Employment in detail

By Leith van Onselen

As reported by the Houses & Holes earlier, the Australian Bureau of Statistics (ABS) has released labour force data for the month of August, which reported a slight reduction in the headline unemployment rate to 5.1% from 5.2% in July, but a decrease in the total number of jobs across the economy.

The economy shedded 8,800 jobs (-0.1%) over the month , with the loss of 9,300 part-time jobs partly offset by a 600 increase in full-time jobs. Last month’s result was also revised down, from a gain of 14,000 jobs to 11,700 jobs.

The labour force participation rate fell, down -0.2% to 65.0%, which explains the reduction in the headline unemployment rate in the face of job losses (see below charts).

The employment market continues to track sideways, which is a reasonable result given the state of the global economy (see below chart).

Nevertheless, there are some significant ‘flies in the ointment’ in this release, which contradict the rosy headline result.

Importantly, employment growth outside of Western Australia has been poor, with minimal jobs growth achieved over the past year. In fact, Western Australia has added 88% of the nation’s jobs over the past year – a highly unbalanced result (see below chart).

Moreover, the aggregate number of hours worked has fallen in every mainland state except Western Australia (see below chart).

Reflecting the above, Western Australia’s unemployment rate is well below the national average, with New South Wales also holding-up well (see below chart).

Nationally, the aggregate number of hours worked continued its recent downtrend, falling -0.4% in August:

On a positive note, the underemployment rate fell slightly in the August quarter, down -0.2% to 12.4%, with falls in Western Australia and New South Wales  partly offset by rises in the other mainland states (see below chart).

Despite the fall in the headline unemployment rate, this is a mediocre report with the economy shedding jobs and the aggregate number of hours worked declining, in addition to the lion’s share of jobs growth confined to just one state – Western Australia – where iron ore prices are now falling heavily.

Twitter: Leith van Onselen. Leith is the Chief Economist of Macro Investor, Australia’s independent investment newsletter covering trades, stocks, property and yield. Click for a free 21 day trial.

Leith van Onselen
Latest posts by Leith van Onselen (see all)


  1. So what happens if the we change the unemployment definition to 10 hours a week worth of work for 1 hour. Do we have access to the data that shows how many people work X hours. EG 50,000 people worked 20 hours week…

    • That would upwardly raise the unemployment rate but it might not affect the change in the unemployment rate.
      It must be accepted that the ABS survey is conducted in a consistent manner regardless of the definition of unemployed, and thus reflects changes within the statistical confidence levels.
      The one hour rule is the international standard which the ABS adopts.

        • The point is that there’s an internationally accepted standard that the ABS is reporting on. The underutilisation rate captures some of the additional information – how many would like to work more – but the unemployment picture is nuanced.

      • dumb_non_economist

        Temple Drake, sorry, but I disagree. The work force has changed a huge amount over the last 30+ yrs. Go back to the early 80s and part-time/casual workforce was mainly limited to the hospitality arena, now it’s pervasive throughout the entire workforce. We have far more part-time casual workers and that is reflected with a bastardised UE rate.

        • Isn’t that the same across the world?

          We need to use the same method for calculating unemployment as other countries, in order for results to be comparable.

          • We need to use the same method for calculating unemployment as other countries, in order for results to be comparable.

            Calculating a number to compare with the rest of the world does not preclude also calculating an additional number that is meaningful.

          • dumb_non_economist


            what I’m getting at is we can no longer compare the UE rate now of 5 odd % with a UE rate of 5% back 30 yrs due to the structural change in employment ie casual/part-time employment.

          • dumb_non_economist

            Yes, because a UE rate has a meaning, which is derived from our past experiences of UE rates and their effect. So we judge rates as good/bad and how we are traveling on the past and to my mind they bear no resemblance.

  2. Important fact: every day that passes, around 250 workers turn 65.

    If they still have a mortgage or their super balance is low, they might keep working a bit longer, but that will only postpone the inevitable.

    • Perhaps your just like us in New Zealand, after all! “A resilient performance by people over the age of 55 is a notable feature of current labour market dynamics, the Ministry of Business, Innovation and Employment report says, noting participation rates by older people are “very high” by international standards.”

    • Yeah, I’ve raised this as well – but then the question would be what is the participation rate measured against; if it looks at working age (15-65) then those people would fall out of the participation rate numbers too.