The jobs market has gone quiet. That is the problem.

THE QUESTION on every economist's calendar is how many jobs the American economy added in August. The answer is, for now, unknown. The Bureau of Labour Statistics will publish its Employment Situation report on September 4th. Until then, the market must grapple with the latest data, which arrived on August 7th and covered July. In that month the economy lost 23,000 jobs. Economists had expected a gain of around 80,000. The miss was not merely a number. It was a sign that the labour market, for all the talk of soft landings, is losing its lift.

The headline payroll figure is the obvious starting point, but the real damage lay elsewhere. The report revised May's employment gain down from 129,000 to 63,000 and June's from 57,000 to 20,000 — a combined downward adjustment of 103,000 jobs. Taken together, the three-month moving average of hiring fell to just 20,000 per month, according to PNC's economics research team. That is barely above the statistical noise and far below the roughly 150,000 a month that would match long-run labour-force growth. The labour market is not collapsing. It is idling.

Which sectors bore the brunt? Education shed 50,000 jobs, much of it in local government, where summer seasonal effects and budget pressures coincided. Mr Kevin Hassett, the director of the National Economic Council, attributed some of the decline to the Department of Government Efficiency, the Trump administration's cost-cutting initiative, and to the end of the World Cup, which had propped up temporary hospitality hiring. Retail lost 19,000 positions and financial activities, 14,000 — the latter sector now 121,000 jobs below its peak in May 2025. Health care added 22,000. The pattern is one of selective weakness rather than broad distress. But the financial sector's sustained decline is worth attention. It suggests that higher interest rates and regulatory uncertainty are having a lasting effect, not a transient one.

The unemployment rate's behaviour makes the picture odder. It fell to 4.1%, continuing a streak of 58 consecutive months below 4.5%. That stability comes from the household survey, which counts people, not payrolls, and is less volatile. The two surveys can diverge, and when they do, the revision record suggests treating the payroll side with more suspicion. The sharp downward revisions to May and June mean the household survey's calm may be the more reliable signal. The labour market is not yet weak by the measure that matters most to households. But the employers' side of the story — the one that drives investment, hiring plans and capacity decisions — is quietly deteriorating.

Wage growth adds another layer. Average hourly earnings held at $37.62, up 3.2% year on year. That is below the Federal Reserve's implicit comfort zone for a labour market that is still supposed to be tight. Temporary layoffs rose by 153,000 to 921,000, a signal that firms are parking workers rather than making them redundant. It is the behaviour of an economy that does not know whether the slowdown is cyclical or structural and is hedging accordingly.

To be sure, a single bad month does not make a recession. July is historically a soft month for hiring, and the seasonal adjustments applied by the BLS can exaggerate or understate the trend. The three-month average of 20,000 would be reassuring if the prior months had not been revised so far down. It would also be more alarming if the unemployment rate were rising, but it is not. The contradiction between stable joblessness and vanishing payroll growth is the puzzle that September's report will need to resolve.

What does this mean for the Federal Reserve? The central bank has been navigating between inflation that has eased but not surrendered and growth that has slowed without yet stalling. A three-month average of 20,000 jobs pushes the balance toward caution. The risk is not that the Fed will be forced into emergency action. It is that the data will keep arriving in this contradictory mode — payrolls thin, unemployment steady, wages soft — leaving the Fed to guess whether the economy is coasting or braking. That is an uncomfortable position for an institution whose mandate depends on clear signals.

The broader lesson is structural. The American labour market has spent the past five years absorbing shocks — pandemic, supply-chain disruption, rate hikes, government restructuring — without breaking. Its resilience has become its own risk. The assumption that it will always absorb more is no longer justified. The financial sector's loss of 121,000 jobs since last year, the rise in temporary layoffs, the softness in retail: these are not random fluctuations. They are the marks of an economy adjusting to higher borrowing costs, tariff uncertainty and a fiscal posture that is less supportive than it once was.

The August report, when it comes, will not answer every question. But if the three-month average stays near where it is, or falls further, the implication will be clear. The labour market's job is to add positions faster than the population grows. At 20,000 a month, it is failing at that task. The soft landing has become something harder to see.

Better to acknowledge the weakness than to wait for the unemployment rate to prove it.