The headline writers for this morning’s payroll reports can’t really help themselves. In a clickbait world, any kind of record or new high or low is bound to make its way into every article title. The unemployment rate fell to 4.6% in November, the lowest since 2007, therefore it isn’t surprising to see reporting on the labor statistics to have focused on that positive aspect.
The problem is what is left out of that decline, especially when combined with leading adjectives like “tumbled.” This article’s headline, for example, from Business Insider declares Unemployment Rate Tumbles To A 9-year Low. All of those words are true, including the “tumble” since the rate was at 5.0% in September only to drop 0.4% since, 0.3% in November alone. A decline in the unemployment rate having such positive economic connotations leaves the impression that this can only be a very good sign for the economy as a whole.
And it is a mistaken one, but one that is a perfect microcosm of how the unemployment rate has brought on so much confusion and malformed expectations the past few years. The Household Survey, which is the numerator, showed only 160k in job gains in November after declining by 43k in October, for a two-month net of just 117k. In the denominator, the official count of the labor force fell significantly in November, by 226k, for the second straight month. There was a decline of 195k in October, for a two-month net total of -421k.
The result is an unemployment rate that drops from 5.0% to 4.6% because of very weak gains on the top and far, far too many people exiting the labor force on the bottom. This is the same problem the economy has, of course, been facing consistently since the Great “Recession.” The past two months show very well the struggles of the unemployment rate to represent what is actually happening. In this specific instance, what the unemployment rate suggests is actually the opposite of the data that creates it.
Because it is at 4.6%, almost every mainstream economist will claim this is “full employment”, a significant accomplishment toward recovery if it were true. At that point we would expect not only robust job gains in the numerator but that those job gains would pull more and more people back into the labor market, increasing also the denominator. That, historically, is what has happened each and every time before.
Going back to the middle of 2014 when the unemployment rate really moved interpretations as well as expectations too far from reality, Janet Yellen, as noted earlier this week, spoke about this as a factor determining actual economic success. It was her expectation at that time, and still seems to be, that the labor force would be increased by economic acceleration in the quarters directly ahead.
My hope would be that as—and my expectation is that as the economy recovers, we will see some repair of that, that many of those individuals who were long-term unemployed or those who are now counted as out of the labor force would take jobs if the economy is stronger and would be drawn back in again, but it is conceivable that there is some permanent damage there to them, to their own well-being, their family’s well-being, and the economy’s potential.
In the more than two years since that speech, the only factor that has drawn a significant number of Americans into the labor force has been the reinstatement of work requirements for SNAP benefits.
If we subtract those six months of clearly anomalous gains (the 6-month change was the largest surge since the recovery in the early 1980’s, meaning that we are supposed to believe that Americans were so confident about labor prospects right smack in the middle of “global turmoil” that they started looking for work in numbers that rivaled the last real, robust recovery in US history; or it was SNAP) the labor force is practically unchanged since January 2015 – right when all this “unexpected” weakness due to the “rising dollar”, or strong dollar as many tried to claim at the time, began to show up in a broad section of economic accounts.
The total labor force was figured to be 157.03 million for January 2015, and is estimated to be 159.49 million for November 2016. Subtracting the ridiculous 2.42 million that showed up between August 2015 and March 2016 when the various states were resuming their work restrictions, the net total for November would be 157.07 million and practically the same level as almost two years ago despite 4.8 million new Americans in the civilian non-institutional population. Again, the only variable that seems to have pulled in “workers” was the prospective loss of welfare benefits, not perceived economic gains in jobs and otherwise. That is a very different view of the labor market than is perceived by the unemployment rate “tumbling.”
The other component for “full employment” is wages as a product of diminished or more likely entirely used up “slack.” As the actual pool of potential workers find their way into the labor force in an actually growing economy, competition for what few prospective workers remain, as well as to obtain employees who are already employed, requires significant wage gains. Thus, the disappearance of slack would lead to clearly accelerating wages, as has also been the case in all prior historical experience at times approaching full employment.
Last month, the payroll report for October 2016, showed rather large wage increases. Average weekly earnings for all employees rose by 4.4% year-over-year, the fastest growth in over a year. For production and non-supervisory employees, the growth was slightly less at 3.5%, but still a much better result. Predictably, the media took these numbers as if confirming the unemployment rate. From the WSJ coverage of analysts and economists last month:
“Hourly earnings are rising at their fastest pace since 2009. It will strengthen ideas that the labor market is getting tighter and pushing up wages. Higher wages are understood to drive core inflation. Core inflation is what the Fed targets.”
And right on cue, wage gains in November “tumbled”, coming in very disappointing across-the-board. For all employees, average weekly earnings were essentially flat year-over-year, up just 0.4% in November. The growth was even flatter for production employees, only 0.3% higher than November 2015. What you never see in the media is that these extremes are typical month to month; as you can clearly see below, wage gains are highly variable, so highlighting only one month when wages are high is misleading.
The average growth in earnings has been continuously subdued, with more notable low months than high since the middle of 2014. There is no sign at all of an end to “slack”, and therefore the idea of “full employment” is once more dispelled from both sides – labor force as well as wages. These are, instead, clear indications of a very weak economy that got weaker as it absorbed the effects of the “rising dollar.”
Whatever the headline Establishment Survey number, and it was once again below average, the economy remains stuck in its lower state. But with only one more payroll report left until the new administration comes in, this is the penultimate version that will be declared strong where it so clearly shows the opposite. I have no doubt that the misleading infatuation with the improper unemployment rate is nearing its end.