The Political Economy of Trade Shocks: China and Local Labor Markets

The Theory Said Everyone Gains

For most of the twentieth century, the economics profession’s position on international trade was about as close to unanimous as economists ever get: trade is good. Not good for everyone in every way at every moment, but good in aggregate — meaning the gains to winners exceed the losses to losers, so that in principle the winners could compensate the losers and everyone would be better off. This conclusion rested on the theory of comparative advantage, first articulated by David Ricardo in 1817 and refined through successive generations of trade models. The logic was elegant and powerful: when countries specialize in what they produce most efficiently and trade with each other, total output increases. Consumers get cheaper goods. Resources flow to their most productive uses. The pie gets bigger.

The profession acknowledged, in theory, that trade creates losers as well as winners. The Stolper-Samuelson theorem, published in 1941, showed formally that trade liberalization would reduce the real incomes of workers whose skills were abundant in the trading partner — unskilled workers in rich countries, in the standard formulation. But this distributional concern was treated as a footnote, a second-order effect to be handled by domestic redistribution policy, not a reason to question the fundamental case for open trade. The canonical textbook treatment devoted a chapter to the gains from trade and a paragraph to adjustment costs. The implicit message was clear: trade is good for the country; if some workers lose out, that’s a problem for the welfare state, not for trade policy.

Then, in the 2000s and 2010s, a body of empirical research emerged that made the footnote impossible to ignore. The central contribution was a 2013 paper by David Autor, David Dorn, and Gordon Hanson — “The China Syndrome: Local Labor Market Effects of Import Competition in the United States” — that documented, with granular empirical precision, what happened to American communities when they were exposed to a sudden surge of Chinese imports. The findings were sobering enough to shift the center of gravity of an entire discipline.

The China Shock

The context was China’s explosive integration into the global trading system. Between 1991 and 2007, Chinese manufacturing exports to the United States increased roughly tenfold in real terms. China went from producing a small fraction of global manufactures to producing a quarter of them. This was driven by a combination of factors: Deng Xiaoping’s economic reforms, the creation of special economic zones, massive rural-to-urban migration within China, foreign direct investment (especially from Taiwan, Hong Kong, and Japan), and, crucially, China’s accession to the World Trade Organization in 2001, which locked in permanent most-favored-nation tariff rates and reduced the uncertainty that had constrained trade.

The scale and speed of China’s rise in manufacturing was historically unprecedented. Previous episodes of trade liberalization — the Kennedy Round, the Tokyo Round, NAFTA — had been gradual and involved countries at similar levels of development. The China shock was different: a very large, very low-wage country entering global markets very quickly, with a massive labor force and a deliberate strategy of export-led industrialization. The result was a supply shock to world manufacturing markets that was, as Autor, Dorn, and Hanson put it, “without historical precedent in terms of its sheer magnitude.”

What Autor, Dorn, and Hanson Found

The key innovation of Autor, Dorn, and Hanson (ADH) was methodological: they looked at the effects of trade not at the national level but at the level of local labor markets — “commuting zones,” roughly equivalent to metropolitan areas or regional labor markets. Because different American regions had different industrial structures, they were differentially exposed to Chinese import competition. A commuting zone dominated by furniture manufacturing or textile production faced intense Chinese competition; one dominated by aircraft manufacturing or health care did not.

By comparing more-exposed and less-exposed commuting zones over time, ADH could estimate the causal effect of Chinese import competition on local employment, wages, and a range of social outcomes. Their identification strategy used a clever instrument: they measured Chinese export growth to other high-income countries (a set of eight nations including Germany, Japan, and Australia) as a proxy for the supply-driven component of China’s export growth, which was plausibly exogenous to demand conditions in specific American regions.

The results were striking. Commuting zones more exposed to Chinese import competition experienced:

Larger declines in manufacturing employment. This was expected, but the magnitude was not. ADH estimated that Chinese import competition directly reduced U.S. manufacturing employment by about 2 to 2.4 million jobs between 1999 and 2011. This was a large fraction of the total manufacturing employment decline over the period, challenging the prior consensus that technology (automation, robots) was the primary driver and trade was secondary.

No offsetting employment gains in other sectors. Standard trade theory predicted that workers displaced from import-competing industries would move into expanding export industries or non-tradable services. This adjustment largely did not happen. Workers who lost manufacturing jobs did not smoothly transition to new sectors. Instead, many left the labor force entirely, moved to disability insurance, or remained unemployed for extended periods. The “adjustment” was downward — lower employment, lower wages, lower labor force participation.

Falling wages across the local economy. The effects were not confined to manufacturing workers. Because displaced workers competed for service-sector jobs and because local demand fell with local incomes, wages declined across the skill distribution in affected regions. The multiplier effects of lost manufacturing income rippled through local economies — fewer customers for restaurants, less property tax revenue for schools, reduced demand for local services.

Rising government transfer payments. Regions hit harder by Chinese imports saw significant increases in Social Security Disability Insurance claims, food stamp enrollment, unemployment insurance payments, and other transfer programs. The federal safety net partially cushioned the blow, but it also meant that the fiscal costs of the trade shock were borne nationally while the employment losses were concentrated locally.

The Adjustment That Didn’t Happen

The most consequential finding was about the failure of adjustment. Neoclassical trade theory doesn’t just predict that trade creates losers; it predicts that those losses are temporary. Workers displaced by imports are supposed to move — geographically, to regions with better opportunities, or occupationally, to growing sectors. Wages adjust, labor markets clear, and the economy settles into a new equilibrium in which everyone is employed, just in different jobs.

This did not happen after the China shock, and understanding why it didn’t is essential for understanding the broader debate.

First, geographic mobility in the United States had been declining for decades. Americans were moving less — between states, between metropolitan areas, even between neighborhoods — and the trend was especially pronounced among less-educated workers. Housing costs in thriving cities (New York, San Francisco, Boston) were prohibitively high for displaced manufacturing workers. The decline in mobility meant that adjustment through migration was far slower and less complete than theory assumed.

Second, the occupational transition from manufacturing to services was not smooth. Manufacturing jobs, even relatively low-skilled ones, had typically offered decent wages, benefits, and some measure of stability, partly because of unionization and partly because manufacturing firms had higher productivity and paid efficiency wages. The service-sector jobs available to displaced workers — retail, food service, home health care — typically paid less, offered fewer benefits, and provided less economic security. “Adjustment” in practice meant downward mobility.

Third, the shock was concentrated in regions that were already economically vulnerable: small and mid-size cities in the Midwest and Southeast whose economies depended on a narrow set of manufacturing industries. These communities had few alternative employers, limited educational institutions, thin labor markets, and declining populations even before the trade shock hit. The China shock did not land on a flat surface; it hit a landscape that was already tilted.

The Downstream Effects

A wave of subsequent research extended the ADH framework to examine the broader social consequences of the China shock. The findings were disturbing.

Health and mortality. Regions more exposed to import competition experienced higher rates of death from drug and alcohol poisoning, liver disease, and suicide — the “deaths of despair” documented by Anne Case and Angus Deaton. The connection was not mechanistic but ran through job loss, income decline, social disorganization, and the collapse of communities organized around stable employment. The opioid crisis, while driven by supply-side factors (aggressive marketing by pharmaceutical companies, the availability of fentanyl), was amplified by demand-side despair in communities where economic prospects had collapsed.

Political polarization. In a 2020 paper, Autor, Dorn, Hanson, and Kaveh Majlesi showed that commuting zones more exposed to Chinese import competition became more politically polarized. In predominantly white communities, the shift was toward Republican candidates, especially those with nationalist, anti-trade, and anti-immigration positions. In majority-minority communities, the shift was toward liberal Democrats. The aggregate effect was a hollowing out of the political center in trade-affected regions, consistent with the broader rise of populism in the United States and other countries that experienced large trade shocks.

Family structure and social outcomes. Research by Autor, Dorn, and Hanson found that trade-exposed regions experienced declines in marriage rates, increases in the share of children born to unmarried mothers, and falling male earnings relative to female earnings. The mechanism was the “marriageable men” hypothesis: when men’s economic prospects deteriorate, fewer of them are attractive marriage partners, and family formation patterns shift accordingly.

What the Research Changed

The China shock literature did not prove that trade is bad. The gains from cheaper consumer goods, the benefits to Chinese workers lifted out of poverty, and the efficiency gains from specialization are all real. What the research demonstrated, convincingly, was that the costs of trade adjustment were far larger, more concentrated, and more persistent than the profession had assumed — and that the standard prescription (“compensate the losers through redistribution”) had not been implemented in practice and might not be sufficient even in theory.

This shifted the debate in several ways.

The distributional question moved to the center. Before ADH, the mainstream position was that trade was good in aggregate and that distributional concerns, while legitimate, were separate from trade policy. After ADH, it became much harder to discuss trade without discussing distribution. The aggregate gains from trade could be small relative to the concentrated losses in affected communities, and the political sustainability of open trade depended on whether losers were actually compensated — which, in the United States, they largely were not.

The “adjustment” assumption was questioned. The finding that workers did not move, did not retrain, and did not find equivalent employment challenged a foundational assumption of trade theory. Economists began to incorporate more realistic labor market frictions — search costs, geographic immobility, skill specificity, hysteresis effects — into their trade models. The new models produced more nuanced welfare conclusions: trade was still beneficial in many cases, but the adjustment costs could be large enough to wipe out the aggregate gains for entire regions over policy-relevant time horizons.

Industrial policy became respectable again. The China shock research contributed to a broader rethinking of trade and industrial policy among mainstream economists. If trade can devastate regions and the market does not provide adequate adjustment, then there is a case for policies that manage the pace and composition of trade exposure, support affected communities, and maintain domestic productive capacity in sectors deemed strategically important. This intellectual shift preceded — and helped legitimize — the Biden administration’s embrace of industrial policy, including the CHIPS Act and the Inflation Reduction Act.

Policy Implications and Ongoing Debates

The China shock literature points toward several policy directions, though none is without controversy.

Trade adjustment assistance. The United States has had Trade Adjustment Assistance (TAA) programs since the 1960s, providing extended unemployment benefits, retraining funds, and relocation assistance to workers displaced by trade. But TAA has always been small relative to the scale of displacement, take-up rates have been low, and the evidence on retraining effectiveness is mixed. The ADH research suggests that TAA needs to be dramatically expanded and redesigned — more generous, longer-lasting, and focused on communities as well as individuals.

Place-based policy. The geographic concentration of trade losses suggests a role for place-based economic development policies: targeted infrastructure investment, business incentives, educational institutions, and public services in affected regions. The United States has historically been skeptical of place-based policy, preferring to rely on mobility (people moving to opportunity). The China shock research suggests that mobility is insufficient and that some places need active support.

Managed trade and industrial strategy. The most controversial implication is that trade itself should sometimes be managed — through tariffs, safeguards, or strategic industrial policy — to slow the pace of adjustment and maintain domestic productive capacity. Economists remain divided on this: traditional trade economists argue that tariffs are costly and distortionary; a growing camp, including former free-traders like Dani Rodrik, argues that some degree of trade management is necessary to sustain political support for an open economy.

The China-specific question. Much of the current debate centers on whether China is a special case — not just a low-wage competitor but a state-directed economy that uses subsidies, currency management, intellectual property acquisition, and industrial policy in ways that distort competition. The distinction matters because the policy response to a one-time adjustment shock (help workers transition) is different from the response to ongoing strategic competition (maintain industrial capacity). The shift from “China shock” to “strategic competition” in American policy discourse reflects this distinction, though it also reflects geopolitical considerations that go far beyond labor economics.

The Broader Lesson

The China shock literature is, at its core, a story about the gap between economic theory and economic reality. Trade theory, in its textbook form, demonstrated that trade was mutually beneficial — and it was right, at a high level of abstraction. But the theory assumed frictionless adjustment, perfect mobility, and lump-sum redistribution that never actually happened. When the theory met the facts — millions of workers in specific communities, over specific time periods, bearing concentrated losses with no meaningful compensation — the abstraction was revealed as inadequate.

This does not mean the theory was wrong. It means the theory was incomplete in ways that mattered enormously for policy and for people’s lives. The contribution of the China shock literature was to force the profession to confront that incompleteness, to take distributional consequences seriously, and to recognize that the political sustainability of an open trading system depends on whether its gains are broadly shared. That lesson, once learned, cannot be easily unlearned — and it remains at the center of debates about trade, industrial policy, and the future of globalization.