There is a growing puzzle among central bankers, one that is both familiar and unique at the same time. When surveying the major employment figures, there doesn’t seem to be anything like recession in them, though that’s what we always see right before the contraction strikes. This time, however, there may be a legitimate reason beyond the customary trend-cycle assumptions of government statisticians.
During any business cycle, managers and owners are quite understandably reluctant to give up on masses of workers. Individuals, sure, yet largescale layoffs are unpalatable for any number of intuitive reasons beginning with what that would say about a firm’s prospects.
More than any other time, in 2023 we continue to hear instances where managers openly confess to maintaining their payrolls well in excess of actual business. Economists call this labor hoarding and though it is typical at business cycle turns this time there’s more to it.
The most common explanation for the behavior is that the costs, perceived or real, are far greater firing then hiring back workers than they would be just keeping employees on even if they’re largely idle. It’s worth it in the long run to take the short run hit to margins and profits.
Obviously, the key to the tactic is “short run.” Conducted over the longer term, this would be madness or some kind of charity, not pure economics. There is and must be a certain informal threshold beyond which the initial cost/benefit conclusion no longer fits.
Back in February, Skynova surveyed 1,010 small businesses and found 91% of them admitted to hoarding workers; 89% of them indicated they would keep it up throughout this year. Broken down by size, of the larger small firms 99% confessed to stockpiling employees in one fashion or another.
According to the same survey, owners self-reported an average of $4,541 in savings by doing this. The calculation is thought to be much higher than usual because of the unique characteristics of the post-COVID world. Whether a real money number or not, it only matters that employers appear to believe in it enough to make these decisions.
What that means is also quite simple: these same businesses as well as many others out there large or small have a much higher tolerance for idling labor than at maybe any other time since the Great Depression (when the original modern concept of hoarding was effectively born). Where employers might’ve pulled the trigger in the past, today they’re going to stick it out even knowing there’s a downturn coming.
This is exactly what many corporate CEO’s have already said. Though the tech layoffs last year and earlier this year made lots of noise for good reason, those have been the exceptions thus far. Dana Petersen, the Conference Board’s Chief Economist, told CNN last week:
“CEOs have been saying for at least a year that they expect there will be a recession in the US, but it’s going to be shallow, and it’s going to be short. So if you think that the recession is not going to be too bad, and it’s not going to be too long, and you spent a lot of money trying to attract and retain labor, you’re less likely to let those people go.”
But there’s the qualifier: “not going to be too bad.”
The economic problem behind labor hoarding is just how procyclical it might be, and fast. You’d think otherwise this would be a positive (as the Keynesians put it), more people holding on to jobs therefore able to continue spending (aggregate demand) right when a weak economy needs this the most.
Yet, labor hoarding can’t pull a troubled economy out of a rut it has already fallen into. What ends up happening instead is that the initial stage(s) of the recession tends to be shallower. Without a realistic escape, once the initial hope is lost the spark of layoffs becomes a self-feeding frenzy.
To anyone watching the main labor statistics, it would be as if a questionable economy suddenly and out of nowhere turned ugly in an instant.
This was, of course, the major macro mistake made by policymakers everywhere fifteen years ago. They commonly cited labor data as the reason they believed the US (or other places) would avoid any recession at all in 2008. Even very late into it, several Federal Reserve policymakers were referencing government payroll data as the basis for these optimistic, even inflationary projections.
Here's just one example from the FOMC as late as August 5 of 2008:
“MR. LACKER. I expect payroll employment to continue to decline for a while at about the current pace—a pace that, as others have noted, is quite modest relative to what we typically see in a recession. I think the most likely outcome is for us to continue to skirt an outright recession.”
In December 2008, Robert Hall, at the time occupying the Chair of the NBER’s Business Cycle Dating Committee, explained why it took so long for the self-styled cycle-daters to make a date determination, saying “Employment declined less than is normal in a recession until about September [2008]. Then so many negative numbers came through that made completely clear this was a recession.”
Quite simply, many businesses had made the same mistake as the Economists and policymakers; hell, they often took officials and academics at their word. Darn near everyone said if there was a contraction in 2008, it would be mild…until it wasn’t.
Even before September when Lehman and AIG hit, the tide had already turned. The payroll reports came in more negative starting in July then August as many business owners and managers gave up waiting for the second half rebound (that never really comes). The monetary crisis and financial meltdown ahead in late September and October just piled more awful onto what was already becoming a procyclical labor disaster-fest.
Had there not been a meltdown in markets during the autumn, it is more than possible – though debatable – the US was in for a rough ride anyway given the way the economy was already trending. A key reason why might’ve been the sudden, quick de-hoarding of workers that summer which produced the devastating positive feedback loop.
Before even getting that far, there were ample warning signs of serious cyclical weakness. The strain of monetary and banking crisis had already negatively impacted the real economy (this wasn’t just true of the US, same for Europe and elsewhere). Alternate government statistics from the same survey as the main employment figure showed that companies across the system had been taking defensive measures along with hoarding workers.
Hours worked, for example, which like the headline payrolls comes from the Establishment Survey, the index for the systemwide aggregate had peaked all the way back in June 2007. Throughout the second half of that year, total hours went sideways indicating a modest contraction in labor utilization while payrolls remained at small positives.
Throughout the first half of 2008, sideways in hours became a slight negative track lasting until around July and August as employers struggled to manage their costs against growing recessionary problems while still largely hoarding employees.
Another cyclical indicator and also from the Establishment Survey, the number of temp workers in the US topped out even earlier, back in January 2007 at 2.65 million. Over the next year, companies would hire fewer temps along with shedding a few they’d already employed; the level dropped to 2.55 million by January 2008.
The rate of decline then picked up substantially so that by August the number was down to 2.34 million.
These obviously aren’t huge numbers which is why it was easy to overlook them. Put them together, however, and what they indicated had been substantial problems placing companies under enormous strain. In short, the labor market, indeed the whole of the economy during that period, especially the first half of 2008, was significantly weaker than many if not most had realized and the major figures had represented.
Once enough company managers woke up to the dangers each was facing, and how they were likely to get worse, combined with the monetary crisis it unleashed a torrent of layoffs unlike anything witnessed since the Great Depression. The much lesser evil was how policymakers and their like were caught so unaware by this.
Here we are all over again and maybe this time the hoarding is substantially more widespread. On top of that, we’re seeing the same cyclical indications.
The hours worked index peaked in December 2022 and is actually down 0.2% through May (a recession warning from the start). Companies have been shedding temporary workers ever since last March, with a particular emphasis also around November and December. There had been 3.18 million temps at the top last year; as of last month, down to 3.04 million.
According to central bankers around the world, however, labor markets continue to be way too “tight.” Just this week when announcing its pause in rate hikes, the FOMC’s statement referenced how, “Job gains have been robust in recent months, and the unemployment rate has remained low.”
The headline payroll number might show this, but is it anywhere close to accurate?
Economists at the Fed would no doubt argue they are reliable considering these days they’ve been far more positive than the numbers had been in 2008; not a single negative has yet been spotted. In fact, according to the Establishment Survey, job gains accelerated after March through May beyond 300k (exactly what the FOMC statement was talking about).
It appears, however, as if not everyone on the committee agrees. How else to explain the Fed’s pause? If what others, the so-called hawks say about the labor market is true, believing (incorrectly) in the Phillips Curve cause for inflation, then there is no time to delay, no apparent reason to pause hikes in the first place.
Perhaps a substantial minority sees the hoarding and cyclicals, understandably arguing for time to see if the rug really is in process of being yanked out from under the workforce.
And if labor hoarding this time is much more pronounced than any other time, should de-hoarding trigger it could very well mean far worse to the downside; especially if combined with the next inevitable stage of the ongoing banking saga. One day the labor market is declared robust, only to find out the next it is undergoing a massacre.
All it will take is a growing perception that the economy is in worse shape therefore holding on to employees isn’t saving money rather costing a whole lot more companies realize they cannot afford.
This is where the recent jump in initial jobless claims could be so much more than it, too, appears. This week for the second week in a row, the seasonally adjusted level reached 261,000. While also a historically low number, it is the highest since October 2021 and if filings keep going in this direction that could potentially represent the spark already igniting the powder keg.
Just considering the possibility would give anyone but the most hardened hawk a legitimate pause.
The real problem at the Fed like the ECB is that the world economy is already showing these familiar signs (see China) before CPIs or PCE Deflators have come down enough for them. That, too, was a problem in 2008 if also to a smaller extent. But if the degree of current hoarding, therefore the potential for all-at-once de-hoarding really does turn out to be true, policymakers will need a miracle not a pause.
