Jai Kedia and Norbert J. Michel
Economists at the New York Fed have published a new paper on how the 2025 tariffs reached consumer prices, and it has drawn deserved attention. In the paper, Mary Amiti, Sebastian Heise, and David Weinstein trace tariffs from the border to the store shelf. That is no small task, as consumer price categories mix imported and domestic goods and don’t line up with tariff schedules.
Some of the findings confirm their earlier work: Foreign exporters absorbed little of the tariffs, and roughly 90 percent landed on US import prices. The bigger contribution is showing that tariffs also raised the prices of goods made in America. Domestic producers paid more for imported inputs, and some raised their markups as competing imports became more expensive.
All told, the authors estimate that about a quarter of the tariff increase shows up in consumer goods, with roughly a third of that coming from domestically produced goods. As they explain in an accompanying blog post, that domestic effect builds slowly and takes about a year to fully transmit. The evidence is valuable, though the authors are careful about what it can and cannot tell us about aggregate inflation.
Properly separating the tariffs’ effects from everything else would mean estimating an open-economy macroeconomic model on data that include the tariff period and then decomposing inflation and output into their corresponding sources. (We’ve previously used such an approach to show that inflation was mostly sourced to factors outside of monetary policy.)
One problem, though, is that models are estimated on quarterly data, and the tariffs have been in place for only a handful of quarters, too few to properly identify how they have moved through the economy. Until more data are available, we’re left with simpler tools such as the ordinary least squares/instrumental variables methods the New York Fed authors employ in their study. And that leads to another problem because those tools always involve contestable choices. But it’s still worth using those tools because they give timely answers to questions the structural evidence can’t yet settle.
Regardless, inflation has dominated the tariff debate, and that’s a mistake, even though most of the early evidence suggests that tariffs have pushed aggregate prices up. Fixating on inflation hands tariff defenders an easy out: If the inflation data look tame, then the tariffs must be harmless. But among other potential difficulties, the cost of a tariff can show up in prices, jobs, and output, or all three.
San Francisco Fed research from 2025, covering 150 years of US tariff policy, found that tariff hikes have historically raised unemployment and lowered inflation, acting more like a blow to demand than a supply shock. Of course, that paper studies a very different period in US economic history, so maybe the latest round of tariffs will act more like a supply shock. Either way, a tariff that doesn’t raise inflation isn’t necessarily doing less harm. In fact, a tariff that doesn’t raise inflation may do even more harm.
More broadly, tariff defenders keep making terrible economic arguments to justify trade restrictions, and virtually all the data have favored removing those restrictions.
For instance, tariff defenders promised that foreigners, not Americans, would foot the bill. White House trade adviser Peter Navarro even argued that foreign exporters would cut their prices to absorb the duties. The New York Fed evidence shows that nearly all the costs landed on American buyers. Neither Navarro nor any of the other tariff apologists ever considered economic forces such as market power or profit margins when declaring that Americans would not bear the cost of increasing domestic trade restrictions.
Then take manufacturing. Factory employment remains below where it stood when President Trump took office, and the recent upswing in output is concentrated in the industries the tariffs spared.
But even if manufacturing jobs had a resurgence, increasing tariffs still would have been bad policy because the long decline in factory employment was never a problem to fix to begin with.
Growth economists call the underlying process that America experienced during the post–World War II era structural transformation. As economies grow richer, activity shifts from agriculture to manufacturing and then to services. The US has gone through this kind of structural transformation before, when farm work gave way to factory work. The same pattern shows up across countries: Manufacturing’s share of employment rises as a country develops and then falls as it grows wealthier. America’s shift toward services is what successful growth looks like.
For the new tariffs to have revived factory employment, they would have had to do so by pulling workers and capital from industries where the US is relatively more efficient. That shift toward services reflects millions of individuals’ decisions, and reversing those decisions by fiat forces people to do what government officials want instead. It is little more than a type of central planning. This government-directed reversal effectively taxes the inputs fueling the next transformation—whether that be the AI build-out or some other industry we are yet to uncover—to prop up activities the current ruling class favors.
The trade deficit tells the same story of failure: Tariffs were supposed to shrink the trade deficit, but it has risen and is basically back to pre-tariff levels. Still, as with supporters’ other favorite metrics, even a smaller deficit would not have vindicated the tariffs. As a measure of welfare, the trade deficit is meaningless. It largely mirrors the gap between what Americans save and what they invest, so a deficit often means foreigners see the US as a good place to put their money. Moreover, deficits tend to shrink in recessions, when Americans spend less and aggregate consumption falls.
More than a year and a half in, the tariffs have raised prices on many consumer goods, taxed American manufacturers’ own inputs, left the country with fewer factory jobs, and failed to shrink the trade deficit. Americans are paying real costs for a policy that has failed even when measured on the tariff defenders’ preferred metrics.
But the strongest case against tariffs doesn’t rest on these data. Tariffs would be bad policy even if they had hit every one of their stated targets because those targets are not measures of economic welfare. That they have missed those targets anyway only makes the verdict against tariffs even harsher.
Congress, which holds the constitutional power to set tariffs, should stop grading them on inflation prints, factory payrolls, and the trade balance. There is no world where increasing taxes, restricting trade, and pushing people into jobs that wouldn’t otherwise exist will increase Americans’ living standards. It’s unsurprising that real income has decreased over the year. It is past time for Congress to reestablish its authority over taxing international trade and reverse the Trump administration’s destructive policies.