Approval and commencement volumes for new dwellings have fallen in the most recent quarter’s data release, while the stock of dwellings already under construction — and the value of work still owing on them — continues to climb, and has done so for over a year. Read together, the two trends point to the same conclusion: the constraint on housing delivery this year sits on the finance and construction-capacity side of the system, not on a shortage of approved projects entering it.
Approvals and commencements are slowing
Building approvals fell across three consecutive quarters into early 2026, from 52,159 to 51,308 to 47,727 (Q3 2025 to Q1 2026). Commencements moved more sharply in the same direction, down from a five-year high of 54,542 to 45,342 in the most recent quarter’s data release — a fall of over 16% in a single quarter. Both series had spent 2023 and much of 2024 recovering from a post-pandemic trough before turning down again through late 2025 into 2026.
The chart below sets the last three years of both series side by side. Approvals and commencements move together for most of the period, as would be expected — commencements are approvals that have progressed to site. The divergence in the most recent quarter, where commencements fell further and faster than approvals, is the more unusual feature of the two.
Building Approvals vs Commencements — quarterly, Q1 2023–Q1 2026
Source: ABS 8731.0, Building Approvals and Building Activity, seasonally unadjusted quarterly totals, Australia. Left axis: approvals. Right axis: commencements.
Local detail: approvals by state and LGA
The national approvals figure is an aggregate of very different state-level pictures. Summing local government approvals data across the country for Q1 2026 gives 47,330 dwellings — within 1% of the national ABS figure of 47,727 used above, a useful cross-check that the LGA-level dataset and the national headline are measuring the same underlying activity.
Building Approvals by State, Q1 2026 (dwellings)
Source: APN LGA Intelligence pipeline, local government approvals data aggregated to state level, Q1 2026 (Jan–Mar).
Victoria and New South Wales together account for just over half of national approvals, with Queensland close behind. At the local government level, the picture beneath the state totals is uneven rather than uniformly soft. Brisbane and Melbourne — the two largest city LGAs by approval volume — both recorded higher approvals in Q1 2026 than in the preceding quarter, while several growth-corridor and regional LGAs pulled back sharply over the same period: Ryde, Wollongong, and Sutherland in New South Wales each fell by more than two-thirds quarter-on-quarter, as did Bundaberg and Toowoomba in Queensland.
Falls of that size at LGA level are consistent with the lumpiness inherent in local approvals data — a handful of large multi-unit developments landing in one quarter rather than the next can move a council’s total by hundreds of dwellings without reflecting any broader shift in conditions. The pattern is worth noting rather than over-reading: the national decline is not being driven evenly across the country, and the two largest city centres are, on this measure, not where the softness is concentrated.
But the stock under construction keeps growing
Set against slowing entry into the system, the number of dwellings actively under construction has risen in each of the last five quarters — from 219,000 to 224,000, 230,000, 237,000, and now 243,000, a level not seen since the 2022 peak. The value of work yet to be done on those dwellings — residential building only, not total building activity — has moved the same way, and faster still: from $66.7 billion in Q1 2025 to $80.3 billion in Q1 2026, an increase of over 20% in a year, and the highest reading on record for this measure.
Dwellings Under Construction vs Residential Work Yet To Be Done — quarterly, Q1 2023–Q1 2026
Source: ABS 8752.0, Building Activity, Australia, Table 50, residential-only column, quarterly, Australia total, original series.
New Housing Supply Forecasts (21520) — the certified composite drawing on both series and on residential work done — currently reads at +1.62σ (Q1 2026), with the residential work-yet-to-be-done component alone at +2.53σ, the highest reading in that series’ history since certification began. The two component readings move in the same direction as the raw figures: dwellings under construction are elevated but within the range of their own history; unfinished residential work value is the more unusual of the two, and has been the primary driver of the composite’s rise over the last four quarters.
What that combination means
Fewer approvals and commencements would normally be read as a demand signal — builders and developers responding to softer conditions by starting less work. That reading doesn’t fit here, because the number of dwellings physically under construction, and the dollar value of residential work still owing on them, are both rising at the same time. A demand-driven slowdown reduces both ends of the sequence together: fewer approvals, fewer commencements, and eventually a smaller stock under construction as the existing work is completed and not replaced. What the data shows instead is the opposite pattern at the completion end — volume continuing to move through the system more slowly than new project activity would otherwise justify. That is more consistent with a constraint on how quickly existing work can be completed than with a shortage of new work being started.
The cost and finance context
On the cost side, the ABS Producer Price Index for house construction has climbed steadily since 2011, and the last three years show that climb accelerating rather than levelling off. The index moved from 179.0 to 182.6 between Q1 and Q2 2026 alone.
PPI House Construction, index level — quarterly, Q1 2023–Q2 2026
Source: ABS 6427.0, Producer Price Indexes, House Construction, Australia.
The year-on-year escalation rate on that index — the pace of change rather than the level — has risen for five consecutive quarters: 1.14%, 1.57%, 2.13%, 1.81%, and now 2.50% (Q1 2025 to Q1 2026). That is a steady, gradual upward trend, well short of the double-digit escalation the same series recorded through 2022, and closer in character to a slow re-acceleration than an acute shock.
ASIC recorded 2,361 construction industry insolvencies in FY2024–25. A FY2025–26 figure is not yet available from ASIC; this brief does not estimate one, and none should be inferred from the absence.
Construction Finance & Capacity (21530) — covering land development lending limits and business lending to the construction sector — currently reads with a composite Z of +2.13σ, with its two component metrics at +2.95σ (development lending limits) and +2.61σ (business lending). Both sit well above their historical average, and both have been elevated for some time rather than moving sharply in the latest quarter alone. These figures were calculated under APN Cortex I, the platform’s earlier analytical methodology; the remainder of this brief draws on the current APN Codex/Cortex II pipeline. They are due to be rebuilt onto that pipeline by the end of the year, and should be read as directional rather than final in the meantime.
Construction Costs & Supply Chain (21260) currently reads at +1.371σ (Q1 2026, all three components present, including ASIC’s quarterly construction insolvency series; the reading is higher than first published because that input was repaired, not because conditions deteriorated). That reading sits below the elevated readings in Construction Finance & Capacity (21530) and New Housing Supply Forecasts (21520) above, consistent with this piece’s reading that finance and delivery capacity — not materials-cost pressure — are the binding constraint this year.
Reading the combination
Three things are true at once in the current data. First, moderate and steadily rising input costs — a five-quarter upward trend in the 1–2.5% range, not a shock. Second, construction-sector lending running well above its historical average on both measures tracked here, and doing so persistently rather than as a single-quarter spike. Third, a construction stock that keeps growing even as new approvals fall, with the value of unfinished residential work reaching its highest point on record.
Together, these point toward finance and delivery capacity as the more likely constraint on how quickly approved housing reaches completion this year, rather than a slowdown in demand for new dwellings or an acute materials-cost shock. New Housing Supply Forecasts (21520), the only fully certified composite among the three nodes discussed here, is consistent with that reading: it is elevated because the volume of unfinished residential work is elevated, not because new activity has contracted. Whether that constraint eases through the remainder of 2026 will depend in large part on construction-sector lending conditions — the one input in this data where the current reading is both the most persistent and the least resolved.
Findings are presented on the basis of data and evidence alone.
