Labor Force Shrinks 400K, JOLTS Hiring Hits 2009 Levels — Payrolls Are Lying

The Payroll Number Is a Magic Trick

The March establishment survey printed +178,000 and the financial media declared the labor market resilient. Here's what they didn't tell you: the household survey — a completely separate measurement of actual employed people — recorded a 400,000-person collapse in labor force participation. JOLTS hiring just matched levels last seen in January and February 2009. The six-month payroll average sits at 15,000. And S&P Global's services PMI just contracted for the first time in three years.

The headline is the trick. The data underneath is the whole show.

Why It Matters

The labor market is the last structural pillar holding up the "soft landing" narrative — and it's cracking at the foundation. When hiring dries up, hours follow. When hours fall, incomes compress. When incomes compress against a rising energy cost base, consumer spending rolls over. That's not a theory. That's the sequence playing out in real time across multiple data sets right now.

The Big Picture

The Fed cut 50 basis points in September 2024 — a panic move that the Japanese carry trade reversal had been telegraphing for weeks before the data confirmed it. That same underlying trend, a weakening labor market dressed up by unreliable headline prints, never actually reversed. It just got papered over by revisions that arrived too late and too quietly to move markets. Now, layered on top of a structurally softening labor market, an energy price shock — five weeks old and already registering in PMI employment data — is accelerating the deterioration. The stock market is looking at +178K and pricing in recovery. The bond market and the real economy are telling a different story.

Key Details

• Labor force participation collapsed 400,000 in March — three of the last four months have shown no increase in the labor force. Workers aren't finding jobs; they're disappearing from the official count entirely. When you add them back in, the adjusted unemployment rate is 5.66%, not the headline 4.3%.

• JOLTS hiring rate matched January–February 2009 levels — the depths of the post-financial crisis labor market collapse. This isn't a blip. Hiring has "absolutely plummeted," and the historical sequence is unambiguous: job openings fall first, then hours, then headcount.

• Six-month payroll average: 15,000 — statistically indistinguishable from zero. The back-and-forth pattern — up one month, down the next — has persisted for ten months. February's already-negative preliminary print of -92,000 was revised further down to -133,000. Average the last two months together and you get roughly 25,000 jobs per month. That's contraction territory.

• S&P Global Services PMI revised to 49.8 for March — down from a preliminary 51.1, marking the first services contraction in three years. The accompanying press release cited "weakest rise in new work since April 2024" and explicitly flagged energy price surges from the Middle East conflict as the driver. Critically, S&P Global has historically been the optimistic outlier on services employment. When they start reporting job cuts, the signal is real.

• Hours worked fell for the third time in four months — the four-month change in the hours index is -0.1%. In a genuine recovery, employers extend hours before they hire. The fact that hours are contracting tells you employers don't believe the demand uptick is durable. They're working down backlogs, not building capacity.

What They Said

"The underlying currents continue to tell us the economy is slowing down. The labor market is weakening — and then all of a sudden you get this blockbuster number that the stock market is just going to eat up, saying everybody needs to buy the dip because the economy isn't going down."

— Steve Van Meter, macro analyst, EuroDollar University — has been tracking the JOLTS deterioration and carry trade warning signals since mid-2024, calls that have since been validated by subsequent data revisions.

"S&P Global, especially on the services side, tends to be the more optimistic outlier. So when S&P Global's employment index starts to contract, that tells you it's consistent with all the rest of it — overall suggesting that in this energy shock, businesses are afraid consumers are going to pull back."

— Jeff Snider, EuroDollar University — correctly identified the August 2024 JOLTS inflection as the signal the carry trade reversal was pricing in, roughly six months before the mainstream acknowledged the labor market was softening.

The Bottom Line

Watch the JOLTS hiring rate and the hours worked index — not the next payroll headline. The sequence from here is textbook: hiring already at 2009 levels, hours compressing, services employment contracting. The only variable is timing. The energy shock is the accelerant, and the tax refund stimulus is a one-time buffer that service providers themselves don't believe will hold — they cut hours and jobs in March despite the refund season tailwind.

Acid Take

The establishment survey is a lagging, heavily revised estimate that the market treats like a real-time GPS. It isn't. It's a rearview mirror