How housing stock composition affects a suburb

• Structure. Separate house, semi-detached or townhouse, flat or apartment (Census STRD).
• Bedrooms. A two-bedroom majority is a different rental and resale market from a three-and-four-bedroom house suburb.
• Tenure. Owned outright, mortgaged, rented. In 2021, rented excludes dwellings occupied rent-free. The ABS warns that this is not comparable to 2016 QuickStats. Do not compute a “rise in renters” across those two Censuses without reading the footnote.
• Pipeline. What is approved or under construction, by class. New apartments change next year’s mix. See apartment supply.

Census 2021 SAL20361 occupied dwellings: separate house 3,892 (35.3 per cent), townhouse 3,235 (29.3 per cent), flat or apartment 3,770 (34.1 per cent). Bedrooms: two bedrooms 43.1 per cent, three 29.6 per cent, four or more 10.5 per cent. Victoria: separate houses 73.4 per cent, apartments 12.1 per cent.

If twelve months of recorded sales in Brunswick are mostly two-bedroom apartments, the suburb median will sit closer to the unit tape. If the same twelve months in a house-heavy suburb are mostly three-bedroom houses, that median is a house tape. Subtracting them is a mix comparison. Split both medians by type, or do not subtract. A labelled illustration: 100 sales, 70 units at a $620,000 median and 30 houses at $1.15 million, can produce an all-dwellings median nearer the units. No house need have cheapened. See comparing suburbs and why medians mislead.

Brunswick occupied dwellings: rented 48.1 per cent against Victoria 28.5 per cent. That is a deep rented stock on Census night 2021. It is not a Districts “investor suburb” label. It is not 2026. Pair it with current listings and the pipeline for the class you would buy. Lone-person households were 34.2 per cent and group households 14.1 per cent, both well above Victoria. That mix supports demand for smaller dwellings. It does not tell you the rent on a four-bedroom house.

These mistakes let the mix of dwellings speak for the house you would buy. Split the stock before you quote the median.

• Reading a falling all-dwellings median as house prices falling.
• Comparing owner-occupier share across 2016 and 2021 without the rent-free definition change.
• Ignoring a large apartment pipeline when the current stock is still houses.
• Using stock mix as a quality ranking.
• Treating incoming apartments as current stock before completion and occupancy.

It does not establish the value of a lot, the quality of a building, or that a high renter share is a problem. Pipeline dwellings are not stock until they complete and are occupied.

Profiles show sales counts beside medians so a thin or mixed tape is visible. They do not always show structure or tenure. Those rows stay on the Census tables unless a labelled source is loaded. Property pages show nearby projects by status, which is incoming mix, not current stock.

Ask whether the median is houses, units or all dwellings. Open STRD and BEDD for the same SAL. Count the pipeline by class on the register. If the sales source cannot split type, treat the median as a mix statistic. If Districts shows a sales count of twenty against an all-dwellings median, read the count before you treat the dollar figure as the house market. Twenty mixed sales can be a handful of houses and a pile of units.