth sets required homes, the construction rate supplies them, the shortfall moves real prices, and a hypothetical buyer – a 22-year-old full-time worker without a university degree, on the median wage for that description – saves a slice of their after-tax pay until they can clear a 20% deposit on an entry-level home (about 65% of the national median – the bottom quartile of the market) and service the loan at 35% of gross income. Every parameter is on a slider or under Assumptions. Move things. Break it. See what it takes to put the bottom rung back on the ladder.

0600k/yr
100k400k/yr
$40k$120k
5%40%
2%9%

Buys at age

–

Annual dwelling gap (year 1)

–

required minus built

Median dwelling, 2040

–

real (2026 dollars)

Price-to-income, 2040

–

median dwelling ÷ buyer's salary

Assumptions & formula
required(t) = (NOM + natural increase) ÷ persons per dwelling + demolitions Δln price(t) = 0.6 · income growth + ε · (required − built) ÷ stock ε = price response entry(t) = 65% · median price(t) — bottom-quartile "first home" buyer saves s · after-tax(income) at 2% real; buys when savings ≥ 20% · entry AND repayments(80% loan, 30yr) ≤ 35% · gross income

Defaults: natural increase and household size from ABS; median capital-city dwelling price and the buyer's wage from ABS series (see Assumptions); price response ε = 2.5 (a 1% shortfall in stock lifts real prices ~2.5%), consistent with RBA estimates of supply-demand elasticity; real wage growth 1%/yr along the ABS age–earnings curve; tax at 2025-26 resident rates including the Medicare levy. The migration-cap toggle applies Casey's Table 2: migration is limited to 100k when real prices are rising, stepping up to 500k when they are falling more than 3%/yr. Casey asserts the cap schedule; data everything else.

What the model says about the doubling claim

Now connect this to the birth rate. The average Australian first-home buyer is now in their mid-thirties. Fertility is not symmetric in time: by 35, roughly 90% of a woman's egg reserve is gone. A housing market that pushes household formation from the mid-twenties to the mid-thirties is, mechanically, a contraceptive. Australia's total fertility rate hit a record low in 2024, below replacement every year since 1976. The people who would have been the 2040s tax base are not being born, and Chapter 3 shows what that does to the ledger. Houses are not intrinsically productive and their material cost is not that high. A country that requires its young to pour the bulk of their lifetime output into servicing loans on structures they cannot live without, instead of investing it in children and businesses, has chosen decline – one auction at a time.