The Tariff Trap: How Protectionism Could Cost the US While Reshaping World Trade

A sweeping tariff strategy aimed at protecting American industry could instead weaken US growth, reduce exports and redirect trade across multiple regions of the world. In the study "Towards a Trade War in 2025: Real Threats for the World Economy, False Promises for the US," published in The World Economy, Antoine Bouët, Leysa Maty Sall and Yu Zheng examine how a large-scale US tariff escalation would reshape production, trade flows, consumer prices and national incomes.

The study uses the MIRAGE-Power computable general equilibrium model to simulate a 2025 trade war in which US tariffs rise by 60 percentage points on imports from China and by 10 percentage points on goods from other partners, with Canada and Mexico exempted. Trading partners respond with equivalent tariff increases on US exports, allowing the model to trace both the initial shock and the adjustments that follow.

The authors estimate that such measures would directly affect more than one-fifth of global merchandise trade. By 2030, world GDP falls 0.5% below the baseline and global trade declines 3.4%, while both the United States and China lose 1.3% of GDP. The scale of the damage is significant, but the study's deeper insight lies in how trade is reorganised rather than simply reduced.

A Tariff Wall Does Not Stop at the Border

Tariffs are often presented as a direct intervention: make imports more expensive, give domestic producers greater room to compete and encourage production at home. The study shows why that sequence becomes much less predictable once international production networks, export markets, exchange rates, imported inputs and retaliatory measures enter the calculation.

Under the key scenario, US merchandise exports fall 22.9% and imports decline 17.5%. China also suffers heavily, with exports falling 8.9%. Higher protection does stimulate activity in selected US industrial sectors, but the gains coexist with weaker activity elsewhere as exporters lose overseas access and companies face more expensive imported inputs.

The household economy also absorbs part of the cost. The simulation produces a 1.2% increase in the US consumer price index alongside declines in skilled and unskilled real wages. Protection creates a distributional question that headline measures of domestic production can obscure: an industry may gain market share while consumers, exporters, workers in other sectors and firms dependent on imported components face higher costs.

Retaliation strengthens this effect. Without retaliation, the model estimates a US GDP loss of 0.7% and an export decline of 15.2%. Once trading partners respond, the GDP loss deepens to 1.3% and exports fall 22.9%. Retaliation does not consistently improve outcomes for the countries imposing it either, underscoring the difference between using tariffs as strategic leverage and treating them as an economically costless response.

The Bigger Shock Is Where Global Trade Goes Next

Chinese exports to the United States fall by more than 80% in the central simulation, while US exports to China decline 58%. Such numbers suggest something closer to a severe bilateral rupture than a conventional adjustment in tariff rates.

Still, declining US-China trade does not translate into a uniform retreat from globalisation. Chinese exporters redirect goods toward markets including France, Germany, Canada and Mexico, while the United States relies more heavily on its North American neighbours. Supply chains and market access are reorganised rather than neatly dismantled.

Canada and Mexico become striking examples of how tariff discrimination can create unintended beneficiaries. Because they are exempted from the central US tariff increase, the model estimates GDP gains of 1.3% for Canada and 6.6% for Mexico, while Mexican exports rise by more than a quarter. Their advantage comes not from a general expansion in world demand but from improved competitive access to the US market as rival suppliers become more expensive.

Their position reverses when the researchers include them in the tariff conflict. Canadian GDP then falls 1.8% and Mexican GDP 1.4%. The contrast exposes the economic value, and risk, attached to preferential access when a country depends heavily on a major neighbouring market.

Developing economies face similarly uneven consequences. India, Sub-Saharan Africa, parts of Latin America and several Asian economies register losses under different scenarios, while some economies gain from the displacement of Chinese exports or shifts in sourcing. The model therefore provides little support for a simple division between developed-country losers and emerging-market beneficiaries. Exposure to China, access to the US market, export concentration and position within global value chains shape the outcome.

Trade Diversion Can Export Political Pressure

The redirection of Chinese goods carries consequences beyond trade volumes. As access to the US market contracts, greater Chinese exports enter other destinations. The researchers estimate, for example, a substantial widening of Europe's trade imbalance with China under the central scenario.

Such a shift could create a second layer of protectionist pressure. European or Asian producers that were not the original targets of US measures could face stronger competition from Chinese suppliers seeking replacement markets. Governments could then encounter domestic demands for safeguards, anti-dumping measures or additional trade restrictions, allowing a bilateral confrontation to spread through policy responses elsewhere.

The study also examines non-tariff measures, which produce a more damaging outcome than tariffs alone. When regulatory and other trade barriers are added to the central scenario, world exports fall 4.8% and global GDP declines 0.8%. US exports contract 31.4%, while US GDP falls by almost 2%.

For policymakers in developing countries, the implication is particularly important. Securing a temporary advantage from trade diversion is not the same as building durable competitiveness. Economies seeking to attract redirected production will still need logistics capacity, reliable infrastructure, investment certainty, skilled labour and access to diversified export markets if relocation is to generate lasting development gains rather than a brief response to geopolitical disruption.

Businesses face a similar calculation. A supplier that becomes cheaper relative to a tariff-hit competitor today may lose that advantage after an exemption disappears, retaliation expands or a third country introduces its own restrictions. Supply-chain resilience increasingly depends on assessing policy exposure alongside wages, transport costs, market size and production efficiency.

Tariff Revenue Cannot Carry the Fiscal Burden

The paper also tests one of the most ambitious arguments for higher tariffs: using customs duties on a scale large enough to replace federal income-tax revenue. Its modelling produces a sharp constraint on how far such a strategy could go.

As tariffs rise, customs receipts initially increase, but imports eventually contract enough to erode the revenue base. The model estimates that a uniform 80% tariff would maximise US customs revenue at roughly $819 billion, far below the more than $2 trillion collected through federal income taxes in the comparison used by the researchers. An 80% tariff would simultaneously reduce world trade by 10.8%, world GDP by 1.6%, US imports by 41.5% and US GDP by 2.7%.

The fiscal result reinforces the broader logic running through the study. Tariffs are not taxes imposed on a static quantity of imports. Consumers change purchases, firms replace suppliers, exporters encounter retaliation, currencies adjust and global production shifts. Raising the tax rate therefore changes the economic base being taxed.

MIRAGE-Power cannot predict the exact path of a future trade confrontation. Its findings depend on assumptions about retaliation, substitution between suppliers, macroeconomic adjustment and the way tariff revenues circulate through the economy. Political negotiations, sector-specific exemptions, corporate adaptation and changing technology could produce outcomes outside the modelled range.

Its analytical contribution is nevertheless substantial. The research demonstrates that protectionism cannot be understood by examining protected factories in isolation from exporters, households, suppliers and third-country competitors. Tariffs change relative prices across an entire trading system, and those adjustments can undermine some of the economic objectives the policy was intended to achieve.

For governments considering industrial security, supply-chain resilience or strategic autonomy, the relevant question is what happens after every major trading partner, multinational company and competing exporter adjusts to the new rules. In the study's simulations, those reactions turn a border measure into a global restructuring of trade, with costs and opportunities distributed far beyond Washington and Beijing.