There are three possibilities about the latest labour market statistics: (a) either today’s 6.4 per cent headline unemployment rate is wrong; or (b) the past several months’ labour market statistics have been wrong; or (c) they’re all wrong. What’s certain is that they’re not all correct. The odds, plus most of the recent other data related to employment, say the 6.4 per cent number just doesn’t add up. Lock in (a), it’s wrong, thanks Eddie.
Of course, that hasn’t stopped forex and bond traders getting a little excited – any excuse for a movement will do when you make your money out of movements. But before anyone else gets too excited, it’s worth trying a little perspective.
Source: ABS First of all, there’s a reason why the Bureau of Statistics prints the trend series first and the seasonally-adjusted numbers second – the trend smooths out some of the sample errors that inevitably work their way into the numbers from time to time. The trend series is much less dramatic, which is why traders and headline writers much prefer the seasonally adjusted numbers that jump around and give us headlines. As the accompanying graph shows, the trend series unemployment rate has worked its way steadily higher to 6.1 per cent, while the seasonally-adjusted rate has been enjoying a more exciting existence. That 6.1 per cent figure is more in keeping with what most private sector economists have been expecting. The official Treasury forecast is for unemployment to peak at 6.25 per cent. Secondly, the ABS itself is wary of putting too much weight on any one month’s employment figures. It sticks a warning about sampling errors just a paragraph or so below the seasonally adjusted figures. Yes, the headline says the unemployment rate increased by 0.3 points, but the ABS says it’s 95 per cent confident unemployment actually did something in between falling 0.1 and rising 0.7 – that 0.3 is just the midpoint. For thirds, as reported, part of the reason for that wide range of possibilities is that last month had a sample rotation in the data collection and the new sample just had a larger proportion of unemployed and a smaller proportion of employed than the sample it replaced. That happens. The bureau says the sample could account for a third of the change in the unemployment rate. And fourthly, such a sudden increase, with the employment numbers flat, is at odds with the trend we’ve been seeing in the ANZ jobs ads survey and in the National Australia Bank business confidence index. As reported here last month, there’s a very nice correlation between employment growth and the NAB employment intentions survey. Today’s figures show employment growth effectively flat when the other data give an expectation that we should be seeing some growth. (For that matter, a real world reading of the nominal dip of 300 jobs last month is a good deal better than it looks, there being an increase of 14,500 full-time jobs and a decrease of 14,800 part-time jobs. That many extra full-time jobs gained certainly beats the bigger number of part-time jobs lost in my book.) But rogue survey or not, today our job market is soft. As the RBA quietly reminds us each month, that’s not going to change any time soon and it’s what the central bank expects will keep inflation down for the next two years.
