The Unfinished Pipeline – What NSW’s housing construction data is telling us about future supply
The housing pipeline
NSW is reported to have almost 60,000 townhouses and apartments under construction. On the face of it, that should be good news for future housing supply.
But the number of completed homes is not keeping pace.
That is the issue at the heart of this paper. Housing policy has focused heavily on increasing approvals and getting more projects underway. But the real test is what happens after that.
In NSW, the relationship between construction activity and completed housing has changed materially, particularly for townhouses and apartments. The question is no longer just how many homes are entering the pipeline. It is how many are actually getting through it.
If the number recorded as under construction keeps rising towards 80,000 or 100,000 dwellings without a corresponding lift in completions, that is not a sign of success. It is a sign that more housing is becoming stuck in the pipeline. The priority should be to understand how much of the current pipeline is genuinely deliverable, how quickly it can be completed, and what action is needed to help viable projects reach completion sooner.
The NSW housing pipeline
The charts below track NSW’s housing pipeline for detached homes and for townhouses and apartments. Each shows rolling 12-month approvals, commencements and completions, alongside the stock of dwellings under construction at the end of each quarter.
In NSW, detached housing provides a useful benchmark because the four pipeline measures generally move together. Approvals rise or fall, commencements follow, the stock under construction responds, and completions move in the same direction. Despite lags between stages, the relationship remains intuitive. By early 2026, the major measures were again moving in broadly similar directions, suggesting the detached housing pipeline continues to behave largely as expected.
The townhouse and apartment pipeline tells a different story. The key measures remained relatively close until higher-density construction expanded in the mid-2010s. Since COVID, however, a larger and more persistent divergence has emerged across the pipeline.
Since the major measures of townhouse and apartment activity peaked between 2016 and 2018, approvals, commencements and completions have fallen substantially. Yet the stock recorded as under construction has remained stubbornly high. Historic ABS approval data suggests around 70–80 per cent of townhouse and apartment approvals have historically been located in Greater Sydney, suggesting that much of this issue is likely to be concentrated in the Sydney housing market.
The divergence is clear when measured from each series’ peak
Three of the four measures in the above table have declined from their respective peaks by between approximately 37 and 46 per cent. Under construction has fallen by only around 14 per cent. This divergence raises an important question about future housing delivery and warrants closer examination.
The problem is not unique to NSW—but the imbalance is now most pronounced there
The chart compares quarter-end dwellings under construction with the mean of rolling annual approvals, commencements and completions in each jurisdiction. Between 2019 Q1 and 2026 Q1, the measure increased in four of the five jurisdictions shown. By 2026 Q1, NSW’s under-construction stock was around 118 per cent above this annual average, compared with 104 per cent in Queensland, 99 per cent in Western Australia, 90 per cent in Victoria and 54 per cent in the ACT. NSW had the highest relative level among the jurisdictions shown.
A large pipeline, but a persistent conversion problem
With almost 60,000 townhouses and apartments under construction in NSW, the pipeline is substantial. That is equivalent to around 16% of NSW’s five-year housing target, or almost 80% of the average annual target. On its own, that looks positive.
But a dwelling under construction is not yet a completed home. What matters is how effectively that pipeline is converting into completions.
The current annual rate of completions is equivalent to around 40 completed dwellings for every 100 dwellings recorded in the under-construction stock.
This is not a measure of whether individual dwellings will ultimately be completed. Rather, it shows the pace of completions relative to the size of the construction pipeline.
The relationship changes over time as market conditions and the mix of projects in the pipeline change. A sustained fall in the ratio means annual completions are lower relative to the recorded stock; it does not, on its own, explain why.
Between 2018–2020, annual completions were equivalent to roughly 70 per cent of the under-construction stock. By 2026 Q1, that had fallen to around 40 per cent.
This matters because a large under-construction pipeline can create the expectation that a substantial amount of housing is close to delivery. The data increasingly challenges that assumption and points to either a housing delivery problem, a pipeline data gap or both.
What could be causing this?
The persistence of a large unfinished pipeline from around 2021 coincided with a much more difficult construction environment. The NSW Productivity and Equality Commission’s August 2024 review illustrates the scale of that change. Its modelling for a typical Sydney mid-rise apartment estimated that the construction component increased by 29% between 2018 and 2023, while financing costs increased by 142%. Although illustrative, these estimates show how significantly project economics had shifted.
Projects that were viable when approved may therefore have become harder to finance and complete. The RBA’s May 2024 findings support this explanation, identifying high-rise developments that had stalled or were not proceeding because construction costs were high relative to achievable sale prices, alongside capacity constraints. However, these findings do not establish how long projects had been stalled, whether construction had begun, or whether they were ultimately abandoned.
Although the ABS excludes projects classified as abandoned, the aggregate under-construction figures do not distinguish actively progressing projects from those making little or no progress. Commencement can include site preparation, with all dwellings associated with a building job counted as commenced when that job starts potentially well before substantial building work has occurred. This raises an important question:
How many dwellings recorded as under construction belong to projects with no clear path to completion?
The issue is not just how large the pipeline is, but what is actually happening within it. The table below brings together the potential explanations and why they matter for delivery.
Sources: NSW Productivity and Equality Commission, Review of Housing Supply Challenges and Policy Options for NSW; CIE, Cost and Feasibility Estimates for Supplying Residential Dwellings; Reserve Bank of Australia, Insights from Liaison; ABS, Building Activity methodology. NSW Parliament, Legislative Assembly Committee on Environment and Planning (2026), Historical development consents in NSW, Report 3/58, July 2026.
More approvals but when will they deliver?
The next few years are expected to bring increased housing approvals, supported by the Housing Delivery Authority (HDA), Low and Mid-Rise Housing Policy (LMR) and other planning reforms. This is the increase in activity NSW needs—but approvals are only the beginning of the delivery process.
The scenario below is not a forecast. It tests what could happen if these reforms generate substantially more approvals, the recorded number of homes under construction stays high and the time it takes projects to move from approval through construction to completion does not materially improve.
In this scenario, approvals double by the end of 2027, while annual completions rise more gradually—from around 23,300 to 53,500 by around the beginning of 2031. Over the modelled period, the stock under construction almost doubles, increasing from around 58,700 to a peak of 112,600 dwellings.
The scenario assumes that dwellings currently recorded as under construction are still viable projects. If some of that stock consists of stalled projects with little prospect of proceeding, the headline pipeline may overstate both the active construction workload and the number of homes likely to be delivered.
This uncertainty cuts both ways. New, financially viable projects could progress faster than the scenario suggests if they are not competing with the full recorded pipeline for construction resources. At the same time, the existing pipeline could deliver fewer completions than assumed if a proportion of those projects never meaningfully progress.
An increase in dwellings recorded as under construction should therefore not automatically be read as a coming increase in completions. More projects entering the pipeline is not enough; they need to be progressing towards delivery.
Distinguishing active construction from stalled projects is essential to understanding both the capacity to respond to new approvals and the number of homes the pipeline can realistically deliver.
What are the next steps?
NSW has spent considerable effort freeing up the planning system to achieve more approvals. That work remains important. But the evidence in this paper suggests the housing challenge is no longer only a planning and approvals problem. It is also about delivery.
Understanding the current pipeline should not just tell us what has gone wrong; it should help us understand how to get the next wave of projects through construction and into completed homes faster.
Almost 60,000 townhouses and apartments are currently recorded as under construction in NSW. But the data does not tell us how many are actively progressing, how many are delayed or stalled, or how many are likely to reach completion within a reasonable timeframe.
Under the modelled scenario, where approvals rise sharply but current delivery timeframes remain broadly unchanged, that pipeline could increase to more than 100,000 dwellings.
If that pipeline is being relied upon as an indication of future housing supply, we need a much clearer understanding of what is actually happening within it and, more importantly, what is preventing it from converting into completed homes.
The priorities should now be:
- Get to the bottom of the under-construction pipeline.
Identify which projects are progressing, which are slowing and which are effectively stalled. If this pipeline is being relied on as future supply, we need to know how much of it is actually moving towards completion.
- Use that evidence to support the next wave of projects.
Understanding why current projects are slowing or stalling should inform how new projects are supported once they commence. The objective should be to identify barriers early and prevent viable projects from becoming part of the next stalled pipeline.
- Develop a housing delivery response.
The main barriers have been identified: feasibility, finance, construction costs, labour and capacity constraints, and longer build times. Policy now needs to focus on removing the barriers that are stopping viable projects from being completed.
- Move housing reform beyond approvals.
Planning reform still matters, but approval is only the first step. The next phase of reform needs to focus on getting projects from approval through commencement, construction and completion.
- Track conversion, not just pipeline size.
Monitor how quickly projects move between stages, how long they remain under construction, and whether the rate of conversion into completed homes is improving.
- Keep the Housing Accord target as the measure of success.
The objective is not more approvals or a larger pipeline. It is to deliver the homes required under the Housing Accord. Approvals and commencements are leading indicators; completions are the outcome.
