Excerpted from https://liquiditytrader.com/index.php/2023/10/04/tepid-tax-collections-mean-its-the-supply/
The BLS has overstated jobs gains for most of this year. The jobs reports have not been consistent with the withholding tax collections, which are raw, unmanipulated data. The withholding tax revenue pattern suggests job gains have been near zero since April. Yet the BLS has consistently reported strong gains.
The BLS nonfarm payrolls survey of employers is dated as of the 12th of the month. While it is supposed to represent conditions in the most recent month, when employers report to the BLS as of the 12th, HR managers are showing conditions mostly from the prior end of month payrolls.
At the end of August, 4-week average withholding was virtually unchanged year to year. At the end of July, it was a gain of +1.65%. That’s a month to month DROP of 1.6% in August, which is the data that HR managers were using when they reported to the BLS as of September 12. Friday’s jobs report should show a crushing decline in employment.
But it’s a crapshoot whether this will show up now or not. The BLS statistical massagery often takes months to adjust to reality. They may not catch up with that weakness for months, perhaps revealing it only in the annual benchmarking, which doesn’t make the nightly news.
9/3/23 At the end of June, the year-to-year change was +1%. The month to month change from June to July was therefore +0.5%. With total payrolls reportedly 156 million, that’s about +78,000. They reported a gain of +187,000. That’s an overstatement of 109,000. Moreover, we already know that the next report should be much, much worse. Consequently, there will be big downward revisions ahead, with the possibility of a one month catch up move that shows up in the monthly report, not as a revision, but as a surprise drop in payrolls for the current reporting month.
Catch-up reports and monthly revisions aren’t the only data games that the BLS plays. Over time the bureau fits the data to Labor Department unemployment claims data via annual benchmarking. This is where the biggest adjustments usually occur. When we look at the BLS data in retrospect, we are looking at data that has been refit to actual conditions. But the real time current release is fantasy.
7/6/23 Interpreting the first monthly release and attempting to relate to stock prices and guesses about Fed policy is a fool’s errand. Obviously everybody on Wall Street wants to participate.
The jobs release only matters for a millisecond as the market reacts to it. Then the market returns to trend. The real significance of the withholding data is not what it tells us about the jobs report. It is what it reveals about current revenues and the trend of revenues. It is the variability of revenues that tells us what to expect about Treasury supply in the near term.
