6 minEconomy
Workers’ share of U.S. income hits record low as AI investment boom accelerates
Labor’s share of U.S. income fell to 52.8%, the lowest since 1947, while corporate margins hit records. Economists warn the AI boom may deepen concentration of gains in capital owners rather than workers.
Workers’ share of American income has fallen to its lowest level since the government began tracking the figure in 1947, even before the anticipated artificial intelligence productivity boom has fully arrived. The labor share dropped to 52.8% of national income in the second quarter, while corporate profit margins reached a record 14.9% of gross domestic product, according to data cited by EY-Parthenon chief economist Gregory Daco.
The divergence comes as Treasury Secretary Scott Bessent and Federal Reserve Chairman Kevin Warsh have argued that AI-driven productivity gains will make the country richer, potentially even deflationary, and large enough to ease concerns about the $40 trillion national debt. Daco, however, cautioned that the productivity improvements behind the current economic picture largely predate the AI boom and have so far protected corporate margins rather than worker compensation.
“Productivity growth protects margins, not income,” Daco wrote in a note Thursday. Economic output grew 1.7% in the second quarter on just 0.3% more hours worked. Compensation rose 2.6%, but with oil-driven inflation over the spring and summer, real wages were flat to slightly contracting, Daco told Fortune in an interview.
Daco said he does not believe 50% represents a floor for labor’s share of income. “As long as you continue to see concentrated gains on the capital side, and within a certain number of firms,” labor’s share could keep falling, he said. The productivity gains behind the current numbers stem from a decade of automation, cost discipline as hiring pulled back after post-pandemic bloat, and capital spending, rather than AI itself.
What AI has delivered so far is further market concentration. “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 parallels support that pattern: during the railroad boom of the late 19th century and the dot-com revolution of the 1990s, large vertically integrated firms initially captured the gains while smaller companies faced persistent cost pressures, policy uncertainty, and higher interest rates.
In the 1990s, a handful of frontier technology companies front-loaded capital investment and reaped the capital gains, but productivity growth from cheaper software spread quickly through the economy, and wage growth eventually followed. There is no guarantee AI follows the same timetable, Daco noted, because the current boom is uniquely capital intensive.
Data center investment is projected to reach $31 trillion by 2050, nearly the size of current U.S. GDP, according to PricewaterhouseCoopers. Construction and manufacturing are currently booming because of data center development; without them, a Chicago manager said in the Federal Reserve’s Beige Book this week, the industry would be in recession.
Yet much of that equipment is not made in America. Net imports of large computers, the Census category for GPU servers, hit a $450 billion annualized pace last month, up from roughly $50 billion a year through 2023, according to Census data compiled by economist Joseph Politano. In GDP accounting, an imported server adds to investment and subtracts as an import in the same amount, so the net contribution to GDP is zero.
That helps explain an unusual feature of the AI economy so far: capital spending is booming, productivity is improving, and corporate margins are enormous, yet hiring is weak, housing struggles under tight rates, and the share of income going to workers keeps shrinking. “While U.S. investment is booming, growth in gross domestic product has been modest,” wrote Jon Hilsenrath, the former Wall Street Journal Fed reporter who now advises hedge funds at Serpa Pinto Advisory.
The situation raises a difficult question for Warsh and Bessent: whether to let the boom run or intervene. Growth is not the same as broadly distributed income. If every dollar of output increasingly accrues to data center owners or shareholders passively collecting checks, the fiscal math becomes complicated. The economy may be getting richer while the tax base and political constituency policymakers usually associate with a boom grow much more slowly.
The investment itself also carries costs and risks. Hundreds of billions of dollars of AI spending competes for capital in an economy where borrowing is getting more expensive. Higher long-term rates make mortgages costly and suppress homebuilding. None of this means the AI productivity boom will fail, Daco said, but it is not immediately obvious how it boosts labor’s share. “I don’t think there’s a floor,” he said.
