From Westpac:
We can debate magnitudes but overall all indicators suggest Australia currently has a robust labour market outside WA
In November, total employment lifted 71.4k compared to a market median of –10k and a top of the range estimate of +20k. Westpac’s forecast was for –20k.
Total employment has now grown 342k (or 3.0%yr) in the year to November, a significant acceleration from the continuing the acceleration from 2.0%yr in September to 2.7%yr in October. Revisions were minor this month.
The mix was also very robust with a 41.6k rise in full-time employment (206.3k or 2.6%yr) and an 29.7k rise in part-time employment (138k or 3.9%yr). There was also a solid rise in participation (from 65.0% to 65.3%) and the unemployment rate fell to 5.8% from 5.9% (5.85% from 5.90%).
This is a positive update on the labour market and one the RBA would welcomed as it looks forward to its summer recess. There are also a number of reasons to think that grounds for a robust labour market are quite clearly defined. Firstly, other labour market indicators while not as robust as the labour force survey are still definitively positive. Secondly, the improvement in the labour market has been the greatest in NSW where the unemployment rate has dipped to 5.2%. This is consistent with a rebalancing of growth away from production (and in particular resources) to services. And thirdly, it is consistent with the recent positive update on economic growth.
So while we always try to take data release at face value, and agree with an overall assessment that the labour market is robust and, overall, is on an improving trend driven by better outcome in NSW there are some issues that make us question the magnitude of strength of the survey.
Firstly, the annual pace of employment growth of 3.0%yr look way to strong compared to our indicators of the labour market (such as the business surveys, our Jobs Index and unemployment expectations). These indicator are pointing to employment growth more around 2%yr. While such divergences are not unusual in history, normally the pace of employment growth has fallen back to the Jobs Index rather than the other way around.
Secondly, there has been a surge in female participation, particularly in NSW, which is driving the recent lift in overall participation. This started in May 2015 and there are some suspicions that this may be more about sample rotation than any underlying change (it is unusual from one gender from one state to drive such outcomes). Given that this group has now been in the sample for eight months, they will be rolled out in December survey. If the theory that the May 2015 sample did pick up group of females in NSW that were more attached the labour force than normal, thus helping to boost both participation and employment, we should find out in the December survey as they roll out.
Thirdly, it still looks as if the ABS is having a hard time seasonally adjusting the data post the changes they may to the supplementary surveys in 2014. It appears that the bulk of the employment gain was in the matched sample, that is it was not due to sample rotation. But note the ABS only releases the gross flows data in original terms which Westpac has to seasonally adjust. The seasonally adjusted picture is the complete opposite based on the unadjusted series suggesting the seasonal adjustment process (and any ABS adjustments to account for changes in the survey) is again critical to interpreting this month’s data.
Overall it is a robust update even if there are some question marks around the magnitude of the strength shown in the survey. In addition, the gross flows suggests that the quality of the employment gains are not great as they are coming mostly from a reduction in voluntary separations. More worrying is that it appears we are seeing a sharp drop in new hirings which for now is being more than offset by fewer separations.
We are eagerly waiting for an industry breakdown (due on December 17th) to see if it can provide any further insight into the labour market.
That’s a sensible take and that chart says it all.
