The debate behind Tariffs Can Lose in Court and Still Win in the Market

Advocate A’s strongest empirical point is that tariffs impose real costs while they are in force; I do not dispute that. The weakest point is the leap from “tariffs cause pass-through and rerouting” to “protectionism wins even after courts invalidate the tariffs.” That inference treats all adjustment as a durable scar and all rerouting as socially wasteful. Amiti, Redding, and Weinstein do find complete pass-through and a $1.4 billion monthly real-income loss by late 2018, but that is evidence of contemporaneous cost, not evidence that later legal relief and commercial re-optimization cannot substantially unwind the shock. A’s own refund examples show friction, not futility: the Supreme Court ruling stopped the legal basis for IEEPA collections, CBP announced it would cease collecting those duties, importers can use post-summary corrections, protests, and Court of International Trade litigation, and the government had indicated it would not oppose court-ordered reliquidation if it lost. That is not a perfect consumer-restitution mechanism, but customs law has recognized remedial channels for the party that legally paid the duty. The fact that GM had not yet received a $500 million refund days after reporting expected recovery is not strong evidence of lasting market transformation; it is evidence that large public refunds are administratively staged.

My first affirmative point is that businesses routinely manage policy shocks through pricing, sourcing, contracting, and inventory. That does not mean tariffs are costless; it means “uncertainty produces adjustment” is not the same as “the old market structure is unrecoverable.” If a tariff-induced supplier move remains after the tariff falls, the better inference may be that firms discovered a commercially viable alternative, not that the original tariff continues to dominate the market.

Second, some supply-chain shifts away from China are strategically relevant. USTR’s 2024 four-year review concluded that Section 301 tariffs reduced imports from China and increased imports from alternate sources, including allies and partners, potentially supporting diversification and resilience. It also concluded that aggregate welfare effects were small and negative while production in directly affected sectors rose. That is a mixed record, not a simple protectionist victory; it supports calibrated use of tariffs or threats where the United States seeks leverage against persistent technology-transfer and industrial-policy practices.

Third, Beijing’s Africa move should not be romanticized as pure predictability. Brookings notes China has run a persistent trade surplus with Africa since 2015, Africa’s deficit with China is about $60 billion, and African exports to China remain concentrated in minerals and raw or low-value products. Zero tariffs can be useful, but they do not by themselves solve standards, logistics, industrial upgrading, or bargaining asymmetries. China is offering access, but also preserving a geopolitical and commodity-supply strategy. The U.S. should reduce gratuitous volatility, but it need not abandon tariffs as leverage in an unstable trading system.

A’s rebuttal concedes the key ground: “not every tariff-induced adjustment is necessarily a deadweight scar” and some moves away from China may be resilient and commercially viable. That concession is decisive because A’s headline-level claim depends on treating persistence itself as evidence that protectionism has “won.” If a firm keeps a Vietnamese, Mexican, or allied-country supplier after a tariff falls because that supplier is now competitive, the lasting effect is not necessarily an improper residue of an illegal tariff; it may be ordinary business learning under policy pressure. A therefore must show net durable harm after refunds and re-optimization, not merely that prices and sourcing changed while the tariff was active.

A’s refund argument also overstates what its evidence proves. The AP figures—about $166 billion paid by more than 330,000 importers on more than 53 million shipments, with 60–90 day processing after approval—show administrative scale, not remedial failure. The relevant legal comparison is not “instant consumer restitution” versus “no remedy”; customs duties are legally paid by importers of record, and post-summary corrections, protests, and Court of International Trade litigation are recognized channels for recovery after the IEEPA ruling. McDermott and Thompson Coburn both describe normal customs mechanisms for refund claims, and Thompson Coburn notes the government had indicated it would not oppose court-ordered reliquidation after an adverse Supreme Court ruling. That undercuts A’s implication that legal invalidation is mostly symbolic.

A’s distributional point about discretionary exemptions is stronger, but it addresses a different institution. Political influence in Section 301 exclusions is a serious criticism of discretionary relief; it does not prove that court-ordered or statute-like refund administration after invalidation will be allocated by lobbying access rather than legal eligibility. A needs evidence that the IEEPA refund process itself is being politically rationed. So far, A has evidence of complexity, not evidence that improper collections cannot be substantially corrected.

On supply chains, A again converts a mixed record into a one-way indictment. USTR’s 2024 review says the Section 301 tariffs reduced imports from China and increased imports from alternate sources, including allies and partners, potentially supporting diversification and resilience. USITC separately found Section 232 and 301 tariffs reduced imports and increased U.S. production and prices in many affected industries. Those findings do not make tariffs costless; they show the policy can have intended strategic effects alongside costs. In an unstable trading system, leverage against technology-transfer practices, industrial subsidies, or coercive dependencies may be a legitimate policy objective.

Finally, A’s China-Africa contrast remains incomplete. Brookings’ analysis that Africa runs roughly a $60 billion deficit with China and exports remain concentrated in raw and low-value products means Beijing’s tariff-free outreach is not simply “predictability as power.” It is also a market-access tactic that can deepen commodity dependence and geopolitical asymmetry. The U.S. should avoid gratuitous volatility, but A has not shown that contested tariffs inevitably create irreversible net harm or that China’s alternative is a superior model rather than a competing strategy with its own risks.

A’s strongest argument is that a tariff can do economic work before a court ever reaches the merits: prices rise, contracts are rewritten, sourcing relationships move, and the party legally entitled to a refund is often not the party that ultimately bore the cost. A is also right that China does not need to be benevolent to exploit U.S. volatility; a simple, visible zero-tariff offer to African states can be diplomatically useful even if China-Africa trade remains asymmetric.

My position is still stronger because A has not supplied the missing causal step: evidence that the post-ruling world leaves a net, durable protectionist distortion that ordinary business adjustment and legal recovery cannot substantially correct. The NBER and USITC evidence shows real contemporaneous tariff costs, including pass-through and higher prices, but those findings do not prove irreversibility. They show that tariffs matter while they operate; they do not show that firms cannot re-optimize once duties end, that refund mechanisms fail at scale, or that retained non-China sourcing is economically inferior to pre-tariff sourcing. Persistence is not itself harm. If a company keeps a supplier in Vietnam, Mexico, India, or the United States after a tariff falls because the supplier is now proven and competitive, that is not protectionism “winning”; it may be learning, diversification, or resilience.

On refunds, A fairly emphasizes imperfect incidence: consumers and downstream firms may not receive a check. But the legal question is whether improper collections can be substantially corrected for the legal payor, not whether courts can recreate a complete welfare counterfactual. AP’s figures—about $166 billion across more than 330,000 importers and 53 million shipments—show scale and administrative friction, not futility. Trade lawyers’ guidance after the IEEPA ruling describes normal channels such as post-summary corrections, protests, and Court of International Trade litigation. Those remedies are not perfect, but A needs evidence that they systematically fail or are politically rationed in this refund setting.

A’s distributional evidence about discretionary Section 301 exclusions is serious, but it does not transfer cleanly to court-driven IEEPA refunds. Exemptions involve agency discretion over prospective relief; refunds after invalidation are more rule-bound and turn on entry records, liquidation status, and procedural deadlines. That still favors sophisticated firms at the margin, but it is not equivalent to political allocation.

Finally, China’s Africa policy is a tactic, not proof that predictability has defeated U.S. leverage. AP reports tariff-free access for 53 African countries, but Brookings notes a persistent structural imbalance: Africa’s exports to China remain concentrated in raw or low-value goods, and the continent runs a large deficit with China. The U.S. should avoid gratuitous volatility, but tariffs and tariff threats can remain legitimate bargaining tools when targeted at coercive dependencies, unfair practices, or supply-chain concentration risks.