Renaissance-style illustration of a bull.

A Return to Pre-2000 Normalcy

An exercise in the fall required to return Australian house prices to their pre-2000 relationship with household income and rent.

November 13, 2026 — Shane Hull

Australian house prices have outrun income and rent for the last three decades. The relationship between prices, income and rent was treated as obsolete. Prices are now falling. This exercise asks how far they would have to fall to return to the pre-2000 relationship, using the 1996 level as the destination.

Two measures do the work. Price-to-income compares the house price index with household disposable income per capita. Price-to-rent compares the same price index with rent, the investment earnings of the dwelling. They sit near the top of their history, but the rent measure is further from its 1996 level.

The measures

Price-to-income divides the Organisation for Economic Co-operation and Development (OECD) nominal house price index by Australian Bureau of Statistics (ABS) household gross disposable income per capita. The income is seasonally adjusted and divided by estimated resident population. The ratio is indexed to 1996 = 100.

Disposable income is after tax and transfers, before housing costs. It includes rent received by household landlords and excludes corporate, institutional and foreign investors’ income. The series begins in 1981.

The measure is per capita. Population is a stable denominator over the long run. Household formation responds to house prices, so population is the cleaner base for a long series. This is a relative price measure. It is not a household affordability measure.

Price-to-rent divides the same house price index by the OECD rent price index. The ratio is the inverse of the gross rental yield. It is indexed to 1996 = 100. A higher index means more years of gross rent to buy the dwelling. The series begins in 1972, and its all-time low sits before the figure below starts.

The gross measure is the usual long-run comparison. A net cap rate after vacancy, operating costs and depreciation would be lower. The ranking is similar if the expense ratio is broadly stable over long periods, which this exercise assumes.

Two panels: price to income index and price to rent index, 1996 = 100, with 25th and 75th percentile lines, income to 2026-Q1 and rent to 2025-Q3

Indices, 1996 = 100. Data to 2026-Q1 for price-to-income and 2025-Q3 for price-to-rent. Dashed lines mark the 25th and 75th percentiles of the period shown. The figure starts below 100. The rent series' earlier observations are lower. Sources: OECD and ABS. Live charts: price to income and price to rent.

Where we are

Measure Latest Date Percentile
Price-to-income 200.6 2026-Q1 100th
Price-to-rent 267.6 2025-Q3 98th

Price-to-income is at its series high. Price-to-rent is about 6% below its 2022 peak. Both sit near the top of their history.

Why the old relationship

The pre-2000 index levels are the reference because the rules changed after that period. Higher rates are the historical norm. The low-rate period after 2000 was the anomaly. The fall in rates fuelled leverage throughout the financialised economy. Most of the price shift came from that leverage. Lower interest rates made debt cheaper and raised borrowing capacity. Financial deregulation let lenders and borrowers act on that capacity. The two reinforced each other, and household debt grew much faster than income from the 1990s. Negative gearing and the 50% capital gains tax discount (from 1999) amplified demand for investment property, but they were secondary to the fall in rates. The exercise uses the pre-2000 relationship because it is the last period before the stretch. That is the sense in which it is normalcy. It is the old regime. It is not a natural constant.

The destinations

The central question is how far prices would have to fall to reach the 1996 level. Price-to-income falls 50%; price-to-rent falls 63%.

Lollipop chart of the price fall required to reach each destination: 50% for price-to-income at its 1996 level and 63% for price-to-rent at its 1996 level.

Price fall required to reach the 1996 level.

The 1996 level is the last major low before the current high, and the common reference for both measures.

The ladder

The ladder shows the percentile rank at each 10% price fall. Every row holds disposable income and rents at their current levels; only the price of the dwelling stock moves. The 0% row uses 2026-Q1 for price-to-income and 2025-Q3 for price-to-rent. Percentiles use the full history of each series.

Price fall Price-to-income Price-to-income percentile Price-to-rent Price-to-rent percentile
0% (latest) 200.6 100th 267.6 98th
10% 180.6 87th 240.8 91st
20% 160.5 77th 214.0 85th
30% 140.4 52nd 187.2 76th
40% 120.4 46th 160.4 58th
50% 100.3 18th 133.9 55th
60% 80.2 0th 107.1 50th

The income measure’s 50% row is 100.3, just above the 1996 level. Its 60% row is 80.2, below the income series’ minimum. The rent measure’s 60% row is 107.1, near its 1996 level. The rent measure is further from its last major low.

Time does part of the work

A fall is not the only path. Prices can stay flat, or fall part of the way:

Measure Assumption Years to the 1996 level
Price-to-income disposable income +3.5% a year 20
Price-to-income 30% fall, then flat, income +3.5% 10
Price-to-rent rents +4% a year 25

The income measure needs about 20 years of flat prices, or a 30% fall and 10 years of income growth, to return to its 1996 level. The rent measure needs about 25 years of flat prices at 4% rent growth, because the gap to its 1996 level is larger.

Caveats

Prices, income and rents are nominal. The indices are inflation-neutral. In the ladder, income and rents are held at their current levels; the time estimates state their growth assumptions. The ladder measures distance in today’s prices.

The data is quarterly and published with a lag. Price-to-income ends 2026-Q1 and price-to-rent ends 2025-Q3.

The exercise measures the distance. It treats the historical relationships as a useful guide; it does not require prices to return there. The 1996 level is a reference point. The likely mechanism is debt: the price falls relative to income and rent when credit stops expanding or becomes more expensive.

And they will print

Policy is currently resisting a fall by pushing the cost onto younger buyers, as argued in Dousing the Pyre. Younger buyers’ mortgages at today’s prices let existing owners sell and lenders keep the loans performing.

Since that article, policy has also begun priming the fall: the 2026-27 Budget abolishes negative gearing for established homes from 1 July 2027 for properties bought after 12 May 2026, and replaces the 50% CGT discount with cost-base indexation and a 30% minimum tax. Resistance and priming pull in opposite directions, but the post-2000 low-rate regime is the historical anomaly behind the distortion of property prices.

The exercise asks how far prices would have to fall to return to the pre-2000 relationship. The answer is a long way. If anything remotely close to this scenario occurs, it is this author’s opinion that the government will print the difference.

If the government turns to monetisation, the price level rises. The debt is fixed in nominal terms, so its real value falls. But historically, low rates and repression filter into financialised assets, including investment housing, which bids up prices and worsens the ratios. The effect on the ratios is ambiguous. But one would hope, for the next generation, that this time they go down and stay down.

Download the data: JSON, CSV.