Heading into today's payrolls report, there was some "whisper" expectation that the July print would be a blowout due to a spike in census hiring, however that did not happen and instead the BLS reported that last month 164K jobs were added, right on top of the 165K expected.
The strong headline print however, was weakened by substantial historical downward revisions, to wit: the change in total nonfarm payroll employment for May was revised down by 10,000 from +72,000 to +62,000, and the change for June was revised down by 31,000 from +224,000 to +193,000. With these revisions, employment gains in May and June combined were 41,000 less than previously reported.
What was perhaps most notable is that the report made no mention of hiring in advance of the 2020 census, and even more inexplicably federal payrolls rose just 2,000.
Looking at the 6 month moving average, it's obvious that the US economy peaked some time in 2017 and is all downhill from there:
- 2019: 124k
- 2018: 222k
- 2017: 227k
- 2016: 176k
- 2015: 205k
- 2014: 233k
- 2013: 154k
- 2012: 135k
Offsetting the weaker downward revisions was the average hourly earnings which came in hotter than expected, as the monthly increase came in at 0.3%, above the 0.2% expected, while the annual increase of 3.3% was also above the 3.2% expected, and once again approaching the cycle highs.
Alas, not even this data point was strong as the only reason why hourly earnings rose is because average weekly hours worked dropped again, sliding to 34.3 from 34.4, the lowest since 2011.
Commenting on this development, Renaissance's Neil Dutta said that the "drop in the workweek is disappointing. Aggregate hours worked up just 0.4% at an annual rate over the last three months."
Meanwhile, the far less notable unemployment rate was unchanged from last month, at 3.7%, just above the 3.6% expected. Of note: Hispanic unemployment rose modestly to 4.5% if still near all time lows.
Manufacturing payrolls climbed larger-than-forecast 16,000, reflecting 7,200 job gains in the motor vehicle industry. This was the biggest increase in mfg jobs since January.
More good news: Business jobs which include real estate, insurance and finance, had the biggest gain since February 2018.
And even more good news: the number of people working part-time for economic reasons plunged by 363K to 3.984 million, the lowest since April 2006.
And now for some not so good news: the prime-age employment to population fell 0.2%pts from 79.7% to 79.5%,'' notes Skanda Amarnath, director of research and analysis at the Employ America think-tan
Some more details from the report:
- Professional and technical services added 31,000 jobs in July, bringing the 12-month job gain to 300,000. In July, employment increased by 11,000 in computer systems design and related services; this industry accounted for about one-third of employment growth in professional and technical services both over the month and over the year. It looks like everyone is learning to code.
- Employment in health care rose by 30,000 over the month, reflecting a gain in ambulatory health care services (+29,000).
- Social assistance added 20,000 jobs in July.
- Financial activities employment rose by 18,000, with most of the gain occurring in insurance carriers and related activities (+11,000).
- Mining employment declined by 5,000 in July, after showing little net change in recent months.
- Manufacturing employment changed little in July (+16,000) and thus far in 2019. Job gains in the industry had averaged 22,000 per month in 2018.
Commenting on the report, Bloomberg's Eliza Winger writes that "The jobs data highlight the resilient labor markets, but the trend-like result is overshadowed by the latest escalation of trade tensions. Still, the more important aggregate income proxy remained relatively strong, supportive of sustained economic expansion."
So how does this affect the Fed's thinking or its reaction function? Well, it's a toss up - on one hand, the headline NFP print should not impact thinking either way, although on the margin, the downward job revisions are likely to offset the increases in hourly earning which were entirely due to less hours worked, as such this might help the doves’ case for another insurance cut.
Additionally, as Bloomberg notes "the three-month payroll average increase of 140,000 was the slowest in almost two years, and that's consistent with a slowing but still solid economy." At the same time, there is no sign the job market is overheating, with the unemployment rate steady and labor-force participation rising. "That suggests a strong jobs market is drawing more people into work, which is a good thing", and also suggests that the Fed has quite a bit of room to cut rates - if it so chooses - before pushing wage inflation sharply higher (if ever, considering the Phillips curve is now officially dead).
Here is CIBC's Avery Shenfeld with a somewhat non-committal take: "The Fed cut rates without much evidence that the U.S. economy was in serious trouble, and that's still the case after looking at July's payrolls figures...the data are consistent with our call for just one more Fed cut, given that the U.S. actually needs to see growth no higher than 2% to avoid overheating the labor market."
But the best analysis is probably that of Andrew Hunter who said that "overall, this report won't be enough to move the needle much in either direction as far as a September rate cut is concerned, but it reinforces our sense that another move next month isn't yet as sure a thing as the markets are now pricing in."