June saw a surprisingly strong nonfarm payrolls report, on the surface, and stood in stark contrast to the weakness we saw earlier in the consumer confidence, ADP and ISM reports. From the survey data, we had detected that for every piece of positive anecdotal evidence for the labour market in June, there were three negative reports. But the employment data Friday painted a much rosier picture, and the futures market quickly priced out any notion of a 50-basis-point rate cut at the end of the month but left a 25-basis-point move intact. The stock market, being in a “bad news is good news” mentality, has not exactly responded favourably to the “good news,” since for equities, it’s really all about the Fed. Who needs a strong labour market?
The headline payroll tally was +224k in June, not just beating the consensus estimate of +160k but also topping the high end of the range (which was +220k). There is really no sign of any cooling off, either, with the average of the past three months of +171k essentially unchanged compared to the average of the prior three months. The private diffusion indices improved dramatically to 60.7 per cent, from 57.6 per cent for the entire private sector, and to 59.9 per cent, from 52.0 per cent, for manufacturing. The world economy is slowing down precipitously, but the BLS is trying to convince everyone that the U.S. labour market is its own little oasis of prosperity. We saw the weakest ISM manufacturing report since October 2016, and yet factory payrolls in this release (+17k) showed the best tally since the turn of the year. We know that construction spending has remained on a decisive downtrend, and yet construction employment managed to jump 21k in June. None of this spells great news for productivity in these two goods-producing sectors, if these employment numbers are to be believed.
Maybe they aren’t to be believed. Outside of students, multiple-job holders, government workers and fictitious additions from the birth-death model, there wasn’t really a whole lot here to get excited about. In fact, if you are a male between the ages of 25 and 54, you have grounds to be outright depressed.
There are valid reasons to take the payroll data with several grains of salt, as is always the case at turning points in the economic cycle. Notably the BLS birth-death model, which is pure guesswork and adds an estimate of jobs from net new-job creation, accounted for roughly half the payroll gain in June. If this was just about the existing sample of companies in the Establishment survey, in other words, that headline would have been closer to +150k or near the lower end of the consensus. In fact, since February, nearly every single job that has been added to nonfarm payrolls has come via the birth-death model – employment growth from the actual sample of firms in the survey has completely stalled out.
The companion Household survey almost matched the headline payroll result with a 247k employment gain. That said, there has been virtually no growth in this metric through the first half of the year. But even here, there were some sore spots beneath the surface. Those working part-time because of “slack work or business conditions” rose 61k. That should not be happening if the economy is actually in solid shape.
Of that 247k headline Household survey job spurt, 222k – 90 per cent – took hold in the public sector (perhaps census-related). Private sector wage and salaried employment rose a grand total of 31k in June after a three-month cumulative 172k plunge; perhaps that is the metric we should be focused on today as markets price out a more aggressive Fed ahead. Multiple-job holders, a contra-cyclical indicator, soared 301k in June. Think about that – all of the jobs and then some came from people taking on a second (or third) job. Not exactly a bullish economic signpost.
Nearly 60 per cent of the Household survey job increase also emanated from the ranks of the self-employed (+143k), which again is a metric that moves with an inverse correlation to the direction of the overall economy. The number of ‘discouraged’ workers shot up 18 per cent from a year ago. I should add that we also had 60 per cent of the total employment gain coming from teenagers (16-19) – a 137k expansion, so the data seem to also have been somewhat inflated by an seasonal influx of summer students. Those who graduated with a college degree, or have at least some form of post-secondary school, were less fortunate, as their employment level declined 205k and has fallen now in two of the past three months.
The real question is how to square the seemingly solid payroll report with the details of the Household survey. Employment for the prime-adult 25-54 year cohort eked out just a 29k increase after a three-month slide, and the number of jobs for this breadwinner group has plunged 168k so far this year. This is strong based on what measure? Prime-adult male employment actually sagged 97k in June and has collapsed 338k over the past three months, which hasn’t happened since September 2009. Of course, I don’t expect many other folks to lift open the shiny hood and check out the sputtering engine.
The average work-week for production and non-supervisory workers also stagnated in June and has declined or flattened now in five of the past six months. Of all the labour market segments, this is the one that has the most reliable leading characteristics – and it peaked in January and has shrunk at a 1.4 per cent annual rate since that time. The second-best leading indicator from within the employment data is the temp agency sector, where job growth practically vanished in June and peaked last December – and has contracted at a 1.4-per-cent annual rate since then. It’s never too encouraging when the headhunters start cutting their own head count. Something tells me that 50 basis points from the Fed may end up coming back on the table before too long.
The various measures of labour market underutilization also showed a bit more slack coming back to the fore, with the U-3 unemployment rate inching up to a three-month high of 3.7 per cent from 3.6 per cent in May; the U-6 rebounded to 7.2 per cent from 7.1 per cent. Still a tight labour market, don’t get me wrong, but less so. Wage growth, as a result, remained very well contained at +0.2 per cent sequentially, the year-over-year trend in average hourly earnings at +3.14 per cent is the slowest in nine months, and the three-month trend at +2.77 per cent suggests that the near-term trajectory is cooling off even more substantially – again, not a typical hallmark of the robust jobs market that the headline payroll number is flashing.
David Rosenberg is chief economist with Gluskin Sheff + Associates Inc. and author of the daily economic newsletter Breakfast with Dave.
David Rosenberg