The U.S. Labor Market Is Weaker Than Consensus & Recession Risk Is Higher
Job growth has ground to a halt (and could be negative) and plummeting labor demand points to continued weakness in the future, thus raising the risk of a recession in 2026
It feels like an eternity since we’ve had a regularly scheduled ‘Jobs Week’ (remember Labor Day?) and while this week doesn’t yet qualify as one in any event, the lights are finally starting to come fully back on from the data black-out we’ve been operating under for some time. As we wrote in early November and early December, we are forecasting that the U.S. economy lost 5,000 jobs in October and gained 45,000 jobs in November.
But more importantly, and with the assumption that the numbers coming out later this morning will come in approximately as we expect, not only are we forecasting further job losses in December, we believe that the job market is significantly weaker than most people believe. As a result, we continue to assign a higher probability of a recession, of any kind, in 2026 along with a higher probability of a severe recession.
Last week’s rate cut wasn’t much of a surprise given that we’ve pretty much been sounding the alarm on the moribund labor market and deteriorating economy for most of the year. It did nothing, however, to resolve the insane levels of uncertainty and bitter division surrounding the state of the economy, the balance of risks around inflation and unemployment, and what’s in store for the coming year. In fact, the fractured vote, with dissents and shadow dissents arising for different reasons including hawkishness, curiosity, and sycophancy, exacerbated tensions even further.
As the WSJ noted, ”Recent divisions reflect a dilemma the Fed hasn’t faced in more than 15 years: stubborn inflation that calls for higher rates paired with a softening job market that suggests a need for lower rates. “So what do you do?” Powell said Wednesday. “You’ve got one tool. You can’t do two things at once.”
Not only that, the WSJ also noted that the “division reflects a committee split over a fundamental question: not whether it could be on the brink of a policy error, but which one. Doves see a softening labor market despite 1.5 percentage points of rate cuts over the past 15 months—now 1.75 with Wednesday’s move. They worry that waiting for proof of weakness will mean acting too late to prevent it. Hawks see a central bank cutting into an economy stronger than it looks, risking a repeat of the “stop and go” mistakes of the 1970s.”
As always, data holds the key and the central bank indicated as much by amending its policy statement to say that it would assess the incoming economic data, evolving outlook and the balance of risks “in considering the extent and timing of additional adjustments.” But Powell also “warned that even when the flow of official data fully resumes, it should be looked at with a “skeptical eye.” The government shutdown disrupted data collection, which could make the initial reports confusing for what Mr. Powell called “very technical reasons.”
Add to those technical issues declining survey response rates, persistent underfunding of the BLS, the risks of endless government shutdowns as far forward as anyone can see, and material risks around the politicization of the BLS and it’s clear that the efficacy of official data is, at best, under attack and at worst, in serious peril.
So debates are raging around essentially every single aspect of the economy these days. And while that is nearly always the case, particularly in the covid era, the present battles feel especially intractable and center around where inflation and unemployment are heading, which of the two presents a greater risk to the economy, and what the neutral rate is.
On that last point, and despite the most dissents in 15 years, Powell indicated that with last week’s cut, “we are now in neutral territory,” a contention refuted by others on the committee and to which the WSJ immediately retorted, “That’s debatable. When the rate cuts began more than a year ago, officials predicted they’d get inflation back to the Fed’s 2% target by 2026. That deadline keeps getting pushed back, and in the Summary of Economic Projections released with Wednesday’s policy decisions, the 2% arrival date is now 2028.”
And before we even get to arguments around the dual mandate, it’s worth noting that debates rage around the Fed, Fed Capture/Takeover, the appointment of a new Chair, the Fed’s reaction function to ever-changing conditions, and even the fundamental efficacy of the Fed’s single/primary tool in their arsenal in today’s economy.
But leaving all of that aside and unpacking inflation first, sub-debates center around the impact, severity, timing, and reversals of tariffs, fiscal policy and all things emanating from D.C., immigration and deportations, consumer resilience, the inflationary/deflationary impact of AI, the wealth effect of the markets, geo-politics, and the Chair selection saga, to name just a few.
And layered not too far below the surface is growing skepticism around the 2% target itself and increasing calls for either a target range or a higher target altogether. Along those lines, even the Fed tripped over contradictions last week in expressing its commitment to a 2% target even while cutting rates in the face of persistently elevated inflation closer to 3%.
As Bloomberg’s Joe Weisenthal wrote:
The committee says it remains focused on the 2% inflation target. But of course the committee is going to say that. The [FOMC Statement], though, says that inflation is not only elevated but rising, and here they are cutting rates anyway. So I don’t think it’s crazy when people say that implicitly the Fed has already abandoned its target.”
Although I freely admit that I am infinitely less qualified to weigh in on inflation than I am on employment, I’m sticking with our consistently-held view that inflation is going to remain in a +/- range of 0.5% around 3% and all of the issues listed above will net out such that inflation remains higher forever.
Turning to employment and here again, before we even get into the actual conditions themselves, I’d note that endless battles rage around how we frame the job market and determine the measures by which we assess its condition (e.g., unemployment, job gains, vacancies, layoffs, wage growth, real income, supply/demand balance, sentiment, seasonal adjustments, alternative data, etc.).
Digging into the condition of the job market, then, sub-debates include BLS credibility, data quality, market fragility, market liquidity, the impact of AI/Robotics/Technology, productivity, immigration/deportation, sentiment, tipping points and reinforcing cycles (e.g. the Sahm and Perkins Rules), The Phillips Curve, The Beveridge Curve, globalization, tariffs and trade, and on/off-shoring, to name a few.
The overwhelming consensus view, at the moment, is that while the job market is ‘softening’ or ‘cooling’ with flat labor demand and a modest decline in monthly job gains, it remains balanced with retiring baby boomers, low firing, low initial claims, and no meaningful uptick in unemployment and will, furthermore, be perpetually buoyed by resilient consumers whose credit limits and spending know no end, corporate earnings that will remain perpetually above trend, markets that never go down, and AI that will solve all of civilization’s ills.
We couldn’t possible disagree more with that overly sanguine, pollyannish read.
By nearly every measure, the job market is deteriorating at a rapid and accelerating clip. Labor demand is plummeting, job gains have turned negative and will continue to drop, unemployment will shoot up rapidly and may already have, layoffs continue to tick up every week, sentiment is abysmal, Washington has taken a flame torch to nearly every cylinder in the engine that drives the economy, and the risk factors around the economy are growing everywhere all the time.
For the time being, with more to come in subsequent posts, we’ll focus on job growth and labor demand as the two primary indicators of the horrendous shape of the job market.
So far this year, quarterly job growth has dropped by over 65% and if the Fed Chair is correct that job growth has been overstated by 60,000 jobs per month since April of 2025, job growth has actually disappeared altogether since then.
As the WSJ noted regarding the potential overstatement of job growth…
Fed Chair Jerome Powell pointed on Wednesday to a job-market risk that economists have been worried about for months: Official statistics [related to the BLS birth-death model] could be drastically overstating recent hiring.
Powell said that Fed staffers believe that federal data could be overestimating job creation by up to 60,000 jobs a month. Given that figures published so far show that the economy has added about 40,000 jobs a month since April, the real number could be something more like a loss of 20,000 jobs a month, Powell said.
“We think there’s an overstatement in these numbers,” Powell said in a press conference following the central bank’s two-day policy meeting.
Subtracting 60,000 jobs each month shows just how ugly the job market might actually be.
We’ll see later today what things look like through November but because there will be subsequent revisions and continued questions around the technical details of job growth, combined with the fact that we aren’t going to get October unemployment data, the release today may not provide sufficient clarity to resolve any of the debates around the job market.
What is very clear (and backed by actual data), on the other hand, is that labor demand has plummeted since March of 2025. Looking at LinkUp’s U.S. job vacancies data indexed daily directly from company and employer websites globally, labor demand has dropped roughly 15% since the beginning of April. In stark contrast, JOLTS data from the BLS indicates that labor demand has risen 7% through October.
This is not the first time our data has deviated materially from JOLTS data. While the two data series generally track one another over time, there have been regular periods in the covid era where the two have diverged significantly, most notably in 2023 and 2024. In both of those years, our data showed large, persistent increases in labor demand while JOLTS data indicated large, steady declines in labor demand.
Given the fact that during those years the job market repeatedly surprised everyone (except us), again and again and again, to the upside relative to consensus, there’s no doubt which data provides a better, more accurate, and timely indicator about the present and future state of the job market and, therefore, the broader economy.
Given the sharp decline in job vacancies in the U.S. in November, we’re also forecasting more grim news for December job growth.
So clearly the job market is in much worse shape than consensus and there are no indications in the data that things are going to turn around anytime soon. In fact, the rapidly deteriorating job market most assuredly sits atop the list of known risk factors that could spiral the economy into a recession.
And beyond the decrepit labor market, the list of additional risk factors, seemingly endless already, is growing by the day and includes, among others:
Endless tariff uncertainty
A market bubble pop
Fed independence (or imminent lack thereof)
Rising debt everywhere
Rising inflation & painful stagflation
Private credit and shadow banking
Crypto mania
Consumers running out of gas
Tectonic shifts in the world order and global geopolitics
Wars on every continent
AI decimation
Climate chaos
Epically insane (and rising) income inequality
Immigration & Deportations
Not surprisingly, then, the probability we’d assign to a recession in 2026 is far higher than the present consensus estimate of 25%.
And as exhaustive and terrifying as the list above is, we’ve left off the list the insanity spewing out of Washington, OBBB, corruption, lawlessness, the decimation of democratic institutions, weaponizing the federal government, and the growing rumblings and increasingly mainstream attention being paid to rising authoritarianism and, literally, the risk of a civil war and/or a revolution of some sort. As we’ve said all year, whatever anyone’s view is regarding how bad things will get, especially heading into an election year, we’ll take the over.
Happy Holidays everyone!