Queensland unemployment recently plummeted – or did it?
Whether you’re a Queenslander getting ready to vote in the January 31 state election, or simply someone interested in employment trends in Australia, Queensland provides a great case study of which jobs figures you should be paying the most attention to.
Why some jobs figures grab the headlines
Politicians love making promises about jobs, even though they’re hard to deliver. But they also know they risk a backlash when unemployment rises. So they obsess about whether monthly job figures from the Australian Bureau of Statistics (ABS) show them in a good or poor light. Whatever the latest statistics, governments and oppositions are sure to try to apply the best or worst possible spin.
Sifting through that spin isn’t easy. And journalists often don’t help by focusing on the volatile and unpredictable “seasonally adjusted” estimates, at the expense of the more boring “trend”. More clicks, and more attention, flow from reporting the bigger swings of the seasonally adjusted numbers.
That’s OK if both figures show the same thing. But often they don’t – and Queensland has just seen a stark example of that during the state election campaign.
Seasons and trends
Employment and unemployment follow seasonal patterns. Fruit pickers get work in fruit season. School leavers look for work every summer. Each month follows a pattern of regular rises or falls in jobs or joblessness.
The ABS figures come from a monthly household survey. Like all surveys, they are subject to random sampling error: even without any underlying change, a different survey will give you (sometimes very) different results.
“Seasonally adjusted” figures from ABS overcome the first of those problems, but not the second. They stop you getting surprised by job-seekers in December. But you still can – and journalists often do – get surprised by the monthly volatility in job figures.
The trend estimates overcome that problem, by smoothing out random fluctuations in “seasonally adjusted” figures. They have some limitations – for example, they can be slow in showing “turning points”.
But for assessing the performance of an economy over time, they’re definitely what you’d rely upon. Back in 2013, the Queensland Treasurer rightly referred to trend estimates as “more reliable”.
As statisticians often say, “the trend is your friend”. And the smaller the economy, the larger the relative variation in seasonally adjusted estimates. So it’s even more important to focus on the trend for state than national economies.
Unemployment in Queensland
What do the numbers show? In December 2014, the seasonally adjusted unemployment rate in Queensland was 6.1% – fully 0.7 points down from 6.8% a month earlier. Yet two months before that it was only 6.3%. And 6.6% a month before that.
Does it really make sense that the underlying unemployment situation would bounce around so quickly? Not really.
Now look at the trend: 6.6% in December 2014, 6.6% a month earlier, and 6.6% two months before that – a much clearer indication of where the state labour market is at.
That trend unemployment figure for Queensland is up from 5.5% in March 2012. It is above the national average and the rates for many other states, and just 0.2 percentage points below the highest (Tasmania, on 6.8%). On the volatile seasonally adjusted figures, Queensland was much better than South Australia in December. But on trend, it was worse.
Employment figures to watch
However, there is much more to a labour market than the unemployment rate. That’s because it can be distorted by changes in beliefs about the benefits of job search. If people think there’s no hope of finding a job, they stop “actively” looking for work (the ABS has fairly strict criteria for what that means), and the unemployment rate goes down.
This doesn’t mean the ABS fudges or deliberately understates unemployment. It follows standard international practice, rigorously and competently. This is just one of those peculiarities of labour market measurement.
So increasingly, economists and bodies like the OECD are also focusing on the “employment rate”. That means the number of employed people as a proportion of the working-age population. The higher it is, the better. It is unaffected by changes in people’s beliefs about job-seeking.
It’s more useful than just looking at the number of jobs, because the population in most places increases anyway, so you need some job growth just to stop a rise in joblessness.
The trend employment rate in Queensland in December 2014 was 60.9%. This was slightly above the national average, as it has been since 1991.
But, notably, it is down from 63.1% in March 2012. The employment rate in Queensland is the lowest in more than 10 years. The fall in Queensland was well above the national fall and the equal worst of any state or territory over that period (alongside South Australia).
One reason is the drop in full-time employment. In trend terms, this was 8,100 (about 0.5%) lower in Queensland in December 2014 than in March 2012. (Part-time employment grew by 8%.)
Queensland was the only state to experience a fall in full-time employment over that period. Its full-time employment rate (full-time jobs divided by the working-age population) fell by more than any other state.
Another, less commonly used, indicator of labour market changes is total hours worked. In December 2014 this was 1.5% lower in Queensland than it had been in March 2012. All other states, apart from South Australia, showed an increase over this time. Queensland was the only state where hours worked dropped over the year to December 2014.
While the end of the mining boom is sometimes blamed for the current state of the Queensland labour market, it is worth noting that mining is one of the six industries in which jobs grew over the year to November 2014. (Employment fell in 11 industries. This is based on original data as the ABS provides no trend industry estimates.) The deterioration in the labour market is broadly based.
Looking ahead, labour markets have supply and demand sides. On the supply side, the Queensland population grew by 1.9% last year, and that is likely to continue. On the demand side, job vacancies were 6% lower in November 2014 than a year earlier in Queensland – the only state with a fall over this period. This combination predicts further falls in the employment rate, which would be matched by rising unemployment unless more people give up actively looking for work.
All this is very different to Campbell Newman’s promise in January 2012 of 420,000 new jobs over six years and to get Queensland’s unemployment down to 4%. Employment has instead grown by only 45,000, well below the rate of population growth or the minimum needed to stop unemployment from rising.
But I’ve never understood why people believe job growth promises from state politicians anyway. The truth is, federal factors are the biggest single influence on a state economy, though state governments certainly make a difference.
Regardless, jobs targets remain a state election staple from both major parties. Former Labor premier Anna Bligh promised jobs for “100,000 breadwinners in 100,000 Queensland homes”. It turned out that target included part-time jobs, prompting Bligh’s Liberal National Party opponents to criticise her spin. Ironically, all the net job growth since the LNP came to office has been in part-time work.
Whether it’s 100,000 or 420,000, one thing is certain: while setting a target is easy, if it isn’t met politicians simply point to circumstances “beyond our control”. But that doesn’t mean we shouldn’t hold all sides of politics accountable for what they promise.
Read more of The Conversation’s Queensland election 2015 coverage.