America's vanishing workforce is a policy choice, not a mystery

AMERICA'S WORKFORCE has shrunk by more than a million people in the past year. The headline is alarming but the real story is subtler. The decline is not a sudden exodus of discouraged workers or an early sign of recession. It is the product of two older forces—one demographic, one political—that have lately collided. The baby-boomer retirement wave has been building for decades. The second force is newer: immigration enforcement on a scale not seen in half a century. Together they have removed the supply of working-age people that American employers, consumers and GDP growth have long taken for granted.

The numbers require some care. In June 2026 the labour-force participation rate—the share of people aged 16 or older who are working or looking for work—fell to 61.5%, the lowest outside the pandemic years since 1976. In June alone, about 720,000 people left the labour force, as the prime-age participation rate for those aged 25-54 dropped 0.47 percentage points in one month. Yet part of the decline is statistical. The Bureau of Labour Statistics introduced an unusually large population-control revision in January 2026, which the Federal Reserve Bank of St Louis estimates accounts for 43% of the year-to-date drop. The revision corrected the share of the population aged 65+ upwards by 0.62 percentage points and lowered the share of prime-age adults by 0.51. Because older people work far less than younger ones, this mechanical shift in weighting lowered the aggregate rate even if nobody changed their behaviour. The implication is that the 2025 participation rate was overstated; the 2026 drop is partly the data catching up to a corrected demographic reality.

But the remaining decline is real. Even after stripping out the revision, participation has fallen because there are fewer prime-age people and because some of those who remain have stepped away. The Federal Reserve Bank of San Francisco notes that monthly labour-force growth among the foreign-born plunged from roughly 119,000 in 2023 to a decline of 6,000 in the first half of 2025. The Baker Institute at the University of Texas reports that more than 1.2m immigrants left the workforce between January and July 2025. Net migration, which has long been the primary engine of American labour-force growth, turned negative for the first time in at least 50 years: the Brookings Institution estimates it was between −295,000 and −10,000 in 2025, driven by both a sharp fall in arrivals and a surge in removals and voluntary departures.

This is not an incidental side-effect. It is what one expects when an administration that campaigned on and then executed large-scale immigration enforcement treats the labour market as an afterthought. Foreign-born workers participate in the labour force at a rate of 66.3%, compared with 61.6% for native-born Americans. About 70% of the foreign-born population is in the prime-age bracket, versus 63% of those born in the United States. Removing them from the labour supply is, arithmetically, a direct subtraction from the workforce.

The more disquieting part of the picture is how easily the unemployment rate has disguised the damage. In July 2026 the unemployment rate dipped to 4.1%, even as the economy shed 23,000 jobs. That stability is an illusion produced by the fact that job growth has slowed at roughly the same pace as labour supply. The Frisco Fed describes it as a joint slowdown: when both supply and demand shrink, the unemployment rate stays steady. The headline looks benign. The structure underneath is more fragile. Job growth has become concentrated in education and health services, while manufacturing continues to lose ground and construction has stalled. The steadiness of the unemployment rate does not mean the labour market is healthy. It means both sides of the ledger are shrinking.

To be sure, the administration's political logic is coherent. Mr Trump's supporters do not view immigration enforcement as a labour-market problem. They view it as an enforcement-of-law problem, an assertion of sovereignty, or a means of protecting native-born wages. There is evidence that construction, agriculture and food processing have suffered the most immediate shortages, but the promised spillover into higher wages for native-born workers has not materialised in any broad way. The Economic Policy Institute, a left-of-centre think-tank, estimates that mass deportation... would destroy roughly 5.9m jobs, including 2.6m held by American citizens, through complementary effects: fewer immigrant construction workers means less demand for native-born electricians and suppliers; fewer immigrant childcare workers means more native-born parents, often women, leave the workforce to care for their own children. Whether one accepts those numbers or not, the mechanism—immigrant and native labour are complements, not substitutes in many sectors—is sound.

The deeper problem is arithmetic. GDP growth is the product of hours worked and productivity. A shrinking labour force translates, one for one, into slower growth unless productivity picks up the slack. The Congressional Budget Office projects the US-born population aged 20-64—the core labour force—will shrink in every year of the next decade. The BLS projects the participation rate will fall another 1.5 percentage points by 2034, equivalent to 4.3m fewer workers than if the rate had held steady. The Trump administration's own budget office forecasts real GDP growth of 2.9%, well above the CBO's 1.8%. Bridging that gap would require productivity growth of roughly 2.3% per year, double the CBO's forecast and above any full business cycle since 2000. The OMB is asking American workers to produce more per hour than they have in a quarter of a century.

The better answer is not to pretend the trade-off away. Immigration is not merely a border issue; it is a macroeconomic input. The CBO estimates that halving net immigration reduces annual GDP growth by 0.2 percentage points; zero net immigration reduces it by 0.4. Those figures are small in isolation but they compound over a decade. The politics may make large-scale reform difficult, but the alternative is an economy that grows more slowly, runs hotter on inflationary pressure from supply shortages, and places the cost on consumers who pay for scarcer labour without seeing higher wages in return. America has chosen a tighter workforce. The question now is whether it can accept the price.

That is not a rhetorical one. A smaller labour supply is a smaller economy. The arithmetic does not care about the politics that produced it.