More people working, higher unemployment: reading August's jobs figures

Australia added employed people in August, but its unemployment rate also rose. That combination can look contradictory when two headlines appear beside each other. It is possible because the labour market is not a fixed group of people moving between two boxes.

The Australian Bureau of Statistics' August labour-force release, published on 24 September 2026, puts seasonally adjusted employment at 14,836,600, up 39,500. The unemployment rate reached 4.6%, while participation rose to 67.1%.

Those figures describe different parts of the same picture. Reading them together is more useful than choosing whichever headline sounds most reassuring or alarming.

Start with the people being counted

The ABS labour-force framework separates people into employed, unemployed and outside the labour force. Employment and unemployment together form the labour force. A person without a job is not automatically counted as unemployed: active job search and availability matter, subject to the survey's detailed rules.

This distinction means the unemployment rate is not the percentage of all Australians without paid work. Children are not part of its denominator, and neither is every adult who is retired, caring for family or studying without seeking work. The number asks a narrower question about people participating in the labour market.

It also counts people, rather than every vacancy or individual job. A person with several jobs does not become several employed people in this headline measure. That is why wording such as 'employment increased' is preferable to assuming every change represents a newly created position.

A small example makes the arithmetic clearer

Imagine a simplified community with 1,000 people in its labour force. Of these, 950 are employed and 50 are unemployed. Its unemployment rate is 5%.

Now suppose another 100 people enter the labour force. Sixty find work and 40 are still looking and available. Employment has increased to 1,010, an improvement of 60. Unemployment has also increased, to 90 out of a labour force of 1,100. The unemployment rate is now about 8.2%.

These are invented numbers used only to explain the calculation. They are not estimates for Australia, and they do not identify what caused August's movement. They show why rising employment and a rising unemployment rate can coexist without either statistic being wrong.

The example also explains why participation deserves attention. If we ignore people entering or leaving the labour force, we can misread the meaning of a changing unemployment rate. The size of the group matters alongside the number finding work.

Look beyond the headline total

In the August release, part-time employment increased by 45,800 and full-time employment fell by 6,300. Seasonally adjusted monthly hours worked rose by 14 million to 2,009 million. The underemployment rate was 6.2%.

These measures answer questions that an employment total cannot settle on its own. Someone may have a job but still want more hours. Another person may deliberately prefer part-time work. It would be a mistake to treat every part-time position as a problem or every increase in the employment total as proof that households have the hours they need.

For a reader trying to understand their own situation, a useful question is: what would count as an improvement for me? It might be finding any paid work, securing predictable shifts, increasing hours or moving into a role that uses a qualification. Those are different goals, even when they all sit beneath a national employment headline.

One release cannot explain every change

The ABS flags a change to supplementary survey collection that may affect the August seasonally adjusted figures. It says the data remain fit for purpose and recommends trend figures when assessing underlying labour-market behaviour.

That is a reason to read the release notes, not a reason to dismiss the results. A survey estimate and an explanation for that estimate are separate things. The headline does not, by itself, establish that one policy, one industry or one group of people caused the movement.

A sensible reading routine is to check the reference month, the release date and whether the figures are seasonally adjusted or trend. Keep the same measure when comparing periods. If a social post switches between series without saying so, its conclusion may be stronger than the comparison supports.

What a job seeker can take from this

Use the national figures as context, then build a more specific picture of your search. Keep a short record of suitable vacancies, applications, interviews and the reasons employers give when they provide feedback. Review that record over several weeks rather than judging the whole market from one difficult afternoon.

If applications produce no interviews, review how clearly they address the role's requirements. If interviews do not produce offers, consider practising the questions that have been difficult. These are practical ways to examine a search, not promises of an outcome or explanations for why any particular person is unemployed.

Employers can take the same approach from the other direction. Before interpreting a national number as evidence that recruitment should be easy, examine the pay, location, hours and skills required in the actual vacancy. A national average cannot tell a business whether its particular offer is competitive.

August's release gives us a shared set of measurements. It does not replace the experience of someone looking for work. The useful next question is how employment, participation and available hours move together over time, and how that evidence compares with the part of the labour market you are actually dealing with.