When looking at the Labour force survey for August 2015, I noticed the seasonally adjusted figure showing the number of employed people in July (they show the month before) was 52,400 lower than had been quoted in the July release – the release which caused Joe Hockey to crow about creating 38,500 jobs in the month. How could this be?
A few phone calls and some reading later I got my answer. (And thanks to the people at the ABS who are always happy to help despite funding cuts).
In February 2015 the ABS went back and revised employment estimates for the previous 44 months (back to the 2011 census) based on the population benchmarks from the Australian Demographic Statistics, Jun 2014. They were then to continue this practice of revision on a quarterly basis (leaving out May 2015 – I assume the government was too busy to give them the info) revising estimates for the previous 25 months each time. I do not know if they were directed to do this or whose decision it was.
When this new approach was introduced in February the accompanying notes said
“Changes to the population benchmarks impact primarily on the level of the Labour Force estimates (i.e. employed, unemployed and not in the labour force) that are directly related to the underlying size of the population.”
The main driver of change in estimated resident population (ERP) is net overseas migration (NOM). The ABS gets this information from the Department of Immigration and Border Protection. The notes from February also said
“The release of Overseas Arrivals and Departures, Australia (cat. no. 3401.0) for the period October 2014 to April 2015 will be delayed. The delay is due to passenger card processing issues as announced by the Department of Immigration and Border Protection (DIBP).”
One wonders what the hell they were doing that, for 7 months, DIBP couldn’t process passenger cards – a fairly crucial part of immigration one would think.
When I asked how, after updating the estimates in February, there could be such a huge discrepancy by August, I was told the ERP of people aged 15+ in July been revised down by 77,000. Barring a plague or natural disaster, this indicated that the information used in and since February about NOM was way out.
These revisions have a cumulative effect as they go back in time. The trend figure for the number of people employed in September 2013 now has gone down from 11,646,800 to 11,452,500 – a drop of 194,300. This makes a significant difference to claims about job creation.
Speaking of which, I asked the people at the ABS about job creation numbers. They said that the number of jobs created would be less than the increase in the number of people employed because most of the newly employed people were part time and not necessarily in new jobs.
So what I have gleaned from this exercise is that the DIBP need less spin doctors and more people actually doing their job, and that the number of people employed, and hence any inference about job creation figures, can be manipulated in the blink of an eye by “rebenchmarking” – a process which changes history significantly dependent on migration figures that the DIBP either can’t get around to doing or can’t get right.