For roughly seventy years after World War II, the story of the world economy was one of falling trade barriers. Average tariffs came down decade after decade under the GATT and later the World Trade Organization, and cross-border supply chains grew dense enough that a single car or phone might cross borders dozens of times before it was sold. In 2025 that direction reversed. The United States imposed the broadest set of import tariffs in generations, dozens of trading partners retaliated, and by September 2025 the average effective U.S. tariff rate had climbed to about 17.4 percent — the highest since 1935 [source: Yale Budget Lab, 2025].
Then, on 20 February 2026, the U.S. Supreme Court did something few expected: it struck down the legal foundation of the largest tariffs, ruling 6–3 that the emergency-powers law the administration had used "does not authorize the President to impose tariffs" [source: Congress.gov CRS, 2026]. Overnight the trade war entered a strange new phase — lower duties, unanswered questions about $166 billion in possibly refundable payments, and a scramble for other legal tools. This is a good moment to step back and ask the plain questions underneath the headlines: what is a tariff, who actually pays it, and what does the evidence say it does to prices, supply chains, and jobs?
- What a tariff actually is (and isn't)
- How we got here: the 2025–2026 escalation
- The legal earthquake: the Supreme Court steps in
- Who actually pays?
- The price on the shelf: inflation and households
- Supply chains, growth, and jobs
- What to watch
What a tariff actually is (and isn't)
A tariff is a tax on an imported good, collected at the border. The crucial and often-misunderstood detail is who hands over the money: the importer of record — a domestic company bringing the goods in — pays the duty to its own government, not the foreign seller [source: World Trade Organization, 2025]. A U.S. retailer importing sneakers pays the U.S. Treasury when the container lands. What happens next — whether that retailer absorbs the cost, pushes it onto shoppers, or squeezes its supplier into cutting prices — is the whole economic question, and we return to it below.
It helps to separate a tariff from its cousins. A tariff is a broad tax on things coming in. That is different from an export control, which restricts what a country allows to leave (for example, limits on critical minerals) and is a distinct policy tool aimed at denying goods to rivals rather than taxing purchases. This article is about import tariffs as broad taxes and their economy-wide effects.
Tariffs are also old. The U.S. Smoot–Hawley Tariff Act of 1930 sharply raised duties and is widely associated, in the economic-history literature, with deepening the collapse of world trade during the Great Depression [source: U.S. Tariff Act of 1930]. The lesson drawn from that era shaped the postwar push to lower barriers. The 2025 turn is significant precisely because it runs against that seventy-year current.
How we got here: the 2025–2026 escalation
The escalation had a clear starting gun. On 2 April 2025 — branded "Liberation Day" — the United States announced a baseline 10 percent tariff on nearly all imports plus steeper "reciprocal" tariffs on dozens of countries, invoking the International Emergency Economic Powers Act (IEEPA) as its legal basis [source: Congress.gov CRS, 2025]. Within days the confrontation with China spiraled: the U.S. reciprocal rate on Chinese goods was raised in steps to an effective 145 percent, and China retaliated with 125 percent duties on American goods [source: Congress.gov CRS, 2025].
Such rates are not sustainable, and both sides knew it. A 12 May 2025 meeting in Geneva produced a truce that cut the reciprocal rates sharply for 90 days; the pause was extended in August, and after a leaders' meeting on 30 October 2025 the fentanyl-linked surcharge on China was trimmed to 10 percent and the lower reciprocal rate extended for a year [source: White House, 2025]. Alongside the drama, a separate and quieter track continued: tariffs under Section 232 of the Trade Expansion Act of 1962, justified on national-security grounds, pushed duties on steel and aluminum to 50 percent and on automobiles and parts to 25 percent [source: Miller Nash, 2026].
The pressure also produced deals. The United States reached framework agreements setting an all-inclusive 15 percent ceiling on most European Union goods [source: European Commission, 2025], a 15 percent baseline with Japan paired with large Japanese investment and purchase pledges [source: Congress.gov CRS, 2025], and a 10 percent floor with the United Kingdom [source: UK Parliament, 2025]. Supporters presented these as proof that tariffs are effective leverage. Whatever one makes of that claim, the cumulative effect on the tax at the border was unmistakable: the average effective U.S. tariff rate briefly touched about 27 percent in April 2025 — the highest since 1903 — before settling near 17.4 percent by September [source: Yale Budget Lab, 2025].
The legal earthquake: the Supreme Court steps in
The tariffs that drove the effective rate to nine-decade highs rested largely on IEEPA, a 1977 emergency-powers statute that had never before been used to impose tariffs. On 20 February 2026 the Supreme Court ruled, 6–3 in an opinion by Chief Justice Roberts, that it could not be: "IEEPA does not authorize the President to impose tariffs" [source: Congress.gov CRS, 2026]. The decision struck down the Liberation Day "reciprocal" tariffs and the trafficking-related tariffs, and all IEEPA duties terminated on 24 February 2026 [source: White & Case, 2026].
Importantly, the ruling did not end the trade war. Tariffs imposed under Section 232 (the steel, aluminum, copper, and auto duties) and Section 301 (the China-specific measures) rest on different statutes and remain in force [source: Miller & Chevalier, 2026]. The administration also moved quickly to replace what it had lost, announcing a 10 percent global tariff under Section 122 — a tool that is capped at 150 days without congressional action — along with new Section 301 investigations [source: Congress.gov CRS, 2026]. Whether that pivot proves durable is, as of this writing, unsettled.
The decision left an enormous loose end: money. The government had collected roughly $133.5 billion in IEEPA tariff payments by mid-December 2025, and analysts estimate that on the order of $166 billion could ultimately be refundable — the Court did not resolve how or whether refunds happen [source: Tax Foundation, 2026]. Even after the ruling, the average effective tariff rate settled near 9.1 percent, still the highest since 1946 [source: Yale Budget Lab, 2026]. The era of low tariffs did not simply resume; it downshifted.
Who actually pays?
Here is the most contested question in the whole debate, and the one where perception and measurement diverge most sharply. The political claim, repeated often, is that foreign countries pay the tariffs. Economic theory offers a partial defense of that idea: a very large importer can, in principle, force foreign exporters to cut their prices to keep market share, shifting some of the burden abroad. Economists call this the terms-of-trade or "optimal tariff" channel, and it is a real possibility, not a myth.
The question is how large that channel actually is — and that is something we can measure rather than assume. During the 2018–19 trade war, multiple studies found near-complete pass-through of tariffs into U.S. import prices, meaning American buyers bore essentially the entire cost while foreign exporters did not cut their prices [source: Amiti Redding Weinstein, 2019]. A widely cited synthesis in the Journal of Economic Perspectives reached the same conclusion [source: Fajgelbaum Khandelwal, 2022]. The 2025 tariffs tell a similar story with fresher data: economists at the Federal Reserve Bank of New York estimate that U.S. importers bore about 94 percent of the cost from January to August 2025, with foreign exporters absorbing only about 6 percent — a share that eased slightly to roughly 86 percent by November as some suppliers began trimming prices [source: Federal Reserve Bank of New York, 2026].
So the honest synthesis is this: the terms-of-trade channel exists but is small. Theory allowed for foreigners to shoulder a meaningful part of the bill; the measured evidence from two separate episodes is that Americans paid the large majority — on the order of 90 percent. That is the difference between a claim and a verified finding, and it is worth holding onto.
The price on the shelf: inflation and households
If importers pay the duty, the next question is whether they pass it on. Businesses tend to protect their margins, so they generally do pass tariff costs through — but with a lag. Research on retail prices found that higher acquisition costs show up on store shelves roughly seven months later [source: Cavallo et al., 2021], which is one reason tariff-driven price increases can feel delayed and diffuse rather than sudden.
The measured effect on the overall price level is real but more modest than sticker-shock impressions suggest. The Yale Budget Lab estimates that the tariffs remaining after the Supreme Court ruling raise the U.S. price level by about 0.6 percent in the short run — roughly $800 per household on average in 2025 dollars [source: Yale Budget Lab, 2026]. Federal Reserve research finds that tariffs in place through late 2025 lifted core goods prices by around 3.1 percent and added about 0.8 percentage points to core inflation, accounting for much of the excess goods inflation over the period [source: Federal Reserve, 2026]. These are specific, bounded numbers — not the same thing as the vaguer feeling that "everything costs more."
Two further points matter for fairness. First, the burden is regressive: measured as a share of after-tax income, the cost to the lowest-income tenth of households is about three times the cost to the highest-income tenth, because lower-income families spend a larger share of income on imported goods [source: Yale Budget Lab, 2026]. Second, before the tariffs took effect, independent models had estimated even larger household costs under the most aggressive proposals — on the order of $1,200 a year for the Canada–Mexico–China tariffs and about $2,600 for a broad 20 percent tariff plus 60 percent on China [source: Peterson Institute for International Economics, 2025]. Those were forecasts; the realized figures came in somewhat lower, partly because deals softened the rates and partly because the Court intervened.
Supply chains, growth, and jobs
Tariffs also reshape behavior before they ever change a price. In early 2025, companies raced to import goods ahead of the new duties, and U.S. imports jumped roughly 41 percent at an annual rate in the first quarter. Because imports are subtracted in the arithmetic of gross domestic product, that surge helped push measured U.S. GDP down about 0.5 percent annualized in the first quarter — a statistical effect of front-loading, not a broad recession, and it rebounded as imports normalized [source: Federal Reserve, 2025].
The deeper question is whether tariffs achieve their central promise: reviving domestic manufacturing and its jobs. The best evidence comes from the 2018–19 experience, which economists at the Federal Reserve studied in detail. They found that moving an industry from low to high tariff exposure produced three offsetting effects: a modest 0.4 percent gain in jobs from import protection, a 2.0 percent loss from higher costs of imported inputs, and a 1.1 percent loss from foreign retaliation — a net decline of about 2.7 percent in manufacturing employment [source: Flaaen Pierce, 2020]. In other words, the protected industries gained a little, but the factories that depend on imported parts and on export markets lost more. This is a measured finding about one episode, not a prophecy — but it is the strongest evidence available, and it cuts against the simplest case for tariffs.
Zoom out, and the global picture in 2025–26 has been more resilient than the gloomiest early forecasts. The International Monetary Fund projects world growth of about 3.0 percent in 2025 and 3.1 percent in 2026, an upward revision that reflects the front-loading, the softer rates after deals, and easier financial conditions [source: IMF World Economic Outlook, 2026]. World merchandise trade, which the WTO had feared might shrink, instead grew modestly in 2025 but is expected to weaken in 2026 — a sign that the drag was delayed rather than avoided [source: World Trade Organization, 2026]. The costs of a trade war, it turns out, tend to arrive slowly.
What to watch
The trade war is not over; it has changed shape. A few things will decide where it goes. The first is the legal aftermath: whether and how the roughly $166 billion in IEEPA duties gets refunded, whether the Section 122 stopgap survives its 150-day clock, and how far the new Section 301 investigations reach [source: Congress.gov CRS, 2026]. The second is retaliation — trade wars escalate in cycles, and each new tariff invites a response that can hit exporters far from the original dispute.
The third is the promise at the heart of the policy: reshoring. Proponents argue tariffs protect strategic industries, provide negotiating leverage, and raise revenue — all legitimate goals — while critics point to consumer costs, retaliation, and the thin employment payoff the evidence has found so far. Watching whether factory investment and manufacturing jobs actually rise, and at what price to households, will tell us which side had the better of the argument. For now, the fairest summary is the one the data supports: tariffs are taxes that Americans have mostly paid themselves, their costs land slowly and fall hardest on those with the least, and their benefits remain more promised than measured. Those are the numbers to keep watching, whichever way the politics turn.