The Courier-Mail is correct to highlight the large increase in public service employment that has occurred recently in Queensland, and it is on the right track in suggesting public service jobs are a large share of new jobs created, but it is misinterpreting the data when it claims that:

“THE State Government’s massive public servant hiring spree bankrolled 80 per cent of the additional jobs created in Queensland in the last financial year.”

That is not strictly correct. Because some sectors, such as mining, have had falls in employment, you cannot simply calculate the increase in public service numbers (8,500) as a percentage of new jobs in Queensland (10,400 in 2015-16 in trend terms), as the Courier-Mail has done. The decline in mining employment in 2015-16 was around 6,500 jobs (measured in four-quarter moving average terms, due to the data only being available in original un-adjusted form from the ABS). Now, if we take 8,500 new public service jobs, subtract 6,500 lost mining jobs, we are left with net jobs growth of 2,000, which is 8,400 short of the 10,400 increase overall. So 8,400 new jobs must have come from sectors other than the Queensland public service, such as hospitality, aged care and private health and education providers. In other words, Queensland public service jobs may be around half of new jobs created in 2015-16, but not 80%.

The Courier-Mail would be well advised to get a copy of math guru Jordan Ellenberg’s How Not to be Wrong which has a good section on the challenges of interpreting employment growth data and other data where component parts can be positive or negative. Ellenberg’s advice is (on p. 78 of the Penguin edition):

“Don’t talk about percentages of numbers when the numbers might be negative.”