Trade Policy, Jobs, and the Political Economy of Protection

In January 2002, Maytag was still running its refrigerator plant in Galesburg, Illinois, employing about 1,600 people at wages that put a family solidly in the middle class. By 2004 the plant was gone, relocated to a lower-cost facility in Mexico, and Galesburg spent the next decade trying to recover a payroll it never fully replaced. The national economy, according to economists, came out ahead on net. It is very difficult to explain that to someone whose job went with the plant.

This is not a story about whether free trade is good or bad. The profession is about as united as it ever gets on that question: surveys of economists show overwhelming agreement that trade raises total national income.¹ The more interesting question is why that consensus coexists with durable, cross-partisan political pressure for protection, and what that tells us about the gap between how trade works in theory and how it lands in practice.

The asymmetry that explains everything

Start with the political logic, because once you see it, it explains almost everything.

The gains from free trade are real but diffuse. Lower prices on clothing, electronics, and groceries accumulate in small, unnoticed increments across hundreds of millions of households. No one sends you a receipt attributing three dollars off your TV to a trade agreement. The Library of Economics and Liberty entry on protectionism frames this as the foundational asymmetry: the beneficiaries of trade are a vast, unorganized majority who rarely know they are benefiting.²

The costs, by contrast, concentrate. When an industry loses to imports, the damage falls on a specific group of workers in a specific set of towns, all at once, and visibly. Those workers know exactly what happened, they know who else it happened to, and they have powerful incentives to organize, vote, and demand action. The industry facing import competition can fund a trade association; the millions of consumers who would each pay a few more dollars under a tariff will never form a counter-lobby, because no individual stake is large enough to justify it.

This is the political economy of protection in a sentence: a small group with a large per-person stake will reliably out-organize a large group with a small per-person stake, even when the large group's total stake is bigger. Protection wins not because it is sound economics but because it is good politics. The visible factory beats the invisible savings every time.

Why the standard story undersold the pain

For a long time, the economist's reply to displaced workers was reassuring: yes, your job moved, but the economy will create new jobs, you will retrain, and the country comes out ahead. The trouble is that the evidence proved far less comforting than the model.

The landmark research came from economists David Autor, David Dorn, and Gordon Hanson, whose study tracked the U.S. labor markets most exposed to the surge of Chinese imports after China joined the World Trade Organization in 2001.³ What they found contradicted the tidy reallocation story: in the hardest-hit communities, manufacturing job losses were deep and persistent, wages fell, labor-force participation dropped, and the damage lasted for a decade or more rather than healing quickly. Workers did not smoothly move into new industries. Many left the workforce permanently. Many towns never recovered their economic footing.

The aggregate gains from trade with China were real. So was a concentrated, durable, geographically specific wave of harm that the standard models had largely waved away. This matters because it changes the moral and political calculus. If displacement were truly temporary and easily cushioned, the case for unfettered trade would be close to unassailable. Because the displacement was often neither, the backlash that fuels protectionist politics is not simple economic illiteracy. It is a rational response to a real cost that the winning side's rhetoric consistently minimized.

The actual argument for free trade

Let us be precise about what the defensible case for free trade actually claims, because it is more limited than the headline that everyone wins.

The accurate claim, grounded in comparative advantage, is that the winners gain more than the losers lose: the total pie grows, so in principle the winners could compensate the losers and everyone would come out ahead. The Library of Economics and Liberty entry on free trade makes exactly this point: the efficiency case for trade is a case about net gains, which logically implies a compensation step.⁴

That qualifier, in principle, is doing enormous work. The case only translates from theory into shared prosperity if compensation actually happens: if some of the diffuse gains are recycled to the concentrated losers through retraining programs, income support, relocation assistance, or serious regional investment. When that step is skipped, the math still says the country is richer, but the people standing in the empty parking lot of a closed plant are not, and walking them through the national income statistics is both accurate and useless.

Where the compensation broke down

The United States has had a program built for exactly this purpose: Trade Adjustment Assistance (TAA), run by the Department of Labor, which offers retraining and extended benefits to workers certified as displaced by trade.⁵ The principle is sound: capture some of the gains, route them to the losers. The execution has been the chronic weak link.

Evaluations have repeatedly found that TAA is modest in scale and limited in effect. Many displaced workers never qualify. Retraining does not reliably lead to comparable-paying work. Average earnings in post-displacement jobs tend to fall well short of the wages that were lost. The compensation channel that the entire winners-compensate-losers argument depends on has been, in practice, chronically underbuilt.

The consequence shows up in voting patterns. Autor, Dorn, and Hanson's follow-up work, published as NBER Working Paper 22637, found that the regions most damaged by the China shock moved measurably toward more polarized and anti-trade political positions.⁶ The protectionist turn in American politics is not free-floating sentiment. It traces, in part, to specific places where the gains-from-trade promise was made and the compensation half of it was never delivered.

Two things that are both true

Two conclusions follow, and they only seem to conflict if you insist trade is entirely good or entirely bad.

First, be skeptical of protection sold as a national win. The World Trade Organization's case for open trade and the full weight of comparative-advantage analysis still hold: tariffs and quotas mostly tax your own consumers to benefit a concentrated few, and the country as a whole ends up poorer for it.⁷ The popularity of protection is a feature of its politics, not its economics.

Second, be equally skeptical of free-trade advocacy that stops at the aggregate gain and never addresses how that gain is distributed. The losses documented in the China shock research are real, were concentrated in identifiable communities, and were genuinely under-addressed. A trade policy that grows national income while abandoning the communities that paid for that growth is not delivering on its own theoretical promise. It is delivering half of one.

The productive frame is not free trade versus protection. It is free trade plus a serious, well-funded mechanism to share the gains: which is, notably, the version economists have always claimed to support and which has rarely been tried in full. Until that second half is taken as seriously as the first, the political appeal of protection will keep outrunning the economics, for reasons that are entirely understandable to anyone who has watched a factory town lose its factory.