NextFin News - The most expensive trade experiment in American history has produced the clearest possible verdict: it did not do what it promised. Two years after President Donald Trump returned to the White House and pushed the average US tariff rate from 2.6% to 13% — the steepest protectionist shift since the 1930s — the US goods trade deficit closed 2025 at a record $1.24 trillion, manufacturing employment sits 75,000 below its second-inauguration level, and the one thing that moved decisively was the route ships take to reach American ports. The lesson the world has absorbed is not that tariffs are powerful. It is that they are evadable.
The arithmetic is unforgiving. The stated goals of the tariff campaign were three: shrink the trade deficit, revive American manufacturing, and make foreign exporters pay. On the first two, the official data deliver a clean miss. On the third, the invoice-level evidence is nearly as blunt. What did work — collecting revenue — now sits under a Supreme Court order to refund it. That is the paradox at the heart of the episode: the one objective the tariffs actually achieved is the one the White House talked about least, and even that gain may be reversed by the courts.
The Deficit That Would Not Shrink
The central failure of the tariff project is the simplest to document, because the Commerce Department's own numbers make it unavoidable. In 2025, with tariffs climbing through the year, the goods trade deficit widened 2.1% to a record $1,240.9 billion. The combined goods-and-services gap narrowed slightly to $901.5 billion, still the third-largest on record. Exports rose 6% and imports rose nearly 5%. The deficit did not contract; it relocated.
China's direct place in US trade collapsed. Goods imports from China fell 29.9% to $308.7 billion, and the bilateral goods deficit with Beijing dropped 31.8% to $202.7 billion. Stanford researchers tracking the shift put it in starker terms: China's share of US imports fell from about 21% in 2017 to just 8% by the fourth quarter of 2025, back to where it stood when China joined the World Trade Organization. But the deficit that was supposed to vanish as a result simply reappeared under other flags. Mexico displaced China as America's largest trading partner, with $580.6 billion in two-way goods trade in 2025, and the US goods deficit with Mexico widened to $196.9 billion from $172 billion a year earlier. Vietnam's surplus with the US more than tripled from under $40 billion in 2018 to a record $123.5 billion in 2024.
This is not a bug in the policy. It is the mechanism. A trade deficit is not a scorecard of unfairness; it is the mirror image of a savings shortfall. A country that consumes more than it produces, and runs a federal budget deficit of about 6% of GDP, must import the difference. Tariffs change the price of imports; they do not change national saving. When you tax one supplier, demand does not vanish; it migrates to the untaxed one, usually at a higher cost. The White House's own "Great Transshipment Scam" report, released in August 2026, estimated that roughly $67 billion of US-bound goods were routed through China-linked supply chains in Mexico, India and Vietnam in 2025, costing about $28 billion in tariff revenue.
The second-order effect is where the real cost hides. Importers did not simply absorb the tariff; they re-engineered their supply chains around it. Components now travel from China to Vietnam or Mexico for light assembly, pick up a new country of origin, and enter the US under a lower rate. The product is more expensive, the value chain is longer, and the tariff base is narrower. Global Trade Alert estimates that $150 billion of Chinese exports were redirected away from the US market in 2025, with ASEAN absorbing $65 billion to $70 billion and meaningful flows into Sub-Saharan Africa, Latin America and the GCC. China's total exports still grew 5.5% in 2025 even as shipments to the US fell 20%, and Beijing's overall trade surplus reached a record near $1.2 trillion. The lesson for Chinese exporters was not to shrink. It was to diversify.
There is also a monetary second-order effect that the tariff debate usually misses. The US deficit persists at record size in part because the dollar's role as the world's reserve currency — the "exorbitant privilege" that lets America finance large deficits cheaply in its own currency — means the savings shortfall can be funded without a currency crisis. Tariffs can change the flag on the container. They cannot change the identity that makes the container necessary.
Who Actually Paid
The administration's signature claim — that foreign countries pay the tariff — has been tested against invoice-level data, and it fails. The Federal Reserve Bank of New York, tracking import, producer and retail prices through November 2025, found that roughly 90% of the tariff burden fell on US firms and households, with only about 6% absorbed by foreign exporters. In the first eight months of 2025 the US share was even higher, at 94%.
"In sum, U.S. firms and consumers continue to bear the bulk of the economic burden of the high tariffs imposed in 2025,"the New York Fed authors wrote.
About a quarter of the tariff increase passed through to consumer prices within a year, according to the New York Fed's pass-through study — a partial figure that understates the total cost. The remainder showed up as squeezed importer margins, softer wages in exposed industries, and higher prices for domestic producers who faced less import competition and dearer inputs. That is the hidden inflation of protectionism: it does not always appear at the checkout, but it appears in the margin, and it appears in the savings rate of the household that pays it.
The distribution of the pain was not even. Retailers and consumer-electronics importers, with thin margins and highly elastic demand, absorbed more of the duty and passed less through. Producers of steel, aluminum and intermediate inputs, facing less foreign competition, had more room to raise prices — which is why the New York Fed found that roughly a third of the consumer-price effect came not from taxed imports but from domestic producers responding to weaker import competition. The tariff, in other words, functioned partly as a transfer from American consumers to American producers in protected sectors — a transfer that showed up nowhere in the federal budget.
The Manufacturing Promise That Did Not Materialize
The political promise was a manufacturing renaissance. The data show a sector that stopped falling and then went flat. Employment declined every month of 2025, recovered slightly in the first two months of 2026, and has stagnated since. The ISM manufacturing index spent ten consecutive months below the 50 threshold in 2025, contracting through the tariff rollout. Total manufacturing employment stood at 12.598 million in June 2026, down 75,000 since the second inauguration and 305,000 below the January 2023 peak of 12.9 million.
There was, however, a construction boom — and it is the exception that proves the rule. Manufacturing construction spending more than tripled from $76.2 billion a month in January 2021 to nearly $230 billion in early 2025, driven overwhelmingly by semiconductors and clean-energy plants. But that wave began under the CHIPS Act and the Inflation Reduction Act, not the tariffs. By November 2025 spending had peaked at a seasonally adjusted annual rate of $206.1 billion and slipped roughly 8% over the following four months. The factories being built were subsidized, not tariff-induced; and when the subsidy wave plateaued, construction followed it down.
The historical analog is not encouraging for the protectionist case. The Smoot-Hawley tariff of 1930 pushed the average US rate to nearly 20% and is widely associated with the collapse of global trade in the Great Depression, though modern economists debate how much of the downturn it caused. The 2018-19 trade war in Trump's first term produced a smaller version of the same pattern: China's share of US imports began falling years before 2025, and the goods deficit kept setting records. Two experiments, separated by a decade, delivered the same result. A policy that fails twice is not bad luck. It is a structural constraint.
The Structural Shift Beneath the Cycle
The right question is whether the tariff shock changed the structure of trade, or merely rode a cycle. The answer is both, and separating them matters — because getting it wrong flips the conclusion.
The cyclical leg will mean-revert. Tariff rates are policy variables that can be cut in a trade deal, and the Supreme Court's 6-3 ruling in February 2026 already forced much of the 2025 tariff architecture onto weaker legal ground. If Washington negotiates lower rates with Beijing, some direct China-US trade could return. The transshipment markup — the extra cost of routing goods through a third country — creates a standing incentive to restore direct links once the tariff wall lowers.
The structural leg will not revert on its own. The geography of supply chains has changed durably. The $150 billion in redirected Chinese exports represents only 0.6% of global goods trade — small systemically, decisive for the countries that caught it. Mexico, Vietnam and ASEAN have locked in market share; Chinese exporters have built distribution networks across the Global South that will outlast any single tariff line. China-plus-one sourcing, once a contingency plan, is now embedded capital. That is the deepest lesson the world learned from Trump: the American market is large enough to punish, but not large enough to isolate.
The legal fragility of the revenue model compounds the point. Customs duties collected by the Treasury jumped from $79 billion in calendar 2024 to $264 billion in 2025, the highest level since 1947 as a share of imports. In fiscal 2025 alone, the government took in $195 billion in duties, more than 250% of the prior year. But the Supreme Court's February 2026 decision held that the emergency-powers statute used to levy most of those duties could not be used that way, and refunds began accelerating in May 2026. A tariff regime that must be refunded is not a fiscal strategy; it is a loan from importers that the courts can call in.
The Strongest Defense — And Why It Still Overclaims
The most serious defense of the tariff era is that it was never really about the deficit. Its defenders argue the goal was strategic decoupling from China, and on that measure it worked: China's share of US imports fell sharply, critical supply chains moved to friendlier jurisdictions, and the threat itself extracted concessions. The White House has framed the period as "the largest reshoring wave in American history."
That case is stronger than the deficit argument, but it still overclaims. Decoupling from China was already underway before 2025, driven by pandemic lessons and the CHIPS Act. What the tariffs added was cost and legal risk without accelerating the underlying trend. And the "reshoring" that did occur was concentrated in subsidized semiconductor fabs — a technology-security project that a targeted industrial policy could have achieved without a 13% average tariff on everything else. The counter-thesis proves that tariffs can serve as a blunt instrument of geopolitical signaling. It does not prove they are an efficient tool of industrial policy.
There is a narrower version of the defense that survives: tariffs can be a bargaining chip. The threat of 25% duties brought Canada, Mexico and the European Union to the table, and a temporary truce with Beijing followed. But a chip that must be played repeatedly loses its value, and the concessions extracted have so far been procedural — purchase pledges and enforcement mechanisms — rather than structural changes to the savings imbalance that drives the deficit.
What Comes Next
The forward picture splits cleanly by horizon. In the short term, the legal overhang dominates. With the Supreme Court having voided the emergency-powers basis for most of the 2025 tariffs and refunds accelerating from May 2026, importers hold claims that could claw back a large share of the $264 billion collected in calendar 2025. Any tariff policy that depends on emergency powers is a policy the courts can unwind, and the administration's fallback — temporary measures under Section 122 of the 1974 Trade Act — carries its own caps and time limits.
Over the medium term, the deficit will do what savings imbalances dictate. As long as the federal budget deficit remains near 6% of GDP and the dollar stays strong, the goods deficit will stay near record territory regardless of the tariff schedule. The falsifying signal for that reading is specific: if the 12-month US goods deficit falls below $1.0 trillion while the fiscal deficit stays above 6% of GDP, the macro interpretation is wrong.
Over the long term, the structural rerouting is the durable legacy. The winners of the tariff era are not American factories. They are the intermediaries — in Mexico, Vietnam, Malaysia and beyond — who learned to route around the wall. The falsifying signal for the "tariffs are useless" verdict is equally specific: if US manufacturing employment exceeds the January 2023 peak of roughly 12.9 million within 18 months while tariffs remain elevated, the no-renaissance call fails. On current data, that would require a reversal of a three-year decline that has so far shown no such inflection.
The final lesson is the oldest in trade policy, relearned at trillion-dollar cost: a tariff is a tax on your own imports that someone else's supply chain learns to avoid. Donald Trump set out to prove American leverage. He proved its limits instead.
Explore more exclusive insights at nextfin.ai.
