In its September labour force report, which reported the Australian unemployment rate fell to 5.6% from 5.7% while employment fell by 9,800 (and full-time employment fell 53,000!), contrary to the jobs growth the market was expecting, the ABS more or less admitted it had to fudge the data so the numbers did not appear even more odd. The numbers might have shown an even greater reduction in employment, for example. Note that the unemployment rate fell even though employment fell due to a decline in the participation rate. On this point, see Pete Faulkner’s post:
Unemployment rate drops on fall in participation
Queensland actually turned out to be the reason the ABS had to fudge the data. The ABS notes on its website:
“The incoming rotation group in Queensland for September 2016 was considerably different to the rest of the Queensland sample and its influence has been temporarily reduced as part of the estimation process. The data will be further reviewed when October data are available.”
Annette Beacher from TD Securities went as far to say that the latest ABS data are “rubbish” (see this ABC News report), and I pretty much agree with her, as I mentioned to ABC Radio’s Loretta Ryan on the Regional Drive program yesterday afternoon. The information content of yesterday’s labour force data is very low. I am always careful not to read too much into any one month’s figures due to the volatility in month-to-month changes, but yesterday’s data were even more dubious than data in the past. Considering both the unreliability of the labour force data, which are essential to monetary policy formulation by the RBA, and the recent Census debacle, the Government may need to bring in new senior management and boost the budget of a clearly struggling ABS.
Finally, Queensland’s unemployment rate fell from 6.2% to 6.0% in seasonally adjusted terms (see chart below), which I would like to believe but am unsure of due to the unreliability of the data. That said, I told Loretta Ryan there are reasons to be confident Queensland’s economic conditions are improving, especially considering positive developments in mining, tourism, and agriculture.
I wonder how much effort ABS has shown in getting real time data from businesses. One could do it here and of course it wouldn’t displace the survey based methods, but there are so many opportunities for a new kind of approach and the ABS seems to have shown very little interest. You can say that they’re not funded to do this, but if you think about it that’s a lame excuse. The fact is that a whole new set of possibilities have become available and whatever it’s budget ABS should be working out the optimal mix.
Yes, very good point, thanks Nick. They could cut back to a quarterly household survey and estimate monthly figures based on online searches or job ads for example.
Great minds think alike, but yours is more programmatic and closer to the nuts and bolts than my suggestion.
Which goes to demonstrate the benefits of a cognitive division of labour.