September 24, 2020 by Dymphna

Weirdest jobs data ever.

Don’t take the unemployment data on face value.

So there was a lot of confusion around the most recent jobs data. Was it good? Was it bad? Is it all just a vast conspiracy to hide the true extent of the nation’s economic pain?

This really was a funny one. At first it was good. Then it was bad. Then it was good, and then it kinda ends up being about what you’d expect – which, given we’re talking about one of the worst economic shocks in living memory – is kinda bad.

So first up, the unemployment rate, which is your headline rate – which is you’re lets talk about it on the evening news number, – was good. It was unbelievably good.

Despite the biggest contraction in GDP ever, the unemployment rate fell, from 7.5% to 6.8% (which, for the record, is the largest monthly fall ever recorded in the labour force series.)

Economists were shocked. Literally no one saw it coming.

Normally markets would get a solid bump out data that surprised to the upside, but the surprise ended up being so big that I don’t think anyone believed it.

We then started looking for explanations.

First up, many people noted that the increase in employment was not driven by employees, but by sole-traders and contractors.

The argument that started going around was the people had left full-time work to become Uber drivers and so on. Or they were gaming the JobKeeper rules somehow.

Either way, the conclusion was that the surge in jobs was actually in “bullshit” jobs, and it was not cause for celebration at all.

This story was doing the rounds for a couple of days, until some people, like Greg Jehrico, noted that August is almost always a bad month for employees, and a good month for contractors.

You can see it in the data, if you compare every August over the years. Employee numbers tend to always fall:

While the number of contractors and sole-traders tends to increase:

And in that context, the employee numbers were actually great, and the non-employee numbers were ok.

So we’re back to things being surprisingly good.

But that’s still surprising.

So we then thought that it might be about the ‘seasonal adjustment process’ that the ABS does with the numbers. So we know things are always quiet in January, so the ABS adjusts January numbers to try and look through that seasonal regularity.

And that makes sense in normal times, but these times are anything but normal. And it is true that while the seasonally adjusted employment numbers grew strongly, the underlying real numbers were much less impressive.

So that makes you stop and wonder if things are actually that great.

And when you tune in with some of the other data, it seems to confirm a bleaker picture:

For example, if you look at monthly hours worked, we’re still looking at the worst numbers since WWII. And that’s true even if you exclude Victoria, which is having the worst time of things.

That’s grim.

At the same time, if you look at the full-time employment rate of prime-aged working men (which always gets smashed during a recession) we see that there has been almost no improvement from the brutal lows of June:

Greg Jehrico also gives us a labour under-utilisation rate, which gives us a much better feel for the actual lived-reality of the labour market. And on that measure, it’s still clear things are as bad as they’ve ever been:

So there’s a lot of confusion going on here. That’s to be expected. It’s 2020 after all.

But the take-away message is that we’re not out of the hole just yet. It’s not time to be pulling the rug of government support packages out from under our feet.

The labour market still a ways to go yet.