The Approval Gap: Population Growth and Building Approval Intensity Across 70 Australian LGAs
Across the 70 LGAs covered by APN LGA Intelligence — together home to 61.8 per cent of Australia’s population — building approvals track population growth about as tightly as two independently measured things ever do: a correlation coefficient of 0.84 between 2024–25 population growth and dwelling approvals per 1,000 residents over the trailing 12 months to April 2026. Faster-growing LGAs approve more housing, on average, in something close to a straight line.
The story here isn’t that relationship — it’s the 24 LGAs that break it. Eleven established, largely built-out suburbs approve well below what their (modest) growth would predict. Thirteen others, mostly regional cities and middle-ring redevelopment corridors, approve well above it. Both groups are hidden inside an aggregate correlation that, taken alone, would suggest the system is working roughly as intended.
The Overall Relationship
Regressing approval intensity on population growth across the 70 LGAs produces a fitted line of dwellings/1,000 ≈ 2.37 + 2.96 × growth% (r = 0.844, r² = 0.712). Growth explains roughly 71 per cent of the variance in approval intensity — a genuinely strong relationship by the standards of LGA-level data, and the starting point this analysis works from rather than argues against. The scatter below shows that relationship directly, with the 24 LGAs departing from it by 1.75 dwellings per 1,000 residents or more picked out in gold.
The Shortfall Cluster
Eleven LGAs approve at least 1.75 dwellings per 1,000 residents fewer than their population growth predicts. Nine of the eleven carry an IRSAD decile of 9 or 10. Northern Beaches is the sharpest single case: 0.5 per cent population growth against 1.89 approved dwellings per 1,000 residents — the lowest approval intensity anywhere in the 70-LGA set, in an LGA with no capacity constraint visible elsewhere in the data (IRSAD decile 10, unemployment 3.2 per cent, the lowest of any LGA analysed here). Gosnells (decile 6), Blacktown (decile 8) and Greater Dandenong (decile 3) are the exceptions that keep this from being read as a clean advantage story — and the correlation between IRSAD decile and approval intensity across the full 70-LGA set is close to zero (r = −0.10), too weak to support a general claim that affluent LGAs build less. What can be said is narrower: within this specific shortfall cluster, high-advantage established suburbs are over-represented relative to their share of the full sample.
The Redevelopment Cluster
Thirteen LGAs approve at least 1.75 dwellings per 1,000 residents more than growth predicts. Melbourne leads by a wide margin — 3.7 per cent population growth against 19.85 approved dwellings per 1,000 residents, a residual of +6.52, the largest departure from prediction anywhere in the set, consistent with an apartment-heavy, high-churn LGA where approvals reflect ongoing redevelopment turnover rather than net new population capacity alone. This cluster mixes two mechanisms this dataset can’t separate: outer-growth LGAs where approvals may be running ahead of population figures yet to catch up (Wanneroo, Moreton Bay), and established or middle-ring LGAs where infill and knockdown-rebuild activity generates approvals without a matching net population gain (Charles Sturt, Banyule, Salisbury).
What This Doesn’t Explain
Approvals are not completions, and this dataset doesn’t distinguish a development approval from a dwelling that will exist in eighteen months’ time, still less one that will exist and be occupied. Approval intensity is also, by construction, a trailing indicator measured against a point-in-time growth figure; an LGA with a large approval pipeline now may be responding to growth from two or three years ago, not the 2024–25 figure it’s plotted against here. This analysis covers the 70 LGAs currently covered by APN LGA Intelligence, not all 547 LGAs nationally.
Close
An aggregate correlation of 0.84 between growth and approvals is a strong relationship, and on the evidence here, the housing approval system in most of these 70 LGAs is doing roughly what a simple model would expect. The residual is where the story sits: eleven established suburbs, most of them high-advantage and none of them visibly capacity-constrained, approving well below what their own modest growth would predict; and a second cluster, mixing outer-growth corridors with middle-ring redevelopment areas, approving well above it. Neither cluster is visible in a state- or national-level approvals number. Both are visible here.
Methodology Note
Dwelling approvals sourced from ABS 8731.0, LGA small area data, new residential, total sectors (downloaded 2026-07-08), summed over the trailing 12 months to the latest available month (May 2025–April 2026). Approval intensity calculated as approved dwellings per 1,000 residents, using 2025 Estimated Resident Population as the denominator. Population growth is the ABS 2024–25 ERP change percentage as carried in APN LGA Intelligence. Scope is the 70 LGAs currently covered by APN LGA Intelligence (Bedrock composite populated), together accounting for 61.8 per cent of Australia’s population (16,765,118 of 27,126,526 residents summed across all 546 LGAs, ABS Regional Population 2024–25, released 31 Mar 2026); all 70 had matching approvals data, with no NIL RETURNs required. Regression is ordinary least squares, dwellings-per-1,000 on growth percentage (slope 2.961, intercept 2.370, r = 0.844, r² = 0.712, n = 70). IRSAD-vs-approval-intensity correlation across the same 70 LGAs is r = −0.10, reported specifically to bound the claim made in “The Shortfall Cluster” above — it is not strong enough to support a general finding that higher-advantage LGAs build less, only that this particular flagged cluster skews that way. The 1.75 dwellings-per-1,000 flagging threshold is an analytical choice made for this piece, not a previously ratified APN threshold. Findings here describe the 70-LGA sample analysed only and should not be read as representative of the full national distribution of Australia’s 547 Local Government Areas.
