Confident consumers are risk takers. Not only do they spend freely, they freely borrow in order to spend. Jay Powell has done his absolute best (I know) to convince Americans they have nothing to fear insofar as any economic fallout from COVID might be concerned. The Federal Reserve working in combination with the federal government has got every conceivable angle covered. In the multiple trillions. 

So, stop worrying.

Yet, the Federal Reserve also reports that some substantial slice of the consumer cohort is worried. Those within it must be. Revolving consumer credit is a window into whatever is beyond sentimental perceptions; what consumers are actively doing with regard to their own personal perceptions and their own personal risks.

What many are doing now, as they have been since March, is shunning revolving debt. Seasonally-adjusted, the Fed in its latest figures estimates that the aggregate balance of revolving consumer loans declined in July 2020 by a small amount. But that makes it five straight months; three of them with gigantic positives in economic accounts like the unemployment rate joining monetary officials in egging consumers on.



If there really is such widespread concern contained within these July 2020 credit numbers, and there is, what must be behind it?

The answer, as usual, is the labor market underneath the headline unemployment rate and payroll numbers. As we’ve witnessed and catalogued over the past few years of a persistently tumbling unemployment rate, the real state of the economy can be very different from the one referenced in these mainstream employment accounts – especially the unemployment rate (as even the Federal Reserve has finally admitted).

It’s not simply a matter of reopening leading to recovery, it is the possibility that reopening can or even likely well fall well short of one. After the frenzied few months of the first stages of the economy getting back on its feet, was it really sustained massive growth as the unemployment rate’s quickened downward trend has implied?

The BLS’s JOLTS series suggests otherwise. Employers are advertising more for new or unfilled positions than they had been a few months before at the bottom, but still not quite at the same levels as before March 2020. Both the JOLTS version of job openings (JO) as well as other measures of the same aspect of demand for labor, such as the Conference Board’s Help Wanted OnLine (HWOL) index, continue to lag.



Each was about 10% less in July 2020 when compared with July 2019 (which, as you’ll note above, wasn’t exactly a robust time period). Those are the kinds of recessionary results which are consistent with the recession we now find ourselves in. This is not necessary unexpected. 

But it wasn’t JO (or HWOL) alone that might explain July’s consumer credit estimate – and maybe a whole lot more than that. The entire issue has always been whether or not everyone laid off and involuntarily separated from their work by the unnecessary combination of COVID overreactions and GFC2 underreactions get to go back.

The more time passes without the gap shrinking fast enough, the more former workers will realize they had actually drawn the short straws; longer-term problems which undoubtedly create harmful uncertainty, maybe even paralysis, in a larger proportion of the working population beyond strictly those who were unfortunately let go.

Even workers currently working are in this kind of environment pretty keenly aware of the real condition in their slice of the labor market. And if it’s not really all that robust, that can have negative impacts on their behavior, too, even if it doesn’t directly show up in the unemployment rate, Establishment Survey, and Household Survey (which is where the unemployment rate comes from).



Tucked away in the JOLTS data was an alarming deceleration in the rate of Hires (HI) back during July. Reopening had caused several months (May & June) of frenzied re-hiring as shutdown restrictions were lifted and millions upon millions were allowed to go back to their work.

However, that positive trend seems to have hit a snag two months ago. As you can see above, the rate of HI slowed sharply at that time, potentially indicated that, for unspecified reasons, the labor market rebound may have materially downshifted.

In terms of JOLTS, it left the hiring rate (HI) more than 3% below what it had been during July 2019. That’s not a rebounding labor market, certainly not a robust comeback from a deep trough.

And while this may prove nothing more than statistical noise, JOLTS data is inherently messy and noisy month-to-month, what really makes the HI figure alarming is what else we’ve already seen elsewhere about what might have happened during July. There has already been a slew of data which already has indicated a substantial slowing in rebound/reopening momentum.

From the private side of the Establishment Survey and its corresponding number pieced together by ADP, to, more importantly in my analysis, a sudden upturn in federal (not state) unemployment claims. Each of these took a wrong turn like HI during July.



And that’s not all – how many markets have we seen which also appear to have experienced some degree of change in axis right during the same month (or, in the case of jobless claims, the same exact week of that same month)? Too many to ignore.

In other words, the HI estimate isn’t proposing on its own a potentially negative economic setback, it is merely the latest on a lengthening list already confirming and corroborating that very idea.

More and more it looks like a real problem developed during the early part of summer. July. July. July. July.



If the labor market – and by extension rebounding economy – did indeed downshift all the way back a couple months ago, at the very least it will have greatly reduced the chances of everyone getting to go back to work; significantly increasing the risks of further consequences from an economy that doesn’t get right back to normal.

From oil to credit spreads to bond yields, maybe even stocks, there does seem to be a palpable spreading of concern now in September as to those possibilities.

Here we are once again; when a broader survey of employment data even for two months back suggests one thing that the unemployment rate does not. Even if the shape of the economy has, in terms of the usefulness of some of these statistics things never change.