Once again, the BLS just can’t seem to be straight-shooting when it comes to employment data. There are so many quirks in this report that it’s no wonder equities have rebounded in to the green, generally up around +1%, after being down near -1% in pre-market post Payrolls.
I could spend the afternoon digging into and explaining the wackiness of this report, but here’s a bottom line fact… The headline print, at a 6 sigma beat of consensus and higher than any sell-side estimate, sounds strong. Dig in a little, and not so much. How does the BLS explain the sudden upward break in the Jobs trend and the sudden strongest print since January? Well they don’t. If this year has proved anything, it’s that the BLS is politically motivated, not motivated to bring Americans an accurate depiction of the jobs market.
There’s the Establishment survey, which added 336,000 jobs, then there’s the household survey that added just 86,000 jobs (the second lowest of 2023). I’ve talked about the two diverging series throughout the year – establishment survey (with one month’s reporting so low that 75% of it was modelled on an excel spread sheet) much stronger than the household survey. So this huge disparity between the establishment survey’s best print since January and the Household survey’s second worst print of the year (336k vs. 86k) is not surprising.
Despite the best headline print since January, the number of unemployed workers rose to 6.360 million, the highest number since January 2022, the unemployment rate was expected to decline to 3.7%, but stayed at 3.8%.
Vanda Research FX trader, Viraj Patel noted earlier: the official adjusted data showed this Leisure and Hospitality added a whopping +96k jobs. But unadjusted data showed that the sector lost -466k jobs in Sep. This means that the unadjusted private sector payrolls was -399!
So how are so many jobs added with a 399k loss in jobs? Government jobs added 984k jobs (ongoing fiscal stimulus in the form of government jobs). The Payrolls report’s strength is largely a matter of seasonal adjustments.
In looking at the Household survey (considered to be far more accurate), the adjusted data shows that part-time workers accounted for the entire increase in the Household survey, rising by 151K, while full time workers lost -22K jobs in September.
Every month of the last 3 months has seen a trend of full time workers decreasing while part time workers increase.
Taking the BLS’s magical seasonally adjusted numbers out of the mix, we find the establishment survey actually shows an 855,000 loss of full time jobs. That’s the largest collapse in jobs since April 2020 (the COVID lockdown with the majority of the economy shutdown).
So how do we get to job gains with so many full time job losses? Part-time workers exploded by 1.127 million jobs.
The household survey counts multiple job holders, but the establishment survey sees 1 person holding 3 new jobs as 3 new jobs added, although it’s 1 person. On a seasonally adjusted basis, multiple job holders increased by 123K to 8.151 million, the highest since January 2020. And the more accurate (less magic) unadjusted number jumped from 7.778 million to 8.146 million, an increase of 368K, or more than all the 336K payrolls reported by the establishment survey.
In short, the massive job gains of the headline survey establishment print, was due to people working part-time, or adding a second or third job, which is double or triple countered by the establishment survey as new jobs added.
The number of multiple job holders (E.g- having a full time job and adding a secondary full-time job) hit a record high with the September Jobs report!