The June Average Hourly Earnings (AHE) number is what one should expect given “residual calendarity”, seasonal distortions, and sectoral shifts. We remain agnostic about the slope of the Beveridge curve and estimate that gap between the unemployment rate and the vacancy rate needs to fall by at least another 1pp for wage growth to be roughly in line with the pace consistent with the Fed 2% target.
The June AHE reading
The June AHE reading is not biased by seasonality distortions. Figure 1 shows the published AHE MoM across sectors. For the 3rd time in the last 4 months (see our notes here and here), the aggregate reading seems influenced by a large drop in one sector. In the previous episodes, we warned that the AHE reading was probably biased by seasonal adjustment distortions (a call which turned out to be correct by looking at the rebound of “other services” and “information” in April and May). This month, we do NOT think that the drop in “Information” is driven by the same distortion. The reason is that other seasonal adjustment methods have produced the same result (see for instance here). The impression is that, at least in the “Information” sector, nominal wage growth has been low/stagnant for about one year.
Figure 1. Published MoM of AHE (total private) and AHE sectors
Our AHE model
Our AHE model suggests that the June number is what one should expect controlling for “residual calendarity”, seasonal distortions, and sectoral shifts. We have updated our AHE model, results are shown in Figure 2. The model suggests that the June reading is in line with the state of the labor market. For the record, residual calendarity in June was neutral, and we could not identify any significant shift in sectoral employment shares. Therefore, we can take the June reading (MoM saar at 4.4%) as the “true” AHE pace right now.
Figure 2. AHE model – published AHE (MoM) and fitted values.
Note: the figure shows published MoM of AHE (orange line) and the fitted values of a model controlling for own lags, labor market slack, “residual calendarity”, and sectoral employment shares.
Coming down – fast enough?
The US economy is travelling, as expected, along a Phillips curve. Figure 3 shows our “agnostic” AHE Phillips curve. We have labelled this Phillips curve “agnostic” as we use the gap between the Urate and the vacancy rate as a measure of slack (in other words, we are agnostic about the slope of the Beveridge curve). The latest data in Figure 3 is June, assuming that the vacancy rate will remain unchanged. In the last 2-3 months the US economy has made some progress and it is travelling, as expected, along the estimated fit (only in May the economy has travelled 6 tenths on the horizontal axis in Figure 3). However, in order to bring wage inflation to a level consistent with the Fed 2% target, the gap between the unemployment rate and the vacancy rate is expected to drop by (at least) another percentage point (and preferably 1.5pp).
Figure 3. The “agnostic” AHE Phillips curve
Conclusion
Today’s AHE figure changes little for the Fed. From our point of view, we remain unsure whether the Fed will hike this month (yes, we have seen the market odds but there is also a chance to get a 0.2% MoM in core CPI next week). After the skip in June, what has changed to hike in July? We are not sure there is a good answer to this question.