Dodgy jobs figures leave us all in the dark

If you can trust the figures, and that's a big if, jobs rose by 24,000 in October, which just happens to be very close to the consensus forecast. The unemployment rate itself was steady at the new upwardly revised rate of 6.2 per cent. The million dollar question is whether that's an improvement or deterioration? It's not that easy to answer.

Taken at face value, the October rise may just represent a bit of uplift following weakness the prior two months (cumulative falls in employment of 32,000). Read that way, employment growth is on a downward trend, falling about 2,000 per month on average, over the last three months. In that time, the unemployment rate has increased from 6.1 per cent to 6.2 per cent. Both of those would fit into the broader slowdown narrative -- the decline in the terms of trade and the end of the mining boom.

However, another way of reading it might be to interpret the jobs jump as a return to trend, following a couple of months of uncertainty. Prior to the period of ‘instability' (as the ABS calls it), employment growth had been robust. Even on the revised jobs numbers, growth was around 18,000 per month -- above average and a decent clip. Why would that change so suddenly?  Well, recall that September business conditions and confidence fell sharply, coinciding with a marked lift in financial market volatility -- the stock market dropped about 9 per cent. Don't forget we also had a hard push against the property market by regulators over that period as well. That this spooked both developers and owner-occupiers is clear, as we've seen in recent building approvals figures.

With all that going on, it may simply be the case that employers shelved any hiring plans -- awaiting better times. That scenario would also probably fit in better with the fact that of the last 10 months, employment has only declined in three of them (two of them recently) and increased in seven and the average monthly increase in jobs over the period was quite solid at around 23,000.

The problem is, we still can't trust the ABS figures. A full review is only being conducted early next year -- and the ABS hasn't ruled out further significant revisions in that process. There are certainly still some oddities in the numbers. On a state basis for instance, the unemployment rate apparently surged to 7 per cent in Queensland from 6.3 per cent, which is the biggest increase in the unemployment rate since 1991, and the highest rate since 2003.

Equally concerning, the apparent lift in the headline unemployment rate of late is not matched by the unadjusted estimates, which show an unemployment rate at 6.1 per cent and perhaps even as low as 6 per cent smoothing for volatility. Prior to the ABS' recent data issues, the unemployment rate was at 5.8 per cent on the unadjusted figures.

To give a further sense as to how ridiculous the situation is, consider that the ‘95 per cent confidence interval' for the October print -- which is the range of values that analysts can be confident represent the ‘true value' --  ranges anywhere from -33,300 to 81,500 on the employment front. For the unemployment rate, that's anywhere from 5.8 per cent to 6.6 per cent.

That is, you can be confident that the unemployment rate either shot up, or fell sharply this month: Things are either going gangbusters, or we are going into a recession -- and you can be confident of that!

Given the ongoing problems with this survey, which are not going to be resolved in the short-term, it's unlikely analysts or policy makers alike, will get much out of it.  There are good reasons to expect much stronger jobs growth than what we've seen of late -- construction is surging and the retail sector has picked up -- both big employers. Yet we won't know until well into next year. In the meantime, the RBA is unlikely to move the cash rate either way