BLS private nonfarm payrolls increased by 444K, while ADP decreased by 301K and consensus increased by 150K (Bloomberg), a sharp revision from previous months.Do not over-read a lesson in the ADP series (see Employment Situation release).
figure 1: BLS private nonfarm payrolls from January release (bold grey), December release (black), Bloomberg consensus as of 2/1 (turquoise square), ADP (red). Source: BLS, Bloomberg and ADP, via FRED.
Note that employment levels have changed significantly throughout 2021; this is because the January version included annual benchmark revisions, which were release.
ADP private nonfarm does not track BLS private nonfarm very well at the level; there has been no difference in recent months either. (see this postal For analysis in the pre-COVID era. )
figure 2: BLS private nonfarm payrolls change since January release (bold grey), December release (black), Bloomberg consensus as of 2/1 (turquoise square), ADP (red). Source: BLS, Bloomberg, ADP via FRED, and author’s calculations.
US (total) nonfarm payrolls 95% confidence interval +/- 110K (BLS). If the private NFP confidence intervals are of similar size, then we can be pretty confident that the actual change is positive.
Note further that employment change for December and November has been revised up substantially in this benchmark revision.
Looking ahead, in terms of the January jobs revision, we can look back at the recent past. Average absolute revision of (total) nonfarm payrolls from Version 1 to Version 2 of 100,000 by 2021 (for the record) here). We don’t have full year data from the 1st to the 3rd.See more discussion of revisions in recent years, see here postal.
Although this number constitutes a large positive surprise and is outside the 60% prediction interval of my particular BLS-ADP regression (discussed here postal), which is within 95% of the prediction interval, thus confirming the limited information content of the ADP-private NFP of the BLS-private NFP.
