Who really paid for Trump's first tariffs

With big new tariffs at the centre of the political debate, research on the 2018–2019 trade war gives a clear answer on who pays for import duties. Studies by the economists Amiti, Redding and Weinstein found that the tariffs were passed through almost one hundred per cent to American buyers.

What was imposed

The first Trump administration raised import taxes in several waves. Safeguard duties on washing machines and solar panels came in January 2018, followed by tariffs of 25 per cent on steel and 10 per cent on aluminium from most countries. The largest round targeted China, where duties eventually covered most of what America bought, at rates of up to 25 per cent. China, the European Union, Canada and Mexico retaliated against American exports, with farm goods such as soybeans taking much of the blow.

That gave economists something rare: a large, sudden change in trade policy, applied to some goods and some countries but not others. By comparing the prices of tariffed products with similar untariffed ones, and the same goods from targeted and untargeted countries, researchers could trace who ended up bearing the cost.

Americans paid the bill

Foreign exporters did not cut their factory-gate prices to absorb the duties. US import prices rose by the full amount of the tariff. American importers, wholesalers and retailers then either swallowed the cost in thinner margins or handed it on to consumers. The notion that tariffs raise revenue at the expense of foreign rivals does not survive contact with the data.

The finding was not a one-off. A separate team, Fajgelbaum, Goldberg, Kennedy and Khandelwal, reached the same conclusion using detailed customs data, and estimated that the losses to American consumers and importers outweighed the gains to protected producers and the tariff revenue collected. Work by Cavallo, Gopinath, Neiman and Tang found near-complete pass-through at the border, but a smaller rise in shop prices, which meant that importers and retailers absorbed a good share of the cost in their margins rather than passing it all to shoppers.

Legal incidence and economic incidence

The mechanics are simple. An American importer of record pays the duty to US Customs when the goods arrive. Who bears it in the end depends on bargaining power. A foreign supplier with few other customers might cut its price to keep American business, in which case part of the tariff falls on the exporter. In practice, Chinese exporters had other markets, American buyers had few quick substitutes, and the dollar price of the goods at the dock barely moved. The tariff was added on top.

The washing-machine duties offered a vivid case. Research by Flaaen, Hortaçsu and Tintelnot found that washer prices rose after the tariff, and so did the prices of dryers, which were not tariffed. Retailers and manufacturers used the cover of the duty to raise the price of the matching appliance, so consumers paid more than the tariff itself implied.

Winners and losers

Some did gain. Domestic steelmakers raised prices and profits after the metal tariffs, and some American appliance makers expanded production. The costs, however, fell on a far larger group: manufacturers that use steel, aluminium and imported components, who employ many more workers than the industries being protected. Farmers suffered from retaliation and were compensated with tens of billions of dollars in federal aid, which meant taxpayers paid a second time for the same policy.

The strongest objection

The best counterargument concerns exchange rates. Standard trade theory holds that a broad tariff should strengthen the currency of the country imposing it, which lowers the price of imports in dollars and shifts some of the burden abroad. The renminbi did weaken during the trade war, which partly cushioned Chinese exporters. Supporters of new tariffs argue that a universal duty, applied to all trading partners at once, would produce a larger currency offset than the targeted duties of 2018.

The offset is real but partial, and the evidence from the last round is that it was far from complete. A weaker renminbi helps Chinese sellers keep their dollar prices steady; it does not reduce what American importers pay in duty. A universal tariff would also remove the alternative suppliers that let some importers dodge the China duties, raising costs further. Retaliation against American exporters would, as before, add a second channel of damage.

What is on the table now

The proposals in this campaign are far larger than the last round. Donald Trump has floated a tariff of 60 per cent on Chinese goods and a universal tariff of between 10 and 20 per cent on everything else. The 2018 duties covered a fraction of imports; a universal tariff would cover all of them, including consumer goods that the first trade war largely spared. If pass-through looks anything like last time, the price effect would be felt across the shopping basket rather than in a handful of categories.

Nor would Congress necessarily have a say. The duties of 2018 and 2019 were imposed under trade laws, Section 232 on national security and Section 301 on unfair practices, that delegate wide powers to the president, and the courts largely upheld them. A new administration could move quickly, which means the first price effects could arrive within months of an inauguration rather than after a long legislative fight.

The timing of the effect matters for the Fed. A one-off rise in import prices would lift inflation for a year or so before dropping out of the annual rate, and central banks usually look through such shocks. The risk is that a large enough rise, arriving before inflation has fully returned to target, lifts expectations and forces the Fed to keep rates higher than it otherwise would.

If new tariffs arrive, the evidence says American companies and shoppers will pay for them, through higher prices and thinner margins, soon after the duties take effect. A universal tariff on the scale now proposed would show up in the consumer price index within months of its introduction, and the Fed would be left choosing between ignoring it and slowing the rate cuts it has only just begun.