Homes under construction
is not homes being delivered.

Construction pipeline capacity is the number of jobs that can be in production at once, multiplied by how often each of those positions turns over. Concurrent capacity is set by supervision, working capital, trade access and, for licensed residential builders, the total contract value a warranty underwriter permits under construction at any one time. Turnover is set by build duration. Capacity is the product of the two, which makes it a rate rather than a quantity.

That distinction is why homes under construction and homes completed can move in opposite directions without either figure being wrong. Under construction counts occupancy at a moment. Completions count turnover over a period. When each job holds its position longer, occupancy rises and completions fall, and the pipeline looks fuller while delivering less. The Australian Bureau of Statistics publishes both series, and reading only the first is how a congested pipeline gets described as a busy industry.

Why it matters

Because the same confusion runs inside a single building business, where it is more expensive. Work in progress rising looks like growth. More open jobs looks like a good year. Both are occupancy measures, and neither tells a builder whether the business is actually delivering more houses or simply holding more of them. The arithmetic on this page is the same at every scale, and the useful half is the small one: it tells you how many jobs you can carry and which lever changes that number.

The published inputs are straightforward. On ABS figures, average completion time for new private houses rose from 2.2 quarters in September 2019 to 3.3 quarters by June 2024, an increase of 1.1 quarters per house. At the national peak in March 2023 there were 104,315 private new houses under construction, against a pre-pandemic normal near 70,000 (70,306 in June 2018). More houses were open than usual, and each was holding its position about a third longer than it would have before 2020.

Apply the second figure to the first across the three years from late 2021 to mid-2024 and the additional slot occupancy is equivalent to roughly 163,000 dwellings of throughput capacity.

What that figure is, and what it is not

This is our own derivation from ABS inputs. The ABS does not publish it, and it should never be quoted as an ABS figure. The inputs are public and checkable; the calculation and the framing are ours. Anyone using the number should say so, and should carry the two qualifications below with it, because without them it is routinely misread.

• It measures displaced throughput, not missing homes. It describes how much completion capacity was absorbed by slower pipeline turnover across the national detached housing stock under construction in that period. It is not a count of homes that would otherwise exist, and it is not a shortfall in the ordinary sense. The same homes moved through the system more slowly.
• It spans every cause, not one. Materials lead times, labour availability, insolvency and weather all lengthened build durations over the same period, and no part of this figure is attributable to any single one of them. It is a pipeline efficiency loss, and it is silent on why the efficiency was lost.

For context on scale, the National Housing Accord targets 1.2 million homes over five years from July 2024, roughly 240,000 a year, and national dwelling completions on ABS data have been running well below that pace. The reasonable observation is that at the worst point of the blowout, slower turnover was consuming productive capacity on a scale comparable to the annual gap. That is an interpretation offered as a way of reading the data, not a finding, and it does not convert displaced throughput into houses that were never built. All of these figures are point-in-time and should be checked against the current ABS release before use.

This is the half that is actually usable, because a builder can act on it. The model has three inputs and one line of arithmetic.

1. Concurrent capacity. How many jobs the business can genuinely have open at once. Work it out from the constraint that binds first, which for most residential builders is working capital or supervision rather than demand.
2. Average build duration. Commencement to practical completion, in years, taken from your last several completed jobs rather than from your contract build period.
3. Annual throughput. Concurrent capacity divided by duration. Six open jobs on twelve-month builds is about six completions a year. The same six jobs on fifteen-month builds is about 4.8.

That last comparison is worth sitting with. Nothing changed except duration, and annual output fell by a fifth. To recover the lost completions by adding work instead, the business would need seven and a half concurrent jobs rather than six, which means more supervision, more working capital and more warranty capacity, each of which takes quarters to acquire. Removing three months from the average build achieves the same result with resources already in the business.

Where the warranty ceiling comes in

For licensed residential builders the arithmetic has a hard outer edge. Home warranty underwriting caps the total contract value permitted under construction at any one time, commonly discussed as Open Job Value. That cap applies to occupancy, not to turnover, so a rough reading of annual revenue capacity is the cap divided by average build duration in years. A builder capped at six million dollars with twelve-month builds can turn over around six million a year. The same builder with fifteen-month builds is closer to 4.8 million, against an unchanged cap and an unchanged balance sheet.

Treat that as a model rather than an underwriting rule. Schemes differ by state, underwriters assess differently, contract value is not consumed evenly across a build, and no cap behaves exactly like a simple divisor. The point that survives all of those qualifications is directional and it is the one worth acting on: under a fixed financial ceiling, build duration is the variable that sets revenue, and it is the one a builder can change without asking anybody's permission.

The operator's observation is that almost every residential builder who has come unstuck growing did the same thing, and did it for good reasons. Work was available, clients were ready, the pipeline report looked strong, and each new job was signed because turning it down felt like leaving money on the table. What none of them had was a number for how many jobs the business could carry, so there was never a moment where the next contract was visibly the one too many. The overload arrived as a feeling before it arrived as a figure, which is far too late.

Builders who avoid that hold two numbers and check them monthly: how many jobs are open, and what stage each one is at. The second number is the one that does the work. Concurrent capacity is not a count of jobs, it is a count of jobs at a stage, and a business that can comfortably run six houses will fall over with four if three of them hit frame in the same fortnight. Staggering commencements against the constraint stage is the cheapest capacity a builder will ever buy.

The third discipline is measuring your own duration honestly, from commencement to practical completion rather than from the day the client thinks it started to the day you consider it finished. Most builders are surprised by their own number the first time they calculate it, usually because the completion tail is longer than anyone remembers. That figure drives the entire capacity model, so it is worth being unsentimental about.

Where software fits the workflow

None of this can be seen one job at a time. A builder looking at a single programme sees a job; the capacity picture only appears when every open job is on the same understanding of stage, cash and duration at once. In VIABUILD the construction schedule holds every job against its baseline, so open jobs, current stages and drift are visible together, and cash flow forecasting shows what each of those slots is tying up and when it comes back. That turns capacity from a feeling into a number the business can decide against, which is the whole point of the exercise. The decision about whether to sign the next job stays with the builder.

Pipeline figures are quoted heavily in public discussion and are easy to misread. The points below are labelled by evidence class so a reader can see which are measurements, which are conventions and which are judgement.

• Measurement. The ABS publishes commencements, completions and work under construction quarterly in Building Activity, Australia, and average completion time in related articles. Under construction is a point-in-time occupancy count. Completions are a flow over a period. The two answer different questions.
• Measurement. The figures used here, 2.2 quarters rising to 3.3 quarters between September 2019 and June 2024, and a pipeline peak of 104,315 houses under construction in March 2023 against 70,306 in June 2018, are ABS figures current at the time of writing. They are point-in-time and will date.
• Derived, not published. The figure of roughly 163,000 dwellings of displaced throughput capacity is our calculation from those ABS inputs. It is not an ABS statistic, it measures throughput displaced across all causes rather than homes that would otherwise exist, and it should not be cited without both qualifications.
• Convention. Approvals are the weakest of the pipeline measures for anything to do with delivery, because an approval is permission to proceed rather than a house being built. Commencements, completions and time from approval to occupancy describe delivery; approvals describe intent.
• Professional recommendation. Home warranty capacity is assessed by scheme underwriters under rules that differ by state and territory, and any figure a builder models against their own cap should be confirmed with their broker or underwriter rather than taken from a general reference like this page.

Illustrative only, not a benchmark. A custom home builder runs six jobs at a time and is capped by their warranty underwriter at six million dollars of contract value under construction. Average build duration, measured properly from commencement to practical completion across their last eight houses, is twelve months. Annual throughput is about six houses and turnover sits near six million dollars. The business is stable, the supervisor is stretched but coping, and cash arrives at a rhythm the office has learned to plan around.

Three years later the same business is running the same six jobs under the same cap, and average duration has drifted to fifteen months. Nobody decided this. It accumulated from longer waits on windows, slower client decisions, a wet run and an inspection queue that never recovered. Annual throughput is now about 4.8 houses and turnover is closer to 4.8 million dollars, with the same overhead, the same supervisor and the same office. Nothing was lost commercially; every job still gets built and every contract is still honoured. The business simply produces a fifth less each year.

Faced with that, the instinctive fix is to sign more work. It cannot be done under the cap, and if the cap were lifted it would require more supervision and more working capital to absorb, both of which take a year to arrange. Taking three months back off the average build restores the original output using nothing the business does not already have. Run the same arithmetic across a hundred thousand open sites and you have the national picture above. Only the number of slots differs.

Pipeline capacity (concurrent capacity multiplied by turnover), concurrent capacity (how many jobs can be open at once), throughput (completions delivered over a period), slot (one position in the pipeline, occupied for the whole of a job's duration), work in progress (value of started but unfinished work), under construction (a point-in-time occupancy count), completions (a flow measure over a period), commencement (the start reference point the ABS measures from), Open Job Value (the total contract value a warranty underwriter permits under construction at once), constraint stage (the stage where jobs bank up and where capacity is actually set). Definitions for the wider vocabulary live in the construction glossary.

The financial ceiling this page keeps running into is its own subject; the next reference is construction working capital.

• Australian Bureau of Statistics, Building Activity, Australia , the quarterly source for commencements, completions and dwellings under construction. Pull the current release rather than quoting any figure here.
• Australian Bureau of Statistics, Home Building Through the Pandemic , the source of the completion-time and pipeline-peak figures used in the derivation on this page.
• The Australian Government's National Housing Accord, published through the Treasury, for the 1.2 million homes over five years target referred to above.
• Your home warranty scheme underwriter or broker, for how your own capacity ceiling is assessed in your state or territory. The rules differ, and only your assessment applies to your business.
• Your own completed-job records, for average build duration from commencement to practical completion. Every number in the business-scale model on this page depends on that one figure being honest.