Labor's Share of U.S. Income Hits Record Low Amid AI Boom

Workers losing ground as productivity surges

American workers are capturing the smallest slice of the economic pie since the government began tracking the metric in 1947. Labor's share of national income has fallen to 52.8%, even as corporate profit margins reached a record 14.9% of GDP in the second quarter, according to EY-Parthenon chief economist Gregory Daco.

The divergence comes as Treasury Secretary Scott Bessent and Federal Reserve Chairman Kevin Warsh tout an AI-driven productivity boom that will make America richer. But the data raises a pointed question: richer for whom?

Economic output grew 1.7% in the second quarter while hours worked increased just 0.3%. Worker compensation rose 2.6%, but after accounting for oil-driven inflation, real wages were flat to slightly negative, Daco told Fortune.

Productivity gains predate AI

The productivity improvements driving record margins largely stem from a decade of traditional automation, post-pandemic cost discipline, and capital spending—not artificial intelligence. What AI has delivered so far is further concentration of gains among a handful of large firms.

"You tend to have greater concentration and more of a winner-takes-all type of environment when you have these technological advances," Daco said. Historical patterns from the railroad boom and the 1990s dot-com revolution show large, vertically integrated companies initially capture gains from technological shifts.

During the 1990s, productivity growth eventually spread throughout the economy through cheaper software, and wage growth followed. But AI's trajectory may differ significantly.

Capital intensity without domestic production

The AI buildout is uniquely capital intensive. Data center investment is projected to reach $31 trillion by 2050, according to PricewaterhouseCoopers—nearly matching current U.S. GDP. Construction and manufacturing are surging on data center demand; without it, the industry would be in recession, a Chicago manager reported in the Federal Reserve's Beige Book.

Yet much of this equipment isn't made domestically. Net imports of large computers used in AI servers hit a $450 billion annualized pace last month, up from roughly $50 billion annually through 2023, per Census data compiled by economist Joseph Politano. Because imported servers add to investment while subtracting equally as imports, their net contribution to GDP is zero.

This helps explain why capital spending booms while hiring remains weak, housing struggles under tight rates, and workers' income share keeps shrinking. Hundreds of billions in AI spending compete for capital in an economy where borrowing costs are rising, making mortgages expensive and suppressing homebuilding.

Why it matters

If productivity gains increasingly accrue to data center owners and shareholders rather than workers, the fiscal math for policymakers becomes complicated. The economy may grow while the tax base and political constituency associated with prosperity expand much more slowly. Daco said there's no floor for labor's share: "As long as you continue to see concentrated gains on the capital side, and within a certain number of firms," it could keep falling.

These details were first reported by Fortune.