Donald Trump’s tariff experiment is beginning to look less like a temporary negotiating tactic and more like the foundation of a new American trade system. On 31 July, Trump signed an executive order setting revised tariff rates for dozens of trading partners ahead of an August deadline, following months of negotiations triggered by April’s “Liberation Day” announcement. The emerging structure is uneven rather than universally protectionist: countries that negotiated agreements or concessions have received lower rates than those that did not, while a general tariff floor remains and China continues to be treated through a separate strategic confrontation. The European Union has agreed to a 15% tariff framework on most exports to the United States, Japan has accepted a similar 15% rate, Britain previously negotiated a more limited arrangement, while India faces a 25% tariff alongside Trump’s criticism of New Delhi’s trade barriers and purchases of Russian energy. Other countries face substantially higher rates. The important development is therefore no longer simply that America is raising tariffs. Trump is converting access to the world’s largest consumer market from something governed principally by common trade rules into something increasingly priced according to bilateral bargaining, strategic alignment and the negotiating leverage of individual states.
This is a fundamental change from the philosophy underlying the post-war trading system. The General Agreement on Tariffs and Trade and later the World Trade Organization attempted to reduce precisely this kind of discretionary bargaining by establishing common rules and the principle of most-favoured-nation treatment. The idea was not that every country needed identical tariffs or that governments should surrender all industrial policy. It was that trade becomes more productive when firms can predict the conditions under which goods cross borders without renegotiating political relationships constantly. A German manufacturer deciding whether to construct a factory could assume that American tariff treatment would broadly follow established rules rather than depend heavily on which government negotiated most successfully with the current president. Trump is deliberately weakening that model because he believes predictable access allowed foreign governments to take the American market for granted. His alternative is conditional access: if countries want favourable terms, they need to offer America something in return.
The European agreement demonstrates why the strategy can generate concessions. The EU faced the possibility of much higher American tariffs and therefore negotiated a 15% rate while also discussing large purchases of American energy and defence equipment and additional European investment in the United States. From Trump’s perspective, this is exactly what tariff leverage was supposed to accomplish. Europe receives continued access to American consumers at a cost higher than before but lower than the threatened alternative. America receives tariff revenue, investment commitments and greater demand for American exports. The agreement therefore resembles a bargaining surplus divided according to asymmetric outside options. The United States can tolerate losing some European trade more easily than many export-dependent European industries can tolerate losing access to America, giving Washington leverage.
But the economic meaning of “15%” should not be confused with America receiving 15% of European income. The tariff is paid initially by the importer in the United States. The ultimate burden is distributed according to market power. European exporters may reduce prices to preserve market share. American importing firms may accept smaller margins. Consumers may pay more. Domestic American competitors may raise prices because foreign products have become more expensive. The incidence therefore emerges through elasticities, not presidential declaration. A government can determine the statutory tax. Markets determine much of who actually bears it.
This distinction will become increasingly important as the administration argues that foreign countries are financing American government through tariffs. Customs revenue has indeed increased substantially as tariff rates rise, and that revenue can reduce the fiscal cost of other policies. Yet a meaningful portion ultimately comes from economic actors inside America. If an American company pays customs duties and passes half the cost to consumers, describing the entire tariff as money extracted from foreigners is misleading. Conversely, critics claiming consumers necessarily pay exactly 100% can also oversimplify because foreign firms sometimes absorb part of the cost. The correct question is empirical: how much of each tariff is absorbed by foreign producers, American firms and American consumers?
The variation across sectors will be enormous. A foreign producer selling a highly differentiated product with few substitutes may possess enough market power to pass much of the tariff forward. A producer competing in a commodity market may need to cut its export price significantly. An American importer able to switch rapidly towards another supplier can avoid much of the tariff. A company dependent on one specialised component cannot. This is why broad tariff rates conceal thousands of different microeconomic adjustments.
The Japan agreement provides another example of Trump’s strategy because Tokyo has strong reasons to protect access for its automobile industry. Japan can respond to tariffs by increasing investment inside the United States, something Japanese manufacturers have already done extensively for decades. This demonstrates one mechanism through which tariffs can achieve Trump’s industrial objective without eliminating foreign companies. The policy is not necessarily “American companies instead of Japanese companies”; it can become “Japanese companies producing inside America instead of exporting into America.” Nationality of capital and geography of production are separate variables.
This is potentially much more important than tariff revenue. A foreign-owned factory located in America employs American workers, purchases some American inputs and pays American taxes while transferring profits partly to foreign shareholders. From a national industrial-capacity perspective, the physical location of production can matter more than ownership. Trump’s strategy therefore can attract foreign direct investment through tariff avoidance. Companies effectively purchase access to the American market by building inside it.
The risk is that investment becomes distorted towards tariff avoidance rather than genuine productivity. A company might construct an American plant even though global production would be cheaper elsewhere simply because the tariff makes the alternative artificially expensive. Consumers then finance the inefficiency through higher prices. Whether this is worthwhile depends on the external benefits of domestic production: resilience, employment, technological spillovers or national security. Again, not every factory has equal strategic value.
A semiconductor fabrication plant may justify substantial policy support because advanced chips are critical to defence, artificial intelligence and industrial systems. Recreating low-value production domestically simply because imports produce a bilateral deficit is harder to justify. Economic nationalism therefore needs prioritisation or it gradually becomes autarky by accumulation.
India’s position is particularly interesting because the 25% tariff announced against Indian goods collides directly with America’s strategic objective of strengthening India as a counterweight to China. Trump has criticised India’s relatively high tariffs and trade barriers for years, but he has also attacked New Delhi’s continuing purchases of Russian oil and defence equipment. Trade policy is therefore being used simultaneously to pursue commercial and geopolitical objectives.
This creates an important question for India: how much economic alignment with America should be exchanged for strategic autonomy?
India benefits substantially from American demand, technology, investment and defence cooperation. The United States benefits from India’s scale, geography and role in balancing Chinese power. Neither side therefore approaches negotiations as a purely commercial relationship. A trade conflict that significantly damages India’s growth or pushes New Delhi towards deeper economic cooperation with China and Russia would undermine American strategic objectives. Conversely, India cannot assume its geopolitical importance means Washington will ignore protectionism or Russian ties indefinitely.
The optimal outcome should therefore be a broader bargain. India reduces selected tariffs and non-tariff barriers, expands purchases where economically sensible and improves market access. America reduces the threatened tariff burden and preserves India’s role in supply-chain diversification away from China. Both governments maintain enough flexibility that India does not become a formal American ally and Washington does not treat every Indian relationship with Russia as disloyalty.
This would recognise strategic autonomy as a negotiating constraint rather than an obstacle to partnership.
India’s purchases of discounted Russian oil demonstrate why this matters. After Western sanctions reduced European demand for Russian crude, India increased purchases substantially, refining some of that oil domestically. From New Delhi’s perspective, buying discounted energy lowers import costs for a country dependent heavily on foreign oil and helps contain domestic inflation. From Washington’s perspective, those purchases continue providing revenue to a Russian state financing its war against Ukraine.
Both interpretations contain truth.
The disagreement is distributional.
Who should bear the economic cost of isolating Russia?
European countries initially asked the world to reduce Russian energy dependence after having spent decades benefiting from relatively cheap Russian supply themselves. India, with much lower per-capita income, has less willingness to pay an energy premium for a European security objective unless the strategic benefit to India is clear.
This is the burden-sharing problem of sanctions.
Sanctions are most effective when many countries participate.
Each additional country has incentives to defect if discounted goods become available.
A government purchasing cheaper Russian oil receives a concentrated national benefit while the reduction in collective sanction effectiveness is distributed across the coalition.
This is a classic collective-action problem.
Secondary tariffs or sanctions attempt to solve it by increasing the cost of defection.
Trump’s threat towards India therefore follows a coherent coercive logic even if New Delhi strongly dislikes it.
The danger is that excessive coercion encourages countries to build mechanisms reducing American leverage in the future. India has already promoted greater use of local currencies in some trade, participates in BRICS and consistently resists dependence on one geopolitical bloc. American tariffs therefore reinforce an existing Indian argument: diversify economic relationships so no external power can impose intolerable costs.
Again, we reach the weaponisation-of-interdependence paradox. America can use market access as leverage because countries depend on the American market. Every use of that leverage gives countries another reason to reduce dependence.
The same is true for Europe. European leaders accepted the 15% framework partly because the alternative was worse, not because they suddenly prefer American tariffs. The EU will therefore continue pursuing trade agreements elsewhere, strengthening its internal market and discussing greater economic sovereignty. Trump may win immediate concessions while accelerating long-term European diversification.
This does not make the strategy irrational.
A country should use leverage sometimes or the leverage has little strategic value.
The relevant calculation is the discounted value of current concessions minus the future leverage lost as partners diversify.
That is much harder to measure than customs revenue.
It may become the central economic question of Trump’s foreign policy.
The global trading system is consequently moving towards managed fragmentation rather than complete deglobalisation. Trade will not disappear. Modern economies are too interconnected and the efficiency gains too large. Instead, the terms of trade increasingly depend on geopolitical relationships. Strategic sectors receive protection. Allies receive preferential access. Rivals face restrictions. Companies build redundant supply chains. Governments subsidise domestic capacity.
Globalisation becomes more expensive but potentially more resilient.
This is very different from the 1990s expectation that economic integration would gradually make geopolitics less important.
Geopolitics is now determining the architecture of integration itself.
The China relationship remains the largest example. Washington and Beijing have temporarily moderated some tariff escalation while negotiations continue, but the broader strategic decoupling in advanced technology remains. Semiconductors, artificial intelligence, telecommunications, critical minerals and batteries increasingly are treated as security sectors rather than ordinary commerce.
This creates a two-layer global economy. Low-security goods can continue moving relatively freely. High-security technologies become embedded inside strategic blocs.
The boundary between the layers will be politically contested because industries have incentives to define themselves as strategic when doing so produces subsidies or protection.
Steel is strategic.
Semiconductors are strategic.
Pharmaceuticals can be strategic.
Energy is strategic.
Food can be strategic.
Data is strategic.
Once everything becomes national security, ordinary protectionism can hide inside security language.
Governments therefore need a disciplined test.
Would disruption of this foreign supply during a major crisis create severe national-security or economic damage that markets are unlikely to insure against adequately?
If yes, intervention may be justified.
If not, the burden of proof should remain with protectionists.
Otherwise resilience becomes an argument for producing almost everything domestically, sacrificing enormous gains from specialisation.
The World Trade Organization is one obvious casualty of this shift. Its dispute-settlement system was already weakened before Trump’s second term, and unilateral tariffs based on national-security and emergency authorities increasingly bypass the assumption that trade restrictions should be negotiated through multilateral rules.
This creates a return towards power-based trade diplomacy.
Large markets possess more leverage.
Small states need coalitions.
The EU becomes valuable partly because no individual European country can bargain with America or China as effectively as the bloc can collectively.
ASEAN and other regional arrangements gain similar logic.
Ironically, American bilateralism can therefore strengthen foreign regional integration.
Countries pool bargaining power because Washington prefers negotiating from national asymmetry.
Again, power generates counter-power.
The inflation effects in America remain uncertain because several factors can absorb the tariff shock. Foreign firms can reduce margins. The dollar can move. Importers can switch suppliers. Domestic producers can increase output. Companies can run down inventories accumulated before tariffs.
These adjustments can delay visible consumer-price effects.
That does not mean tariffs are costless.
It means incidence can appear with lags.
This is politically important because Trump needs voters to experience the industrial benefits before they experience substantial price increases.
Factories take years.
Price adjustments can occur within months.
That asymmetry remains one of the largest risks to the strategy.
The Federal Reserve therefore will continue watching whether tariffs generate a one-time change in the price level or more persistent inflation. A one-time 5% increase in the price of imported goods is painful but not necessarily ongoing inflation if prices then stabilise. The danger arises if firms repeatedly increase prices, workers demand compensating wages and inflation expectations shift.
This is the difference between a relative-price shock and an inflation process.
Monetary policy should respond differently to each.
The uncertainty itself can weaken investment even before prices rise. Companies now need to forecast not only consumer demand and technology but political relationships between governments. A manufacturer deciding between Vietnam, Mexico, India and the United States needs assumptions about future American tariffs on all four locations.
This increases the value of political-risk analysis inside ordinary corporate strategy.
Trade lawyers, geopolitical consultants and supply-chain specialists become more important.
Those are real economic resources diverted towards navigating policy complexity rather than producing goods.
Economists call this a transaction cost.
A more fragmented global economy therefore can lose efficiency even without dramatic tariff revenue simply because firms spend more effort managing borders and regulatory uncertainty.
The strongest defence of Trump’s approach is that the previous system concealed its own costs. Communities losing manufacturing employment, governments becoming dependent on Chinese supply chains and countries allowing defence-industrial capacity to disappear were also paying costs. They simply did not appear as customs duties.
This is correct.
Efficiency and resilience are both economic values.
The policy mistake of the pre-2016 consensus may have been assigning too little value to resilience and distribution.
The policy mistake of the new consensus could be assigning too little value to efficiency.
The optimal system lies between hyper-globalisation and autarky.
Trump is pushing the equilibrium strongly towards the second direction because he believes the first went much too far.
The empirical question is where the new equilibrium settles.
The deals emerging this summer provide our first evidence.
America clearly possesses extraordinary bargaining leverage.
Countries facing high tariff threats have negotiated.
Foreign companies are considering or announcing US investments.
Governments are offering purchases and market-access concessions.
Those are real outcomes.
The costs are equally real.
Higher effective trade barriers.
Policy uncertainty.
Retaliation risk.
More expensive supply chains.
Potential consumer-price increases.
Strained alliances.
The correct judgement therefore cannot be made from the existence of deals alone.
A coercive negotiation always produces concessions if the stronger side can impose enough pain.
The question is whether the concessions are worth the cost of creating the coercive environment.
That is the metric I would use to judge Trump’s entire tariff project.
Not tariff revenue.
Not trade deficits by themselves.
Not the number of countries arriving in Washington.
Measure productive investment.
Productivity.
Real wages.
Consumer prices.
Supply-chain resilience.
Export performance.
And the durability of America’s economic alliances.
If those variables improve together, Trump will have changed the intellectual history of global trade.
Governments around the world will conclude that the free-trade consensus underestimated national bargaining power and strategic production.
If America gains factories but consumers become substantially poorer, allies diversify away and productivity weakens, the old case for predictable low trade barriers will return much stronger.
What already seems unlikely is a complete return to the world before Trump.
Biden preserved many of Trump’s first-term China tariffs.
Europe is developing its own industrial policy.
India protects strategic sectors.
China has never operated a genuinely laissez-faire trade model.
Economic security has entered mainstream policy everywhere.
Trump is simply pushing the logic furthest.
That means the most consequential legacy of his tariffs may survive even if a future president lowers many of the rates.
The assumption has changed.
Governments no longer begin by asking why trade should be restricted.
They increasingly begin by asking whether dependence itself is safe.
That is a very different world economy.
The tariff schedules announced this week are one expression of it.
The larger transformation is conceptual:
globalisation is no longer being organised principally around the cheapest place to produce something. It is increasingly being organised around the cheapest place governments are politically willing to depend on.
That difference will shape investment, alliances and development long after the current tariff negotiations are finished.
