We know from the real time US Federal tax withholding data that the December jobs gain was NOT just 199k jobs. Withholding had a 3.2% higher annual growth rate in December than in November. That would equate with not 199,000 new jobs. More like 1.99 million.
Now before you get too excited, not all of that increase in withholding was an increase in the number of jobs. The average weekly earnings rose by 0.8% M/M so that the real rate of change was 2.4%.
We also know that withholding taxes include distributions other than regular income, particularly employee 401K and IRA distributions, which are subject to withholding. We can assume that these were higher this year than last year, but we don’t know how much. It’s unlikely that this would account for all of the difference between November and December.
The 2.4% month to month real increase in withholding is the biggest since the initial post pandemic y/y surge in May. That surge settled down in June. December’s is the strongest monthly performance since then. It is a full 1 % better than the previous peak in the year to year growth rate, in October.
The BLS said that October jobs increased by 648,000. We have a 1% stronger annual increase in withholding taxes in December and yet the BLS reports only 199,000 gain in jobs? I mean WTF are they doing? Where are the other 447,000 jobs? Plus 1%. How did they miss that? Forget the tax data.
Their own data says they missed a huge chunk of the jobs that were added last month.
Let’s use the BLS’s own data. Let’s look at December, not seasonally manipulated data, in other words, the actual number that they derived from their survey’s before seasonal adjustment, for the previous 10 years. This is the month to month change that they themselves derived from their own surveys.
What do you notice about this data?
That’s right. Every single December there were fewer jobs versus November.
Except for one. This year.
December 2021 has the only gain in jobs in the last 11 years. It was the best performer, by far, of all of the last 11 Decembers. Yet the BLS managed to see only a tepid gain of 199k in its headline number.
Now let’s see how the BLS ranks this December with the past 10 Decembers on the basis of their seasonally adjusted headline number. The M/M column shows the change from November each year. Rank is how that December ranked among the 11 years. December 2021 ranked 7th, that is, fourth worst of the last 11 years.
How is is that they only managed to see an increase of 199k, ranking the 4th worst in the last 11 years, when last month was actually the best of the 11 years.
How is that reasonably possible. It’s ridiculous. It’s absurd. It’s criminal. And yet Wall Street takes this garbage seriously.
Here’s another way to look at it. Based on the BLS headline number, the December 2021 M/M change increased over December 2020’s performance by 505K. That makes sense because December 2020 was terrible.
But the actual data, NSA, says that it really increased by 591K. The BLS’s own data shows that it missed at least 85,000 jobs. It also shows that the NSA data for December was +74,000 over December 2015 which BLS says was the best December of the previous decade. If last month was 74K better, then the headline number should have been 273k+74k= 347k.
So even the BLS own data shows that they undercounted by anywhere from 85k to 306k.
The only miss here was by the BLS. It missed the creation of a couple hundred thousand jobs. Based on the withholding tax data, they actually missed far more than that.
Bond traders are not fooled by the BLS’s statistical garbage.
The 10 year yield has broken out and is headed for the projections I reported yesterday.
