Don’t get fooled by the disappointing headline. Although August payrolls came in well below expectations and there were substantial lower revisions to the prior two months, the overall message was a reversion to form. The 3-month moving average fell back to the average for the whole of 2016, which is more than twice the pace needed to absorb new workers. While the unemployment rate ticked up, it is now tracking closer to its decline in 2016 than its decline in 2015, when job gains were substantially faster (about 226k per month). Wage growth held steady, but that’s nothing new.
Here’s your cheat sheet:
The diffusion index held at a healthy level (63.8 versus. 64.9), but the industry breakdown showed a shift towards good-producing sectors, perhaps reflecting the last few months of dollar weakness. Both construction (28k vs. -3k) and manufacturing (36k vs. 26k) saw substantial gains, and even mining eked out a small gain (6k vs. 0). Within services, losses were modest and confined to utilities (-500 vs. -600), information (-8k vs. -4k), and government (-9k vs. -13k). Retail trade seems to be leveling off after a few months of losses (800 vs. -1.9k). The meaningful slowdowns were in education & healthcare (25k vs. 54k) and leisure & hospitality (4k vs. 58k). The former seems like a blip given the ongoing aging of the U.S. population, while the latter falls under “could be worse,” given the lower average wages in that sector. A little more concerning was the drop-off in temporary payrolls, which are a leading indicator of job growth.
However, as we have previously noted and explain below, a little moderation in job gains would not be entirely unwelcome at the Federal Reserve. The most unambiguously negative signal in the establishment survey was the downward move in hours worked (34.4 vs. 34.5). That will slow growth in aggregate income earned by the household sector, but recent strength in consumer spending suggests this area of the economy will retain plenty of momentum.
Turning to the household survey, the bounce back in the unemployment rate is not surprising given the pace of job gains in 2017 relative to 2016 and 2015. Part-time work declined, but by too little to bring down the underemployment rate, underlining fears that this indicator will not converge on its pre-recession level relative to the underemployment rate. It wasn’t all bad, though: long-term unemployment declined and unemployment among those without a high-school degree fell by nearly a full percentage point, to 6.0%. Unemployment rose for all other educational backgrounds except college graduates, reinforcing concerns about job polarization.
The stability of wage growth remains a puzzle and while it may be an issue for income inequality and the welfare of workers, it will reassure policy makers at the Fed that they can continue to take a patient, gradual approach to rate hikes. Indeed, this report as a whole was a blessing in disguise for the Fed. The combination of strong job growth and soft inflation has been a challenge for policy makers in terms of understanding the outlook and calibrating their current policy. According to the minutes of the July meeting, this has become enough of a problem that they are now openly questioning the usefulness of their own traditional models. Moderation in payroll growth is actually somewhat reassuring in terms of getting their bearings in this phase of the economic cycle. It also suggests we may be closer to a reassertion of more traditional dynamics. That (along with thin trading) may be why expectations for a December hike seem to have risen modestly this morning.
As iCIMS’ former chief economist, Josh Wright led a team of data scientists in analyzing emerging trends in the U.S. labor market. With publications ranging from academic journals to national media, Wright previously served as a U.S. economist with Bloomberg L.P., and was a staff researcher at the Federal Reserve.