Free Trade and Manufacturing Jobs: The China Shock Evidence

Abstract

Economic theory predicts that free trade raises aggregate welfare even as it causes winners and losers. The "China shock" research of Autor, Dorn, and Hanson documented that import competition from China after 2001 caused large, concentrated, and persistent job losses in affected US manufacturing regions — losses that were not offset by gains elsewhere as standard trade theory predicted. This evidence has substantially changed how economists think about trade's distributional costs and the political economy of globalization.

Background

The case for free trade rests on comparative advantage: countries benefit from specializing in what they produce most efficiently and trading for the rest. This logic predicts that even if some workers in import-competing industries lose jobs, aggregate employment and welfare rise as resources flow to more productive uses. Trade adjustments were expected to be temporary and mitigated by worker mobility across sectors and regions.

The entry of China into the World Trade Organization in 2001 provided an unprecedented test of this theory. China's manufacturing exports to the United States grew from $100 billion in 2000 to over $500 billion by 2015. The scale and speed of this import surge allowed economists to study its effects with unusual precision.

The Evidence

The China shock caused large, persistent local employment losses. Autor, Dorn, and Hanson (2013, American Economic Review) used variation in regional exposure to Chinese import competition — based on each region's pre-existing mix of industries — to identify causal effects. Regions more exposed to Chinese import competition suffered much larger manufacturing employment losses, and these losses did not recover within the study period. Regions that lost manufacturing jobs did not see offsetting gains in other sectors: workers did not smoothly reallocate to service industries or other regions.

The adjustment process was far slower than models predicted. Standard trade models assumed displaced workers would relatively quickly find new jobs at similar wages. The evidence found the opposite: workers in import-affected industries and regions experienced persistently lower earnings, higher unemployment, higher disability claims, and higher mortality for over a decade. Autor, Dorn, and Hanson (2013) estimated the China shock eliminated approximately 2–2.4 million US manufacturing jobs. Acemoglu et al. (2016) extended this estimate to 2.4–2.8 million jobs considering downstream effects.

Geographic concentration magnified the harms. Manufacturing employment was concentrated in particular cities and regions — the furniture factories of North Carolina, the electronics factories of Silicon Valley, the textile mills of the Southeast. When these industries declined, entire local economies declined. Workers could not easily move to other regions (due to housing costs, social ties, and the costs of relocation), and service industry job growth was insufficient to absorb the displaced workers.

Social harms extended beyond direct employment. Autor and colleagues documented increased mortality in hard-hit regions. Pierce and Schott (2020, Review of Economics and Statistics) found that regions more exposed to the China shock saw larger increases in drug overdose deaths. Acemoglu and Restrepo (2020) found that regions exposed to both import competition and automation had especially severe long-term employment declines.

The political consequences were predictable in retrospect. Autor, Dorn, Hanson, and Majlesi (2020) found that voters in regions more exposed to the China shock shifted sharply toward Republican candidates and populist anti-trade politicians between 2002 and 2016. The concentrated economic disruption produced concentrated political mobilization — a pattern that the aggregate welfare framing of trade economics had not anticipated.

Aggregate economic welfare probably did still rise. Consumer prices for Chinese manufactured goods fell significantly, benefiting all consumers but especially those who bought more of those goods. The aggregate welfare calculation likely still favors trade expansion — but the benefits were diffuse (lower prices for everyone) while the costs were concentrated (job loss in specific communities). This distributional structure is the core political economy problem.

Counterarguments

The China shock was unusually large and rapid. China's entry into global trade was a singular event — the sudden integration of the world's largest population into the global manufacturing supply chain. Future trade liberalizations may not produce comparable disruptions.

Trade adjustment assistance and worker support programs were inadequate. Standard economic arguments for trade include provision of adjustment support for displaced workers. In practice, US trade adjustment assistance programs were underfunded and underutilized. The policy failure was not trade itself but the failure to compensate losers adequately.

Automation was occurring simultaneously. Disentangling the effects of import competition from domestic automation is difficult. Some researchers argue that a substantial share of manufacturing job losses attributed to trade were actually caused by labor-saving technology.

What We Can Conclude

The China shock research produced some of the most consequential empirical findings in recent economics, substantially revising how the profession thinks about trade's distributional effects. The costs of import competition fell on specific workers in specific places and were far more persistent than standard models predicted. These concentrated harms had measurable consequences for mortality, social pathology, and political behavior. The aggregate welfare case for free trade remains theoretically sound, but the China shock evidence demonstrates that the distributional consequences of rapid trade liberalization are large enough to require active policy responses — not assumptions of smooth adjustment.

References

• Autor, D., Dorn, D., & Hanson, G. H. (2013). The China syndrome: Local labor market effects of import competition in the United States. American Economic Review, 103(6), 2121–2168.
• Acemoglu, D., Autor, D., Dorn, D., Hanson, G. H., & Price, B. (2016). Import competition and the great US employment sag of the 2000s. Journal of Labor Economics, 34(S1), S141–S198.
• Pierce, J. R., & Schott, P. K. (2020). Trade liberalization and mortality: Evidence from US counties. American Economic Review: Insights, 2(1), 47–64.
• Autor, D., Dorn, D., Hanson, G., & Majlesi, K. (2020). Importing political polarization? The electoral consequences of rising trade exposure. American Economic Review, 110(10), 3139–3183.
• Acemoglu, D., & Restrepo, P. (2020). Robots and employment: Evidence from US commuting zones, 1990–2007. Journal of Political Economy, 128(6), 2188–2244.