One of the most firmly held beliefs in economics is that free trade is good for society. Although it’s long been known that globalization creates both winners and losers, we’ve learned only recently how concentrated and long lasting the losses from trade can be.

We used to think that the factory workers who lose their jobs because of trade (or other events) could easily move to new jobs in new regions, mitigating most of the consequences of job loss. But in truth, less-educated workers aren’t very mobile and this lack of mobility has made the adjustment to globalization painful.

In the early 1990s, China began an export boom the likes of which the world had never seen before. A consequence of the resulting trade shock was concentrated job loss in US manufacturing cities and towns (and boom times for economies capable of supplying the inputs that China needed to grow).

When China joined the global economy, its comparative advantage was focused on a surprisingly narrow set of goods. The regional and national economies that found themselves competing with China were exposed to the brunt of its export surge.

By the late 20th century, many US manufacturing industries were located in smaller towns and cities in the Midwest and Southeast. When China’s export wave was unleashed, it hit these communities like a tsunami.

Manufacturing is one of the few US industries that has allowed workers without a college degree to reach the middle class. When manufacturing jobs disappeared, so did well-paying jobs for the non-college educated.

The immediate consequence of the China trade shock was an increase in imports available to US consumers. In China’s comparative-advantage industries, the country grabbed a substantial share of the US market.

Greater import competition caused US manufacturing factories to close and many manufacturing workers to lose their jobs. Job loss was concentrated in the cities and towns that had long specialized in labor-intensive manufacturing industries.

In regions hit by the China trade shock, job loss in manufacturing was not offset by labor migration to other regions or job gains in other sectors. Overall employment rates declined, with the exodus from work being largest for men without a college education.

Because of the hit to incomes from trade-induced job loss, many households in impacted regions increased their use of government benefits. But the extra income from these transfers replaced only a small fraction of the lost income.

Dig into how the China Trade Shock affected specific regions.

The loss of manufacturing jobs, decline in employment rates, and increased reliance on government transfers have persisted for a full decade beyond the levelling off of the China trade shock. In the typical impacted region, recovery has been minimal.

Persistent joblessness precipitates social dislocation. Trade-impacted regions have seen declines in marriage and fertility rates, more single parenthood, and elevated incidence of children living in poverty and young men dying of substance abuse.

Greater economic despair has disrupted politics. The rise in import competition has increased support for politically conservative viewpoints and candidates, especially among non-Hispanic whites.

The lesson of the China trade shock isn’t that trade is bad for America — it’s that the concentrated and long lasting impacts of job loss require our attention.

As part of an escalating bilateral feud, the US has sharply increased tariffs on goods imported from China. There is no evidence that these barriers have brought jobs back to the communities that were hurt by the China trade shock.

Depressed employment rates in distressed regions have causes well beyond import competition from China. Restoring economic opportunity will require targeting the absence of work, not why job loss happens.