I wonder if Malcolm Turnbull likes listening to Frankie Valli, because after the release of the latest unemployment figures, which showed the unemployment rate had fallen to 5.9%, he would be well within his rights to start singing a few bars of “Can’t take my eyes off of you”.
But while he and the treasurer, Scott Morrison, may enjoy singing “I love you baby” while gazing at the figures, on closer inspection the more pertinent line of the song might be “you’re just too good to be true”.
In the past year or so, looking at the unemployment figures has required some suspension of disbelief. That isn’t really what you want to think when you see perhaps the most important monthly data released by the Bureau of Statistics.
The figures for October, released on Thursday, showed Australia’s unemployment rate in seasonally adjusted terms had fallen from 6.2% in September to 5.9%:
This was a significant drop – the biggest monthly fall in 13 years, since July 2002.
But just as in August I noted that when the unemployment rate jumped to 6.3% things weren’t as bad as they seemed, this time round things are not as good as they seem.
The problem is the seasonally adjusted figures are extremely volatile of late.
The seasonally adjusted unemployment rate changes on average by 0.12 percentage points each month (up or down).
In the past five years it has changed by more than 0.2% points just eight times – and five of those times are in the past 11 months:
It makes it tough to believe the figures – and even the ABS knows it. In the latest issue it pretty much plead with people to ignore the seasonally adjusted and focus on the trend figures – and with good reason.
In seasonally adjusted terms, the ABS would have you believe 58,600 jobs were created in the month – and nearly half in Victoria alone.
The jump of 26,100 jobs in Victoria had the ABS scrambling to defend the figure, stating “this was a strong but not unprecedented movement, with 12 larger movements in absolute terms over the life of the series, including the most recent in December 2014”.
So sure, it’s not unprecedented, but while saying there have been 12 bigger changes sounds a lot, remember “the life of the series” is 453 months, so we’re still talking pretty exceptional moves.
And Victoria wasn’t the only place where jobs were being found in unusually high numbers. Western Australia is no longer the state suffering from the end of the mining boom, it appears. It had its biggest monthly jump in employment since January 2012:
And yet, would you believe, WA’s unemployment rate, in seasonally adjusted terms, rose from 6.1% to 6.4% – higher than it has been since the mining boom began in 2002.
What the heck is going on?
Well, apparently women in WA are joining the labour force like it is going out of style – 18,000 of them entered in the past month alone, and 15,000 got jobs! So large was the increase in employment for women in WA that, despite accounting for just 5% of Australia’s total employment, it created 25% of the increase in jobs in October.
It meant the participation rate for women in the west rose 1.7 percentage points – the second biggest jump in 13 years and the fifth biggest of all time.
And, while the ABS would have us focus on the trend – and as ever in these cases the trend is your friend – the problem is that the seasonally adjusted figures affect the trend.
The trend is essentially a weighted average of 13 months of the seasonally adjusted figures (it’s a bit more complicated, but that’s the crux of it). And so big was the drop in the seasonally adjusted figures in October that it drastically changed the picture of the employment market provided by the trend figure:
Whereas last month the trend line was suggesting the unemployment rate in September was above 6.2% and rising, now the figures suggest that in September the actual picture was that the rate was just below 6.1% and falling.
It all makes for rather a tough time for policymakers when you are trying to gauge how the economy is really faring.
To an extent, it is not surprising that employment is looking solid. Job vacancy numbers have been growing – even if the growth looks to have peaked a bit.
And the Department of Employment’s leading indicator of employment has been positive for some time now – even if it too looks to have peaked:
But one of the more interesting signs is that the growth of hours worked is actually greater than the growth of employment – this would suggest not only that firms are taking on more people but they are giving those in work more hours. This confirms the view of the past two to three years that there have been a lot of people working fewer hours than they would like:
But, interestingly, the gap between the unemployment rate of those looking for full-time work and the overall unemployment rate has widened:
This is unusual as generally when the employment situation is improving and the number of hours worked is increasing the number of those unable to get work full-time declines. The figures we have here are more indicative of a labour market where full-time work is getting harder to get.
Whatever the case, whether or not you believe these figures, the market certainly believes they are enough to stop the RBA cutting interest rates.
Prior to the release of the figures, the market was pricing in a cut of the cash rate to 1.75% by June next year. Now it rates that only a 70% chance to happen at all next year.
So enjoy the unemployment rate but perhaps don’t get too excited. As with all surveys, it is just an estimate and, while we can trust the figures, as Ronald Reagan liked to say, “trust but verify”.
And in this case, perhaps the prime minister and treasurer might be best served to wait a couple months to see if the good news continues before stepping up to the karaoke microphone to begin singing in celebration.
