With less than a week to go in the election campaign the latest employment data is able to reveal which areas of the nation have fared the best and worst since the last election – and reveals that there would be electorates in the mood to punish, while others would be thinking things are going quite swimmingly.
The range of unemployment rate around the country ranges from the relatively unemployment free 2.6% of the Sydney Northern Beaches, and the 2.9% of Sydney East, up to 11.7% in the Queensland Outback, 9.2% in Townsville, 9.1% in Adelaide’s north and 8.5% in Cairns.
Such high unemployment in the northern Queensland regions are a large reason behind the visits by the leaders to seats such as Leichardt, Dawson and Flynn despite them being relatively safe Liberal Party seats. A look at the path of the unemployment rate in Townsville also highlights the problems with looking at even state-based unemployment rates.
For most of this century, Townsville’s unemployment rate largely followed the broader Queensland rate but since the end of 2013, Townsville’s unemployment rate has soared:
Townsville also has seen one of the worst rises in the unemployment rate since the 2013 election. Other areas to have done poorly since the last elections are Murray in NSW, the Hunter Valley (excluding Newcastle). The best areas were in Tasmania, with the south-east of the state seeing the unemployment rate fall from 9.7% in September 2013 to 6%:
Unemployment change since 2013 election
Of course, as I have noted in the past, the unemployment rate can hide a lot of what is going on, but here they are a pretty good reflection. The Townsville area has also seen the biggest fall in the ratio of employment to population since the last election.
In September 2013, 65.5% of the adult population in Townsville were employed – now it is just 55.2%; similarly the Cairns area has seen a fall from 62.0% to 55.1%.
Other areas that are now experiencing lower levels of employment are the Murray area in NSW, the Logan-Beaudesert area south of Brisbane and Mackay, also in Queensland.
As a general rule, the areas with the highest levels of employment now remain the highest levels of employment in future years. Interestingly, the areas of Sydney City and Inner South saw one of the biggest falls in the level of the population employed. The overall level of 66.8% of the adult population employed remains very high – the 14th highest in the nation – but it is a fall from September 2013 when the 72.4% employed made it the second best in the nation (behind Darwin which remains number one).
Employment to population ratio change since 2013
Looking at changes in the level of the population employed is also very good for showing how a state can experience sharp division of economic performance. While Queensland areas have seen of the worst declines they also have some of the best improvements. Gold Coast’s employment to population ratio went from 61.1% to 65.8%, and the inner Brisbane and Brisbane east areas were also among the top 10 best increases across the nation.
Such diversity is also seen in the level of actual employment growth. Of the 87 local areas, only 21 saw employment actually fall in nominal terms, and five out of the worst seven were in Queensland – the outback, Townsville, Cairns, Mackay and Logan- Beaudesert. But at the other end of the scale, the Queensland areas of Moreton Bay-North and the Gold Coast were among the highest employment growth areas:
Employment growth since the 2013 election
On a statewide basis, the ACT, South Australia and Queensland are the worst performing states and territory since the last election, but WA with the end of the mining boom is far and away the worst performer for full-time employment:
The end of the mining boom in an employment sense (if not in exports) is reflected as well in the growth of jobs across industries. Since the last election mining employment has fallen by nearly 42,000 – or 15.6%. The biggest growth industry in percentage terms is the real estate services industry, which rather nicely fits with the narrative of the housing boom.
But in terms of actual jobs, the health care and social assistance industry has easily been the biggest growth industry. There are now 169,000 more employed in that industry than there was at the last election. For context the professional, scientific and technical services industry is the second best growth industry with less than half that amount – 80,200:
And as for actual occupations, the continued decline of manufacturing as well as the mining industry drop-off has meant, not surprisingly, it has not been a great three years to be a machinery operator or driver. Such occupations have seen a 4.7% fall in total employment and a 6.3% fall in full-time work:
Given the increase in work in the health care and social assistance industry, equally unsurprisingly, the biggest growth occupation has been among community and personal service workers.
This is part of a long-term trend which now sees such occupations number more than labourers – something unthinkable 20 years ago:
Employment always weighs heavily among the issues in an election. Employment pretty much equals the economy for most people. The poor performance since the last election of employment in parts of Queensland, South Australia and Western Australia would suggest an opportunity for the ALP.
But as ever it comes down to who gets credit and blame. The Liberal Party loves to take all credit for the mining boom, and yet dismisses the collapse in work in that industry as just the expected “transition” that no one could halt. Voters in WA in 2007, 2010 and 2013 were very happy to vote for the Liberal Party in the good times, will they punish the incumbents now that things are not so rosy?
Similarly, in South Australia the long-term incumbent ALP state government means it has been tough for the ALP to place all blame on the federal government, and perhaps is a major reason why the Nick Xenophon Team has seen a surge in support in that state.
But if employment does have an impact on votes, the great diversity of employment performance within states, let alone across the nation, suggests that it would not be surprising if the swing in votes is equally diverse – and makes it very tough to use a uniform two-party preferred vote to predict the result.
