it won't help you.
For those who care about the truth and the facts, enjoy:
www.zerohedge.com
BY TYLER DURDEN
SATURDAY, FEB 03, 2024 - 10:00 AM
On the surface, it was an blockbuster jobs report, certainly one which nobody expected. Starting at the top, the BLS reported that in January the US unexpectedly added 353K "jobs" - the most since January 2023 (when the print was 482K compared to 131K) , double the consensus forecast of 185K and more than the highest Wall Street estimate (300K from Natixis). In fact, this was a 4-sigma beat to estimate, unheard of in the past year.
The headline data was stellar across the board, starting with the unemployment rate which once again failed to rise - denying expectations from "Sahm's Rule" that a recession may have already started - all the way to average hourly earnings, which unexpectedly spiked from 4.1% (pre-revision) to 4.5%, the highest since last September, and a slap in the face to the Fed's disinflation narrative...
... or it would be if one didn't think of checking how the average rose: well, it turns out that, since average hourly earnings is a fraction, it did not rise due to a jump in actual wages but - since it is earnings over a period of time - "rose" because the BLS decided to sharply slash the number of estimated hours that everyone was working, from 34.3 to just 34.1, which may not sound like a lot until one realizes that the last time the workweek was this low was when the economy was shut down during covid. Excluding the covid lockdowns, one would have to go back to 2010 to find a workweek that was this anemic.
And speaking of revisions, we had a lot of those: in January, the BLS conducted its annual "annual re-benchmarking and update of seasonal adjustment factors." Long story short, what was until December a decline in jobs has now been miraculously transformed into gains, as shown in the chart below.
For those asking, the revisions were unambiguously designed to give the impression that the labor market is slowing much less than it is. Consider this: before the revision, the average monthly job gain in 2021 was largely unchanged (606K pre-revision vs 604K post), and while the average monthly gain in 2022 was revised lower (from 399K to 377K), this was purposefully goalseeked to make 2023 appear stronger, and indeed the average monthly increase in 2023 has been revised from 225K to 255K.
Which would be great, if only it wasn't almost entirely due to the BLS's latest choice of seasonal adjustments, which have gone from merely laughable to full clownshow, as the following comparison between the revised BLS Payrolls number and the ADP payrolls show: the trend is clear: the Biden admin numbers are now clearly rising even as the impartial ADP (which directly logs employment numbers at the company level and is actually far more accurate), shows an accelerating slowdown.
And speaking of seasonal adjustments, the January print was all seasonals, because while the seasonally adjusted payrolls was up 353K, the unadjusted was down 2.635 million, a 3 million jobs delta.
In other words, just a 10% error rate in the seasonal adjustment (roughly where it falls) would wipe out the entire gain and make January increase a decline. Then again, this is the case with every January jobs report, because as shown below, the actual change in jobs in the first month of the year is down anywhere between 2.5 million and 3 million!
But it's more than just the Biden admin hanging its "success" on seasonal adjustments: when one digs deeper inside the jobs report, all sorts of ugly things emerge... such as the latest divergence between the Establishment (payrolls) and much more accurate Household (actual employment) survey. To wit, while in January the BLS claims 353K payrolls were added, the Household survey found that the number of actually employed workers dropped again, this time by 31K (from 161.183K to 161.152K).
This means that while the Payrolls series hits new all time highs every month since December 2020 (when according to the BLS the US had its last month of payrolls losses), the level of Employment has barely budged in the past year. Worse, as shown in the chart below, such a gaping divergence has opened between the two series in the past 4 years, that the number of Employed workers would need to soar by 9 million (!) to catch up to what Payrolls claims is the employment situation.
For those who care about the truth and the facts, enjoy:
Inside The Most Ridiculous Jobs Report In Recent History | ZeroHedge
ZeroHedge - On a long enough timeline, the survival rate for everyone drops to zero
www.zerohedge.com
Inside The Most Ridiculous Jobs Report In Recent History
BY TYLER DURDEN
SATURDAY, FEB 03, 2024 - 10:00 AM
On the surface, it was an blockbuster jobs report, certainly one which nobody expected. Starting at the top, the BLS reported that in January the US unexpectedly added 353K "jobs" - the most since January 2023 (when the print was 482K compared to 131K) , double the consensus forecast of 185K and more than the highest Wall Street estimate (300K from Natixis). In fact, this was a 4-sigma beat to estimate, unheard of in the past year.
The headline data was stellar across the board, starting with the unemployment rate which once again failed to rise - denying expectations from "Sahm's Rule" that a recession may have already started - all the way to average hourly earnings, which unexpectedly spiked from 4.1% (pre-revision) to 4.5%, the highest since last September, and a slap in the face to the Fed's disinflation narrative...
... or it would be if one didn't think of checking how the average rose: well, it turns out that, since average hourly earnings is a fraction, it did not rise due to a jump in actual wages but - since it is earnings over a period of time - "rose" because the BLS decided to sharply slash the number of estimated hours that everyone was working, from 34.3 to just 34.1, which may not sound like a lot until one realizes that the last time the workweek was this low was when the economy was shut down during covid. Excluding the covid lockdowns, one would have to go back to 2010 to find a workweek that was this anemic.
And speaking of revisions, we had a lot of those: in January, the BLS conducted its annual "annual re-benchmarking and update of seasonal adjustment factors." Long story short, what was until December a decline in jobs has now been miraculously transformed into gains, as shown in the chart below.
For those asking, the revisions were unambiguously designed to give the impression that the labor market is slowing much less than it is. Consider this: before the revision, the average monthly job gain in 2021 was largely unchanged (606K pre-revision vs 604K post), and while the average monthly gain in 2022 was revised lower (from 399K to 377K), this was purposefully goalseeked to make 2023 appear stronger, and indeed the average monthly increase in 2023 has been revised from 225K to 255K.
Which would be great, if only it wasn't almost entirely due to the BLS's latest choice of seasonal adjustments, which have gone from merely laughable to full clownshow, as the following comparison between the revised BLS Payrolls number and the ADP payrolls show: the trend is clear: the Biden admin numbers are now clearly rising even as the impartial ADP (which directly logs employment numbers at the company level and is actually far more accurate), shows an accelerating slowdown.
And speaking of seasonal adjustments, the January print was all seasonals, because while the seasonally adjusted payrolls was up 353K, the unadjusted was down 2.635 million, a 3 million jobs delta.
In other words, just a 10% error rate in the seasonal adjustment (roughly where it falls) would wipe out the entire gain and make January increase a decline. Then again, this is the case with every January jobs report, because as shown below, the actual change in jobs in the first month of the year is down anywhere between 2.5 million and 3 million!
But it's more than just the Biden admin hanging its "success" on seasonal adjustments: when one digs deeper inside the jobs report, all sorts of ugly things emerge... such as the latest divergence between the Establishment (payrolls) and much more accurate Household (actual employment) survey. To wit, while in January the BLS claims 353K payrolls were added, the Household survey found that the number of actually employed workers dropped again, this time by 31K (from 161.183K to 161.152K).
This means that while the Payrolls series hits new all time highs every month since December 2020 (when according to the BLS the US had its last month of payrolls losses), the level of Employment has barely budged in the past year. Worse, as shown in the chart below, such a gaping divergence has opened between the two series in the past 4 years, that the number of Employed workers would need to soar by 9 million (!) to catch up to what Payrolls claims is the employment situation.
