What Happened to America's Manufacturing Jobs, And Who Benefited?

THE QUESTION

"Globalization destroyed the American middle class." You've heard it from the populist right and the progressive left alike. The U.S. lost 5.7 million manufacturing jobs between 2000 and 2010. Was that trade, automation, or something else? And who actually came out ahead?

WHAT WE KNOW

Both trade and automation played real roles, economists still argue over the split, and the honest answer isn't clean. MIT research from Autor, Dorn, and Hanson found rising Chinese import competition between 1990 and 2007 caused significant, concentrated job losses, especially in manufacturing and among workers without college degrees, with no comparable net job loss from automation over that same window. Separately, Upjohn Institute economist Susan Houseman challenges the popular "robots did it" story: outside the computer industry (whose stats get skewed by how agencies measure rapid tech improvement), productivity growth wasn't high enough to explain the job losses through automation alone. Other estimates put offshoring's direct share at around 13% of the manufacturing employment decline from 1993 to 2011.

There's a real rebound worth knowing: manufacturing employment stabilized around 2010 and added back close to a million jobs, modest, but the longest sustained increase since the 1960s.

WHAT'S MISSING

"It was mostly robots, not trade" is the politically comfortable version, it doesn't point fingers at specific trade deals or specific winners. The specific-winners question is what gets skipped. As import competition rose, labor's share of national income fell sharply, and that value didn't vanish. It shifted, disproportionately, toward corporate profits and toward consumers who benefited from cheaper imports, while the costs landed hard on specific manufacturing communities and non-college workers who saw few of the broader gains.

WHY IT MATTERS

This isn't just history, it's the foundation under today's tariff fight. If trade competition drove real job loss, tariffs are at least a plausible tool for addressing it. If it was mostly automation, tariffs mostly just raise prices without bringing jobs back. The evidence says it's not either/or, so neither "tariffs fix it" nor "tariffs can't possibly help" holds up cleanly.

THE BOTTOM LINE

Established: both import competition and automation drove manufacturing's decline, with real disagreement among credible economists on the exact split; costs concentrated in specific communities and non-college workers while gains flowed more broadly. Uncertain: the precise split, and therefore how much any single fix, tariffs included, can actually reverse.

SOURCES & FURTHER READING

•      Autor, Dorn, and Hanson (NBER), "Untangling Trade and Technology: Evidence from Local Labor Markets"

•      Susan Houseman (Upjohn Institute), "Understanding the Decline of U.S. Manufacturing Employment"

•      Federal Reserve Bank of New York, Liberty Street Economics, "The (Modest) Rebound in Manufacturing Jobs"

•      Public Citizen, "Reporters' Memo: Data Show Trade Had Greater Impact Than Automation on Manufacturing Job Loss"