THE US BUREAU of Labour Statistics will publish its July employment report on Friday morning. Wall Street expects about 83,000 jobs to have been added, a modest recovery from June's paltry 57,000. The ADP private-sector gauge, released on Wednesday, offered little comfort: 44,000 private jobs, down from 95,000 in June and below expectations.
The headline number will generate a brief flutter in bond yields and dollar futures. It will not, however, tell the reader much about where the American labour market is going. The more important story is already playing out, and it is not about how many people are being hired. It is about how many are disappearing from the available pool.
In June, the labour-force participation rate — the share of working-age Americans either employed or looking for work — fell by 0.3 percentage points to 61.5%, the lowest since March 2021, and barely above the level last seen in the 1970s, excluding the pandemic. Prime-age participation (those aged 25-54) dropped from 83.9% to 83.3%. The unemployment rate ticked down to 4.2% not because more people found jobs, but because fewer remained in the hunt. Household employment, the broader survey that counts how many people actually have a job, showed 507,000 fewer workers in June than in May.
Quick Backtesting Tool
That is the real tension. The unemployment rate looks stable because the denominator is shrinking. The headline masks a labour market in a "low-hire, low-fire" equilibrium, a phrase that has stuck with economists at The Conference Board and J.P. Morgan. Employers are neither expanding nor slashing, hovering in a state of paralysis. Workers are not quitting — the quits rate has been frozen at 2.0% for two months, the lowest sustained level since 2020 — because they do not trust that another job will be better. Analysts at Hiring Lab described the market as "calm on the surface but paddling hard just below the water."
Three forces are converging. First, tariff uncertainty, which has been building since early 2025, has frozen hiring plans in sectors that depend on imported intermediate goods. As Mitchell Barnes at The Conference Board put it, the full tariff impacts were expected to roll through in the first half of 2026, and the data suggest they have done so. Second, AI adoption is decoupling growth from employment in ways that have no analogue in recent cycles. J.P. Morgan's economists note that 25% of announced job cuts in March cited AI as the reason; 64% of American workers in a Pew Research poll expect AI will lead to fewer opportunities for them. Office-worker ranks have seen almost no growth. Third, tighter immigration policy has reduced the inflow of new workers, constraining the labour supply from below.
The JOLTS report for June, published on August 4th, adds texture. Job openings slipped to 7.36 million from 7.54 million. Hiring held at 5.35 million; layoffs at 1.77 million. The openings-to-hires spread of roughly 2 million suggests employers are posting positions without committing to hire, a sign of caution rather than demand. Real earnings growth turned negative in the second quarter, according to Hiring Lab. The labour market is not overheating. It is not, at present, collapsing. But the margin for error is thinning.
This is where Federal Reserve policy becomes the binding constraint. Kevin Warsh, who took over as chair in mid-May, has made clear that price stability outranks maximum employment. He has removed the Fed statement's explicit reference to "maximum employment", expressed scepticism towards the idea that inflation and jobs must be traded off, and announced task forces — including one on productivity and jobs led by Marc Andreessen and Charles I. Jones of Stanford — to reassess how the central bank measures labour-market conditions in an AI-disrupted economy.
Mr Warsh's pre-tenure case was that AI-driven productivity gains would allow growth without inflation, opening the door to rate cuts. The data have not cooperated. Inflation remains above the 2% target for the sixth consecutive year, buoyed by energy shocks from the Middle East and tariffs. The IMF now expects inflation to return to target only in late 2027. Fed Governor Christopher Waller says he can no longer rule out rate hikes.
The structural dilemma is not whether the labour market can support tighter policy. It is whether the shrinking participation rate can be reversed without an accommodation that feeds inflation. A rising unemployment rate would be the Fed's conventional signal that the labour market has weakened enough to ease. But if unemployment rises because workers are dropping out rather than being laid off, the signal is noisy. The Sahm rule — a simple recession indicator based on unemployment spikes — has not been breached. Initial jobless claims remain at two-year lows. The layoff rate is steady at 1.1%. Yet employment is up just 0.3% year-over-year, and the total number of employed people has fallen by roughly 833,000 in 2026.
The July report will be scrutinised for clues about whether June was an outlier or a trend. Economists at Wells Fargo expect something closer to the first-half average of roughly 95,000. Jeffrey Roach of LPL Financial forecasts only 75,000 with the unemployment rate edging up to 4.3%. Vanguard economists are even more cautious, projecting just 18,000.
These forecasts tell the reader less about July than about how uncertain the baseline has become. The revisions matter as much as the headline: April's initial figure of 179,000 was cut to 148,000, and May's 172,000 was revised down to 129,000. The Bureau of Labour Statistics routinely adjusts its numbers, but a pattern of downward revisions suggests earlier prints overstated the recovery.
For investors, the relevant risk is not a sudden labour-market shock but a slow grind. A weaker report would lift Treasury yields downward and ease pressure on duration-sensitive equities. A stronger one would reinforce the case for the Fed holding, or even hiking, and keep the dollar firm. Neither outcome resolves the deeper problem. The labour market's stability is bought with a shrinking workforce and a frozen labour force, neither of which can last.
The better question for policymakers is not what the July number says but what the participation rate means for the dual mandate's future. If fewer Americans are working because of demography, immigration restrictions, or a loss of confidence in the economic outlook, monetary policy cannot fix it. Interest rates are a blunt instrument that affects aggregate demand, not structural supply. Mr Warsh's task forces may eventually recommend redefining how the Fed measures "maximum employment". That is a reasonable exercise. It is not, however, a substitute for fiscal and migration policy that addresses the root causes.
The July jobs report will tell people whether June was a blip. The participation rate tells them that the blip is becoming the pattern. That is a harder problem, and one that the Federal Reserve cannot solve on its own.
