A quick note to say that, according to our estimates, the Average Hourly Earnings (AHE) slowdown in August (from 42bps MoM sa in July to 24bps MoM sa in August) is almost entirely due to “residual calendarity”. In other words, there was no real slowdown in wages, not yet. All else equal, “residual calendarity” will go the opposite way in September when we expect the AHE MoM to round back to 0.3%. Overall, the “true” pace of nominal (AHE) wage growth controlling for shifts in sectoral employment and calendar effects is estimated around 3¾ percent (consistent with our estimate of pi* at 2¾ percent).
(A PPT containing all US wage charts has been updated and posted here)
Residual calendarity
Monthly AHE figures are distorted by calendar effects not corrected by the BLS. We have already explained this effect in the past. The reader can refer to this BLS page, and to our previous notes (here, here, and here). What the reader should keep in mind is the following: months with a higher number of weekend days (Saturdays and Sundays) are associated to a higher AHE MoM growth rate. This effect is called “residual calendarity”. The (estimated) rule of thumb is the following: one extra weekend day boosts the AHE MoM growth rate by about 6-7bps.
The August AHE figure and our AHE model
The slowdown in August is almost entirely due to “residual calendarity”. In August, AHE grew 24bps (MoM sa) from 42bps in July. The key point is that July had 10 weekend days (the highest possible) while August had 8 weekend days (the lowest possible). Therefore, the estimated negative drag from “residual calendarity” in August is 13-14bps. In other words, the drop of the AHE MoM in August is almost entirely explained by “residual calendarity”. Another way of putting this is to say that conditional on (i) where the labor market is and (ii) on the composition of employment across sectors, AHE wage growth is moving roughly sideways.
Figure 1 shows an update of our AHE monthly model. As a reminder, this model explains monthly AHE growth rates using recent past, a variable capturing “residual seasonality”, a variable capturing economic “slack”, and sectoral employment shares. In August (latest point on the chart) the fitted value of the model is nearly identical to the published AHE growth rate. In other words, our model is able to fully explain what happened last month. As mentioned, according to the model, 13-14bps (out of the 18bps) of the observed deceleration in August are driven by “residual calendarity”, and not by a real slowdown of wage growth.
Figure 1. AHE MoM (sa) and UnderlyingInflation AHE model fitted values.
Note: the figure shows published MoM (sa) AHE growth rates (orange line) and UnderlyingInflation AHE model fitted values (blue line). The figure excludes the first 6 months of Covid for scaling issues (although they are included in the model estimation).
(Note: at the sectoral level (figure here), the August report showed a large increase in “Financial Activities” and a large drop in “Mining and Lodging”. As we have previously discussed, most series suffer from seasonal adjustment issues in this moment. For instance, see here for a figure comparing the published BLS level of AHE in “Financial Activity” sector vs a series seasonally adjusted in-house. In any case, the net effect of seasonal adjustment issues at the sectoral level in August was, according to our estimates, close to zero).
What to expect in September
The “true” AHE is growing at an annual pace of around 3¾ percent. September has 9 weekend days, one more than August. Therefore, “residual calendarity” is expected to add about 7bps to the AHE MoM in September. This implies that net of changes of employment shares across sectors, and without large movements of the unemployment rate (or the vacancy rate), it is reasonable to expect AHE to expand 31bps in September (24bps in August + 7bps from “residual calendarity”). It also implies that the “true” AHE growth rate (that is, net of all possible high-frequency distortions) is close to 30bps per month, or 3¾ percent at annual rate. Indeed, this should not be surprising because it is fully consistent with our current estimate of pi* at 2.7% (see here).
Conclusion and implications for the Fed
The Fed knows, and it can be patient. In our experience, the Fed staff knows well about “residual calendarity”. Returning to 2% requires an additional help from the labor market (the reader can refer to the AHE wage Phillips curves in the posted PPT) but, as we have been arguing since June, the Fed can now be patient. (P.S. To us, what is happening on the bond market right now is consistent with the view of a less aggressive Fed that, if anything, keeps the rates high for longer but stops being aggressive with the front-end)