Remarks to a Room That Already Knows the Numbers

Fifty-seven thousand. That's what the US economy managed in June — less than half the 115,000 economists penciled in, and it only looked that decent because May got dragged down to 129,000 on revision. Strip out the spin and you're looking at the weakest print in four months, arriving at the exact moment a Fed chairman with a taste for "family fights" is letting his committee flirt publicly with a September hike. Somebody in this room needs to explain how those two facts coexist, because I can't make it add up, and I've been staring at the household survey component all week — 507,000 fewer people reported working, participation down three-tenths to 61.5%, the lowest since March 2021.

Let's not pretend that's a rounding error. That's the labor force shrinking under its own weight.

Here's the part nobody wants to say out loud in a Warsh-era Fed communications regime that's allergic to saying anything out loud: leisure and hospitality lost 61,000 jobs in June. The sector that was supposed to get a lift from the summer calendar instead went backward. Meanwhile professional and business services eked out 36,000, healthcare added 22,000, and everything else just sat there. That's not a broadening expansion. That's an economy running on two sectors while the rest idles.

And yet — sixty-one percent. That's where CME FedWatch has September odds of a quarter-point hike sitting right now, after three regional presidents dissented in favor of doing it a meeting early. Three dissents. The most since 2016. From officials who looked at 4.2% May inflation and decided the labor market's business is not the FOMC's problem this quarter.

I want to be generous here. I want to believe there's a coherent story where you hike into a cooling labor market and it works out. Tell me what it is. Because from where I'm sitting, the only story that fits is the one where the chairman has decided credibility on inflation matters more than credibility on employment, and he's betting the second mandate can absorb the hit while he rebuilds the first.

Friday gives us the July print. Consensus sits at 85,000 — a "recovery" that would still be beneath the twelve-month trailing average from a year ago, dressed up as good news purely because June set the bar on the floor. If it comes in anywhere near consensus, watch how fast the same commentators who called June "noise" start calling August "confirmation." That's not analysis. That's narrative laundering, and the market falls for it every single cycle.

Now look at what's happening one floor up, in corporate earnings, because the credit-tightening story isn't confined to households. Meta posted a 91% collapse in free cash flow this week. Ninety-one percent. Expenses up 55%, Reality Labs bleeding $4.6 billion, and more than twenty analysts trimming targets in response. This is the company that was supposed to be the AI capex bellwether — proof that hyperscaler spending could keep compounding no matter what the cost of capital did. Instead we're getting a live demonstration of what happens when the cost of capital actually starts to bite: the 30-year Treasury just printed 5.23%, a nineteen-year high, and companies funding trillion-dollar buildouts on that curve are discovering the arithmetic doesn't forgive enthusiasm.

So here's my actual question for the room. If the labor market is softening, and corporate free cash flow at the biggest capex spenders is getting crushed by financing costs, and the Fed's own regional presidents are split three ways on what to do about it — who exactly is the September hike supposed to help?

Because it isn't the worker whose participation just got revised into a four-year low. It isn't the CFO at a hyperscaler watching debt service costs climb while Reality Labs torches four billion a quarter with no revenue to show for it. The only party that benefits from a September hike is the inflation print itself, in isolation, stripped of every other variable that actually determines whether people have jobs and whether companies can afford to keep building.

I'm not arguing the Fed should ignore 4.2% inflation. I've spent enough of my career watching central banks let inflation run to know how that story ends. But there's a difference between fighting inflation and pretending the labor side of the mandate doesn't exist because the chairman would rather not say anything about it either way. Warsh built his whole framework around "conditions, not guidance" — fine, then look at the conditions. The conditions are a shrinking labor force, a corporate capex cycle straining against its own funding costs, and a bond market already pricing in more pain than the data has earned.

Friday's number won't settle this. It'll just tell us which story the market decides to believe for four more weeks. My money's on the one where nobody in that room is actually looking at the same spreadsheet I am.