Lower than any quarter before 2021
In August 2026 Australia had 2.2 unemployed people for every job vacancy. The ABS counted 325,000 vacancies and 722,900 unemployed people, seasonally adjusted; the ratio is the desk’s.
Set against the whole series, that is low. In the 162 quarters with a figure from 1979 to 2020, the lowest ratio was 2.5, in February 2008. All 21 quarters below 2.2 came after May 2021, and the lowest of all was 1.1, in August 2022.
Before the pandemic the ratio was higher still. In August 2019 Australia had 3.2 unemployed people per vacancy, with 221,900 vacancies; vacancies are now 46% above that.
Rising, from a long way down
The ratio has roughly doubled in four years. A year earlier, in August 2025, it was 2.0. Since then the number of unemployed has risen by 80,000, from 642,900 to 722,900, while vacancies went from 329,500 to 325,000.
So the market is loosening because more people are looking for work, not because vacancies collapsed. Vacancies are 31% below their peak of 472,700 in May 2022, and still above the highest level before 2021, 260,000 in November 2020.
State by state, the picture against 2019 is the same in all but one. In seven of the eight states and territories, the ratio is lower than in August 2019; Queensland went from 4.3 to 2.0 and South Australia from 5.8 to 2.3. The exception is the Australian Capital Territory, from 1.3 to 1.9, where vacancies went from 7,600 to 6,100: a fall smaller than the survey’s margin of error.
| State or territory | Unemployed per vacancy, August 2019 | Unemployed per vacancy, August 2026 | Vacancies, August 2019 (thousands) | Vacancies, August 2026 (thousands) |
|---|---|---|---|---|
| Australia | 3.2 | 2.2 | 221.9 | 325.0 |
| Tasmania | 5.1 | 2.9 | 3.6 | 5.4 |
| Victoria | 2.9 | 2.6 | 61.2 | 77.5 |
| South Australia | 5.8 | 2.3 | 11.3 | 19.9 |
| New South Wales | 2.6 | 2.1 | 72.8 | 96.5 |
| Queensland | 4.3 | 2.0 | 39.1 | 69.5 |
| Western Australia | 3.1 | 2.0 | 26.4 | 40.9 |
| Australian Capital Territory | 1.3 | 1.9 | 7.6 | 6.1 |
| Northern Territory | 2.0 | 1.3 | 4.3 | 5.8 |
What the ratio counts
It divides one survey by another. The unemployed come from the Labour Force Survey: people without paid work, available and looking. The vacancies come from the Job Vacancies Survey: jobs available for immediate filling on one day, for which the employer has started to recruit.
It is not the number of applicants per job. An unemployed person can apply for many vacancies or for none, and many vacancies are filled by people who already have a job. The ratio compares the size of two pools.
And the states are noisier than the nation. The ABS publishes a standard error for each vacancy estimate; for the Australian Capital Territory it is 700 in both years, so its fall of 1,500 is within the survey’s margin of error. The state figures are original estimates, compared August with August.
- UNEMPLOYED PEOPLE, from the Labour Force Survey: people without paid work who are available and actively looking, or about to start a job, counted where they live.
- JOB VACANCIES, from the Job Vacancies Survey: jobs available for immediate filling on one reference day, the third Friday of the survey month, for which the employer has taken recruitment action. Not adverts, and not jobs that exist.
- A RATIO OF TWO POOLS, not applicants per job. An unemployed person can apply for many vacancies or for none, and a vacancy can be filled by someone already in work. A lower ratio is a tighter market: a smaller pool for each opening.
So 2.2 does not say that two people apply for each job: it says that in August 2026 Australia had 2.2 people without work and looking for every job an employer was trying to fill on the survey day, and that from 1979 to 2020 that number never went below 2.5.
Two surveys, one divided by the other by the desk, and a reading that forgets what each counts says what the source does not.
What these figures do not say
They name no cause. Interest rates, migration, the mining sector and artificial intelligence have all been offered to explain Australia’s labour market, and none of them is measured here.
They do not say that jobs were lost when vacancies fell. A vacancy is a job an employer is still trying to fill; fewer vacancies are fewer unfilled jobs. Nor do they say that anyone moved from one state to another.
They forecast nothing, and the series starts in 1979: nothing here is a claim about any earlier year.
