Westpac with the note:
The ABS Weekly Payrolls estimate of jobs lifted 0.3% in the two weeks to 22nd May and the fortnightly increase of the previous release (for the two weeks ending May 8) was revised from -0.3% to -0.1%. It is normal for Payrolls to be revised upwards in the weeks following the initial release as some firms, particularly SMEs, are slow to submit their returns to the ATO. Total wages paid by employers rose 0.6% in the fortnight.
Now that we are clear of the Easter/School Holidays distortions the data should have greater clarity around the impact of the ending of JobKeeper and the recent moderation in employment growth, particularly given that Payrolls appear to be lagging behind leading labour market indicators such as the various business surveys, job vacancies and job ads.
With revisions, Payrolls point to a modest rise in May employment
Payrolls data have been a consistent guide to the direction, if not always the magnitude, of changes in Labour Force employment (see figure 2) through 2020. Following the May 8th release, Payrolls was pointing to a meaningful fall in employment, possibly as large as -70k. However, Weekly Payrolls are subject to meaningful revisions due to late submissions so we decided to wait for the 22nd of May release before finalising our May Labour Force forecast. With those revisions Payrolls for the first two weeks of May (the reference weeks for the May Labour Force Survey) are now up 0.3% on the first two weeks in April. Our forecast for employment in May is +30k/0.2% which in original terms (Payrolls are not seasonally adjusted) is a +74.0k or 0.6% rise; May has tended to have a seasonal rise of around 0.3%.
Weekly Payrolls has a short history but since January 2020 the percentage change in employment in the Labour Force Survey has tended to be smaller than the percentage change in Payrolls. The only exception to this was in October 2020 when employment outpaced Payrolls and in May 2020 (post the covid lockdown crash in April) when Payrolls moved in the opposite direction to employment. We are looking for employment in May to be boosted by business and workers returning from the April school holidays plus making some allowance for some upward revision to current Payrolls due to late submissions. Nevertheless, based purely on where Payrolls are right now, it appears that the risk to our +30k forecast for employment is to the downside.
Collapse in SME Payrolls since March 27 points to a JobKeeper effect
Following the release of the May 8th Payrolls we noted that since the week ended March 27th (the last week before JobKeeper ended) Payrolls were down 1.5% while wages were down 3.5%. Given that at SMEs are likely to be more dependent on the cash flow support of JobKeeper than larger firms we noted that SME Payrolls had fallen 6.5% since March 27th, medium firms’ Payrolls were down 1.6% while large firms’ were up 1.2%. However, as noted earlier initial Payrolls can be subject to meaningful revisions, particularly for SMEs and especially around school holidays, so we decided to wait for the May 22nd Payrolls so we would get an update on May 8th data as well as being clear of any holiday distortions before we could be definitive on any JobKeeper effect.
With the May 22nd data, even with revisions, Payrolls suggest the ending of JobKeeper support did hit SMEs particularly hard. While the level of SMEs’ Payrolls at May 8th was revised up 2.8% (medium firms had a +1.0% revision while large firms were revised down 0.5%) the level at May 22nd was down 5.8% from March 27th, medium firms were down 0.5% while large firms’ Payrolls are up 1.9%. It is extreme unlikely that any future revisions will make up for an almost 6% fall in SME Payrolls since the ending of JobKeeper and so we see this as indirect evidence that the loss of the JobKeeper cashflow support had a meaningful impact on employment. Robust gains in larger firm payrolls since the ending of JobKeeper suggest that these firms were not as dependent on the cash flow support and since the ending of the programme their gains have mostly offset SME losses with total Payrolls down just 0.6% since March 27th.
Industry-level data also suggests a JobKeeper effect
In the 22nd May release of Weekly Payrolls, we continued to see a divergence in outcomes between industries in the post-JobKeeper period. Moreover, this divergence remains consistent with the distribution of reliance on JobKeeper across industries.
One way to infer an industry’s reliance on JobKeeper is to look at subsidy payments as a share of the total compensation of employees (COE). Using this metric, the March 2021 national accounts indicate that the most reliant industries were Arts & Recreation (JobKeeper payments equivalent to 13.2% of COE), Accommodation & Food services (8.1%), Other Services (7.8%), Agriculture (5.2%), Construction (5.0%), and Transport (4.7%).
It is also the case that these six industries have seen the strongest deterioration in payrolls since the 27th of March (the conclusion of the subsidy). Compared with the week before JobKeeper finished, payrolls in the Construction industry are down 1.8%, Transport down 2.2%, Other Services (-3.2%), Accommodation & Food (-3.4%), Arts & Recreation (-3.8%), and Agriculture down 7.2%.
Figure 5 (click read more below) shows the reliance on JobKeeper (as measured by JobKeeper payments as a share of COE), and Figure 6 shows the change in payrolls since the subsidy’s conclusion. These two indicators are plotted together in Figure 7, revealing an observable correlation. The close mapping between reliance on JobKeeper and payroll outcomes is another indicator that the conclusion of the subsidy is having an impact on employment growth.
The larger states hit harder by the ending of JobKeeper
There has been a significant fall in SME Payrolls in NSW and Victoria since March 27th, -7.1% and -6.8% respectively. Outside of ACT, where Payrolls are down 7.1% the smaller states have done better with a 4.2% decline in Qld, a 3.7% decline in SA and a 4.7% decline in WA.
The recent weakness in SME Payrolls in NSW and Vic means jobs with small business in those state have dropped below where they were pre covid, -1.9% in NSW and -1.2% in Victoria. By contrast SME jobs in WA are 5.0% higher, 3.6% higher in SA and 1.6% higher in Qld (figure 8).
Conclusion
The reference period for the upcoming May Labour Force Survey (LFS) is the fortnight ended May 15. This latest release of ABS Weekly Payrolls covers that fortnight along with two weeks’ worth of revisions for the week ended May 8 and an extra week for late submissions for the week ended May 15. As such after allowing for some further upward revisions due to late submissions from SMEs, and a rebound from the drag of the school holidays, we are forecasting a +30k rise in May employment.
We also note the significant fall in SME jobs, plus the greater loss of jobs in the sectors more dependent on JobKeeper, since the ending of the labour subsidy programme suggests some firms are facing meaningful headwinds with the loss of a significant cash flow support. It has been the robust jobs growth in larger firms, and in the sectors less dependent on JobKeeper, that has provided the offset to jobs losses from SMEs.
