Between the 2016 and 2021 Censuses, the share of Australian households renting their dwelling barely moved — 31.8 per cent to 31.1 per cent of households stating a tenure. Outright ownership held near-flat too, 31.8 per cent to 31.4 per cent; mortgaged ownership actually rose fractionally, 35.4 per cent to 35.5 per cent. Read at the national level, the intercensal period looks like a non-event for tenure mix. It was not. The stability is compositional, and the composition is the story: across 517 comparable local government areas, 455 recorded a decline in renter share over the period; only 62 recorded an increase. The median LGA moved 2.6 percentage points toward ownership; the mean moved 3.5 points. The national figure holds level only because the small number of LGAs that moved toward renting are, disproportionately, among the largest in the country by dwelling count — their weight in the national aggregate outruns their weight in the LGA count.
Full sample, 517 LGAs. Gold points mark the metropolitan, amenity-wave and mining/remote cohorts discussed below; grey points are the remaining sample.
The shape of the drift
The distribution beneath the national average is heavily one-sided. Of the 517 LGAs in the sample, 452 recorded a fall in renter share (rounding accounts for the small difference from the 455 cited above against the unrounded series), three were unchanged to one decimal place, and 62 moved toward renting. The bulk of the sample sits in a narrow band either side of the median: 299 LGAs recorded a shift between −4 and 0 percentage points, a broad, unremarkable drift toward ownership that describes most of the country. The tail is where the analytical interest concentrates — on the ownership side, 26 LGAs moved more than 10 percentage points, all of them mining or remote-community LGAs discussed below; on the renting side, the largest positive shift in the entire sample is Burwood at 4.7 points, an order of magnitude smaller than the largest negative shifts. The distribution is not merely skewed toward ownership; it is asymmetric in scale, with the ownership tail reaching roughly seven times further than the renting tail.
Bin width 2 percentage points. Full sample of 517 LGAs; the three tail bins below −18pp (Menzies, Wiluna, Leonora) are grouped into the leftmost bar shown.
Renting concentrated into metropolitan corridors
The LGAs that shifted toward renting form a recognisable geography. Of the twenty-five largest positive shifts, eighteen sit in Greater Sydney — Burwood, Parramatta, Ryde, Strathfield, Penrith, Liverpool, Fairfield and Cumberland among them — joined by the City of Perth, which recorded the second-largest positive shift in the country at 4.1 points on 12.1 per cent population growth. Several of the Sydney LGAs recorded the shift on comparatively modest growth: Fairfield’s renter share rose 2.6 percentage points on population growth of only 4.9 per cent, and Canterbury-Bankstown, Georges River and Sutherland Shire all posted positive shifts on single-digit growth. The pattern is consistent with rental demand concentrating into established, already-dense metropolitan corridors through the period, arriving as intensification within the existing dwelling stock rather than distributing evenly with population growth across the metropolitan footprint.
The amenity wave arrived as ownership
The high-growth coastal and lifestyle LGAs of the intercensal period moved in the opposite direction, and moved further than the metropolitan corridors moved the other way. Augusta-Margaret River grew 17.8 per cent and its renter share fell 5.2 percentage points; Bass Coast grew 24.3 per cent and fell 4.9; Queenscliffe, Noosa, Fraser Coast, the Gold Coast, Surf Coast and the Sunshine Coast all combined above-median population growth with a contracting rental share. Population growth in these markets was absorbed as owner-occupation, not rental supply. As a share of stock, dwellings moved out of the rental pool at precisely the destinations receiving the strongest inflows — a compositional fact that preceded, and plausibly contributed to, the rental tightening subsequently recorded in several of these markets through the post-2021 period.
Eight representative LGAs per cohort, selected by magnitude of shift within each group. Full underlying data in the §6 SQL.
The relationship with growth is real but weak
Across the full 517-LGA sample, the correlation between population growth and renter-share shift is r = 0.335 — positive, but weak to moderate. Faster-growing LGAs were somewhat more likely to shift toward renting, at roughly 0.18 percentage points of shift per percentage point of growth, from a baseline drift of about four points toward ownership. Correlation of this magnitude explains barely more than a tenth of the variance in outcomes; growth rate is a minor input into tenure-mix direction, not a determining one.
The residual spread either side of the trend line is the analytically interesting object, and it runs in both directions. On the positive-residual side, several outer-growth-corridor LGAs in Melbourne — Wyndham (34.5 per cent growth, +2.0pp), Casey (22.0 per cent, +1.7pp) and Melton (32.1 per cent, +0.9pp) — absorbed strong growth with rental share holding or rising slightly, a different mechanism to the Sydney/Perth infill pattern but a similar direction. On the negative-residual side, several high-growth Western Australian mining and resource towns recorded large ownership shifts despite strong population growth: Ravensthorpe grew 20.3 per cent yet its renter share fell 13.5 points, and Mount Magnet grew 35.5 per cent yet fell 12.2 points. These sit far below the trend line — substantial population growth converted almost entirely into owner-occupation, the opposite of the naive expectation that growth and rental demand move together. Correlation does not establish causation, and reading the exceptions is where the compositional story is actually told.
A separate story in remote Australia
The largest movements toward ownership in the entire sample are not mainstream markets, and the mechanisms behind them are distinct enough to warrant separate treatment rather than folding into the metropolitan or amenity narratives above. Mining LGAs recorded double-digit falls in renter share as workforce-housing arrangements were restructured from their 2016 configuration: Karratha (−16.2pp), Port Hedland (−13.7pp), East Pilbara (−10.9pp) and Isaac (−9.1pp) are the largest of this cohort. The single largest movements in the entire 517-LGA sample, however, are remote Indigenous-community LGAs in Western Australia and the Northern Territory: Menzies (−32.6pp), Wiluna (−29.2pp), Leonora (−26.4pp, low-confidence flagged), Unincorporated NT (−24.4pp) and MacDonnell (−20.3pp) head a cluster of falls an order of magnitude larger than anything recorded in mainstream markets. These substantially reflect changes in how community and government housing tenure was recorded between the two censuses rather than a behavioural shift in household tenure choice, and are excluded from the mainstream-market reading developed above.
Methodology Note
Renter share is calculated against households stating a tenure type — R_Tot_Total ÷ (Total_Total − Ten_type_NS_Total) — controlling for the fall in tenure non-response between censuses (2.71% of dwellings in 2016 to 1.47% in 2021). 2016 figures are ABS Census GCP data apportioned to 2021 LGA boundaries via a validated Mesh Block transfer matrix, routed through the ratified 2016 G33 ↔ 2021 G37 concordance (72 of 72 variables matched by title); 2021 figures are native. Three small territories with no native 2016 source (Christmas Island, Cocos Islands, Darwin Waterfront Precinct) are excluded, along with pseudo-codes, Unincorporated Other Territories, and LGAs below 500 persons in 2021. Full sample of 517 LGAs shown across all three charts; correlation r = 0.335; correlation does not establish causation.
Five LGAs carry a low-confidence flag on at least one 2016 tenure variable (Camden, Campbelltown NSW, Hope Vale, Leonora, and one further LGA) and are retained in the full sample. Mining and remote-community LGAs are shown in all three charts and discussed separately above; their movements reflect distinct mechanisms — workforce housing cycles and intercensal changes in tenure/landlord classification — rather than the mainstream ownership-drift pattern that describes most of the country.
