God knows what the ABS will throw up today but Westpac takes a wild swing:
The ABS has come clean that there was a problem with the Aug survey related to a change in the underlying seasonality in the survey. We suspect that the source of the problem was a change to the Aug survey that previously had encouraged part-time workers to report as “not working” to avoid the extensive additional questioning. To fix the problem the ABS has set the seasonal factor to 1.0 (turned off seasonal adjustment) for July, Aug and Sep. This means there are significant revisions to the Aug labour force estimates.
Given no other changes by the ABS (something we can’t guarantee) Westpac has estimated the following revisions to total employment.
Jul will be revised from –4.1k to +3.5k
Aug will be revised from +121.0k to +32.1k. (Westpac had forecast a rise of 15k for Aug.)
The revisions also take the annual pace of employment growth down from an unbelievable 2.2%yr to a more respectable 1.5%yr which is more in line with Westpac’s Jobs Index.
From this revised starting point we clearly can’t hold to our earlier forecast for a minus 70k print on total employment in the Sep Labour Force survey. That forecast was based on our observation that underlying seasonality had clearly changed and that the seasonal factors had not been changed to match. As such, this would have induced a significant correction in Sep.
Our Jobs Index suggests that an annual pace of employment growth of around 1½%yr is about right. But we also note that the recent improvement in the employment indicators from the business surveys, and in particular the various AiGroup surveys, suggest that employment growth could accelerate to around 2%yr by year end or early 2015.
For that reason we are forecasting +25k for total employment in Sep which will lift the annual pace to 1.6%yr. We should also note that this will produce a total rise over the two months to Sep of +57.1k. In the two months to Sep 2013 total employment in original terms (remembering the ABS has turned off the seasonal factors for Jul, Aug and Sep this year) was 60.9k.
Assuming a small rise in participation to 64.8% (as it was in Jun) and a trend gain in working age population then the unemployment rate drops to 5.9%.
