The China Shock

China's admission to the World Trade Organization in December 2001 was accompanied by projections from mainstream economists that adjustment to Chinese competition would be manageable -- that workers displaced from import-competing industries would find equivalent employment in other sectors as had occurred, if imperfectly, in previous episodes of trade liberalization. These projections proved dramatically wrong, as a body of research by economists David Autor, David Dorn, and Gordon Hanson (2013) documented with precision that has since transformed economists' understanding of trade's distributional effects.

The "China shock" research showed that commuting zones -- local labor markets -- that were more exposed to Chinese import competition experienced larger declines in manufacturing employment, more persistent unemployment, greater uptake of disability insurance and other transfer programs, and lower wage growth over the subsequent decade. The adjustment process that trade theory predicted -- displaced workers moving to other sectors -- largely did not occur. Workers who lost manufacturing jobs either left the labor force, moved to lower-paid service employment, or subsisted on government transfers. Entire communities, particularly in the industrial Midwest, experienced long-term economic decline with no signs of the recovery that the standard model projected.

Autor, Dorn, and Hanson estimated that Chinese import competition reduced US manufacturing employment by approximately 2.4 million jobs between 1999 and 2011 -- a substantially larger effect than previous estimates based on less granular data. The broader employment effect, accounting for the failure of workers to find alternative employment, was larger still. The research helped explain the political geography of the 2016 election: the communities most exposed to the China shock swung most sharply toward Donald Trump.

We had always assumed the labor market adjustment would happen. The people would move, get retrained, find new jobs. What we found was that it did not happen, or happened slowly and incompletely, in ways that left lasting damage in exposed communities.David Autor, MIT, 2016

The China shock research did not demonstrate that trade liberalization was a mistake -- the aggregate gains were real, consumers benefited from lower prices, and export sectors expanded. It demonstrated that the adjustment costs were larger and more concentrated than mainstream economic models had projected, and that the compensation mechanisms that should have offset them -- trade adjustment assistance, regional development programs, labor market policy -- were inadequate to the scale of the disruption.

The policy implications are not a return to protectionism but a more honest accounting of trade's distributional consequences and more serious investment in the institutions that manage adjustment. Countries that have navigated trade liberalization with less collateral damage -- Germany, the Nordic countries -- have done so with stronger labor market institutions, more generous social insurance, and more active industrial policy than the United States has been willing to maintain.