Why Trump's Tariffs Haven't Crashed the US Economy — Yet: The Delayed Damage Thesis

When Donald Trump returned to office in January 2025, most economists expected catastrophe. The new president had campaigned on sweeping tariffs, and within months he delivered: according to the Yale Budget Lab, the average effective tariff on US imports rose from 2% to 18% — the highest level since the 1930s, surpassing even the infamous Smoot-Hawley Act of 1930 in severity and disruptiveness. Yet by the end of 2025, consumer price inflation stood at 2.7% for the twelve months to November, the same rate as in late 2024, and unemployment had risen only modestly from 4.1% to 4.6%. Writing in The Guardian, Harvard economist Jeffrey Frankel — a former member of President Clinton's Council of Economic Advisers — argued that the damage has not been avoided but merely delayed, and that 2026 is when the bill comes due.

The numbers at a glance

• Average effective US tariff: rose from 2% to 18% in 2025 (Yale Budget Lab)
• CPI inflation (12 months to November 2025): 2.7% — unchanged from late 2024
• Unemployment: rose from 4.1% to 4.6%
• Government shutdown: 1 October to 12 November 2025, disrupting data collection
• Front-loading savings: up to $6.5 billion through May 2025 (Penn Wharton)
• Retail price increase on tariffed goods: approximately 5.4% since April
• Contribution to overall CPI: +0.7 percentage points
• Mexico/Canada exemption: granted 6 March under USMCA
• November rollback: tariffs reversed on grocery-price-sensitive goods

The scale of the tariff shock

The 2025 tariff escalation dwarfed anything in the post-war trading system. Trump imposed levies that violated international agreements and abandoned the Republican Party's decades-long professed commitment to free trade. The policy changes were frequent and often inexplicable, creating an environment of radical uncertainty that may have been as economically damaging as the tariffs themselves. Businesses could not plan, invest or price with confidence when the rules changed weekly.

Yet the macroeconomic data for 2025 told a surprisingly benign story. Inflation did not surge. Employment did not collapse. Growth probably slowed toward the end of the year, but the picture was obscured by a government shutdown from 1 October to 12 November that delayed data collection and left significant gaps in the statistical record. Frankel identifies four specific reasons why the worst predictions did not materialise — at least not yet.

Reason one: measurement problems

The federal government shutdown stretched from 1 October to 12 November 2025, disrupting the Bureau of Labor Statistics' ability to collect price data, particularly for October. Some CPI components are missing entirely. Even in November, there is reason to doubt the reported zero inflation in housing costs — if accurate, it would bias the overall CPI estimate downward. Meanwhile, the Bureau of Economic Analysis fell behind schedule on GDP releases, with third-quarter data postponed. In short, the official statistics may understate the true economic damage simply because they are incomplete.

Reason two: incomplete implementation

Many of the most damaging tariffs announced were never fully implemented. Trump postponed some repeatedly. On 14 November he rolled back others because they were driving up grocery prices — a politically sensitive category. Most significantly, on 6 March he exempted goods from Mexico and Canada from a 25% levy that had been in effect for just two days, recognising that the integrated North American auto industry would have been devastated. Goods from both countries now face no penalty if imported under the US-Mexico-Canada Agreement.

Frankel observes that this softening was predictable: "Trump regularly stakes out extreme negotiating positions, only to back down when the heat is on." Investors have internalised this pattern — the assumption that "Trump always chickens out", known colloquially as "Taco", has become a market taunt. But the tariffs that were implemented remain very high, and the pattern of escalation-then-retreat does not eliminate the damage already done to supply chains and business confidence.

Reason three: front-loading

As soon as Trump won the November 2024 election, companies began front-loading imports — rushing to accumulate stocks of goods before the anticipated tariffs took effect. The Penn Wharton Budget Model estimates that this strategy saved US importers as much as $6.5 billion (£4.8 billion) through May 2025, equivalent to 13.1% of the new tariff bill. Gold from Switzerland and weight-loss pharmaceuticals from Ireland were among the most aggressively stockpiled categories.

After tariffs entered into effect, most retailers did not immediately raise prices because they had not yet depleted their pre-tariff inventories. This is standard retail practice — firms sell through existing stock at old prices before adjusting. But it means the consumer price impact of tariffs was delayed by months, creating a false sense of normalcy in the inflation data.

Reason four: corporate absorption

The most important factor, Frankel argues, is that importers continued to absorb much of the cost increase even after depleting pre-tariff inventories. Research by Alberto Cavallo and co-authors, using real-time data from large US retailers, found that prices of goods subject to the new tariffs — both imported products and their domestic substitutes — rose by approximately 5.4% at the retail level since April 2025. This was enough to add roughly 0.7 percentage points to overall CPI inflation. But it represents only a small fraction of the costs that could potentially be passed through at current tariff levels.

Crucially, the prices importers pay did rise proportionately with tariffs, contradicting Trump's repeated claims that foreign exporters would bear the cost by lowering their prices. It is US companies that have been absorbing the duties, much as they typically do when the dollar depreciates. This partly reflects uncertainty about how long the tariffs will last — Trump might reverse course, or the Supreme Court might strike them down. That same uncertainty helps explain why affected companies have largely refrained from laying off workers.

Why 2026 is different

Frankel's central argument is that the benign 2025 data is misleading. Companies will not allow tariffs to erode profit margins indefinitely. Once pre-tariff inventories are exhausted, once the uncertainty resolves in the direction of permanence, and once shareholders begin demanding margin restoration, the pass-through to consumer prices will accelerate. Assuming the tariffs remain in place, the United States can expect more price increases and downward pressure on real incomes in 2026.

The implications for monetary policy are significant. The Federal Reserve, which held rates at 4.25–4.5% through most of 2025, faces a supply shock it cannot easily offset. If tariff pass-through pushes inflation above 3% while growth slows, the Fed will be trapped between its dual mandate objectives — unable to cut without validating inflation, unable to hike without deepening the slowdown. This is precisely the stagflationary scenario that economists feared in early 2025 but that appeared not to materialise. Frankel's analysis suggests it was not avoided — merely postponed.

Assessment

The 2025 tariff experience reveals several structural truths about the US economy. First, the corporate sector has significant capacity to absorb cost shocks temporarily, drawing on margins, inventories and pricing power. Second, the political system constrains tariff policy in practice even when rhetoric suggests otherwise — the Mexico-Canada exemption and the November grocery rollback demonstrate that domestic pressure forces retreats. Third, statistical measurement during a government shutdown is unreliable, and policy conclusions drawn from incomplete data are premature.

But the most important lesson is about timing. Economic damage from trade policy operates with lags — typically six to eighteen months between implementation and full pass-through. The 2025 data captures only the early phase. The full reckoning, Frankel implies, belongs to 2026. Whether it arrives as a sharp inflationary spike or a gradual erosion of real incomes depends on how quickly corporations decide they can no longer subsidise the tariff burden from their own margins. Either way, the bill has been written. It has simply not yet been presented.

Historical parallels and differences

The closest historical parallel is the Smoot-Hawley Tariff Act of 1930, which raised the average US tariff on dutiable imports to approximately 20% and is widely credited with deepening the Great Depression by provoking retaliatory measures from trading partners. The 2025 escalation reached a comparable level — 18% on all imports, not merely dutiable ones — but the global response was markedly different. Rather than blanket retaliation, the European Union and China pursued targeted countermeasures while continuing negotiations. The US-Mexico-Canada Agreement provided a structural exemption that did not exist in 1930, and the integrated North American supply chain proved too costly to disrupt entirely.

Another critical difference is the role of services. In 1930 the US economy was predominantly industrial; today services account for approximately 80% of GDP and are largely unaffected by goods tariffs. This structural shift provides a natural buffer that did not exist in the interwar period. However, it also means that when goods tariffs do bite — through higher input costs for manufacturers and retailers — the effects are concentrated in specific sectors rather than dispersed across the whole economy, making them politically visible even when macroeconomically modest.

The Supreme Court variable

One factor that could alter the 2026 outlook entirely is the pending Supreme Court challenge to the legal authority underpinning many of the 2025 tariffs. If the Court rules that the president exceeded his statutory authority under the International Emergency Economic Powers Act or Section 232 of the Trade Expansion Act, the tariffs could be struck down retroactively. This possibility — explicitly acknowledged by Frankel as a reason companies hesitated to pass through costs — introduces genuine uncertainty into any forecast. A ruling against the administration would remove the tariff burden, validate corporate patience, and likely produce a deflationary impulse as accumulated cost savings are passed to consumers. A ruling in favour would remove the last hope of reversal and accelerate the pass-through that Frankel predicts for 2026.

Sectoral impacts and the uneven geography of tariffs

The tariff burden in 2025 was not evenly distributed across the US economy. Consumer electronics, apparel, furniture and automotive parts faced the steepest levies, with rates on Chinese goods reaching 60% or more in certain categories before exemptions and rollbacks. Domestic manufacturers that rely on imported inputs — steel, aluminium, semiconductors, chemical intermediates — found their production costs rising even as final goods producers faced competitive pressure from unsubstitutable imports. The agricultural sector, traditionally a Republican constituency, bore disproportionate costs through retaliatory measures on soybeans, pork and corn exports, prompting the administration to announce compensation packages reminiscent of the 2018-2019 trade war subsidies.

Geographically, the impact concentrated in coastal states with high import dependence — California, New York, New Jersey and Texas — and in manufacturing corridors that rely on cross-border supply chains, particularly Michigan, Ohio and the Southern auto belt. Inland states with service-dominated economies felt less direct impact, creating a political asymmetry: the constituencies most harmed by tariffs were not necessarily those most supportive of the policy. This asymmetry, Frankel implies, is part of why the political system produced exemptions and rollbacks rather than full implementation.

The labour market response was similarly uneven. Aggregate unemployment rose only 0.5 percentage points, but sectoral data suggest manufacturing employment stagnated while logistics and wholesale trade shed jobs as import volumes declined. Retail employment held up better than expected, partly because retailers were drawing down existing inventories rather than placing new orders — a temporary buffer that cannot last indefinitely. The Bureau of Labor Statistics data gap during the shutdown means the true sectoral picture for October and early November may never be fully reconstructed.

Policy implications for 2026

Frankel"s analysis carries clear policy implications. For the Federal Reserve, the delayed pass-through means that 2025"s benign inflation data should not be extrapolated forward. If tariff costs begin flowing through to consumer prices in the first half of 2026, the Fed will face a classic supply-shock dilemma: tolerate above-target inflation to support employment, or tighten into a slowing economy to anchor expectations. Neither option is attractive, and the Fed"s credibility depends on communicating clearly which it will choose before the data forces its hand.

For fiscal policy, the tariff revenue — estimated at over $200 billion annualised at the 18% average rate — creates a tempting but dangerous source of income. Using tariff revenue to fund tax cuts elsewhere, as some in the administration have proposed, would embed the trade distortion permanently and make future liberalisation politically impossible. For businesses, the lesson is that supply chain resilience has a cost, and that cost is now being socialised through higher prices rather than borne by individual firms. The companies that invested in diversification away from China in 2024-2025 are better positioned than those that waited, but even they face higher input costs than in the pre-tariff world.

The central question for 2026 is not whether the tariff damage will materialise — Frankel"s analysis makes clear that it will — but whether it arrives as a one-time level shift in prices or as a persistent drag on growth. A one-time shift would be painful but manageable; a persistent drag, compounded by retaliatory measures and supply chain fragmentation, would represent a structural reduction in US potential output that no monetary or fiscal policy can fully offset.