What did the Supreme Court actually rule, why aren't consumers seeing relief, and what does the pattern reveal about how tariff costs are absorbed, passed on, or captured?

My position is that the refund windfall is not an administrative footnote; it is the clearest money-trail evidence that tariff incidence largely ran through American and Canadian households rather than being durably absorbed by corporate margins. The Supreme Court’s ruling in Learning Resources, Inc. v. Trump held that IEEPA—the International Emergency Economic Powers Act, a 1977 emergency-powers statute—does not authorize the president to impose broad customs duties. That matters because the ruling invalidated the 2025 IEEPA tariffs, including emergency drug-trafficking tariffs on Canada, Mexico, and China and worldwide trade-deficit or “reciprocal” tariffs, but it did not invalidate Section 232 national-security tariffs on steel, aluminum, autos, or Section 301 unfair-trade-practice tariffs on China. So the Court fixed a separation-of-powers problem, not the household-incidence problem. CRS and SCOTUS materials explicitly noted that refunds would go to importers even though some had already passed costs onward. BEA later described the refund obligation as roughly $166 billion to affected importers and businesses, excluding interest. (supreme.justia.com)

The first empirical pillar is pass-through. “Tariff pass-through” means the share of a tariff that appears in downstream prices rather than being absorbed by foreign exporters, importers, wholesalers, or retailers. The best 2018 evidence does not support the claim that foreign countries paid the tariffs: Amiti, Redding, and Weinstein found complete pass-through into domestic prices of imported goods and estimated a $1.4 billion monthly real-income loss by the end of 2018; Fajgelbaum, Goldberg, Kennedy, and Khandelwal similarly found targeted import prices did not fall, implying complete pass-through to duty-inclusive prices. Cavallo and coauthors complicate the story by finding less retail-store pass-through in some categories, implying some retail-margin compression, but even that is not a vindication of tariff policy: it means the burden fell on domestic firms and consumers, not foreign governments. (ftp.aeaweb.org)

The second pillar is the refund distribution. Refunds legally follow the importer of record, not the household that paid a higher shelf price. Public reporting and Senate letters identify large actual or expected refunds: Walmart reportedly up to $10.2 billion, Apple $2.2 billion, Target up to $2.2 billion, Amazon about $600 million in Q2 2026, and Energizer around $50 million; Forbes reported more than $5.3 billion disclosed by top retailers and that executives mostly described reinvestment rather than consumer rebates. That is exactly the double-gain mechanism: raise prices when tariffs hit, receive a government refund when tariffs are voided, and face no automatic obligation to reverse the original price increase. (warren.senate.gov)

The third pillar is Canada. Bank of Canada work found Canadian firms were planning substantial tariff pass-through, with roughly three-quarters of firms expecting to pass on more than half of tariff-related cost increases, and later estimated that if tariffs and counter-tariffs stayed in place, past experience suggested about 75% pass-through over roughly 18 months. Canada’s counter-tariffs are not IEEPA refunds, but they demonstrate the same incidence logic on the northern side of the border: tariffs become consumer prices unless policy deliberately rebates households. (bankofcanada.ca)

B’s strongest point is genuinely important: border pass-through and retail-shelf pass-through are not identical. Amiti, Redding, and Weinstein’s “complete pass-through” result is primarily about duty-inclusive import prices, while Cavallo-style retail evidence and the Federal Reserve’s 2025 work show that retailers often adjusted gradually and absorbed part of the shock through margins, inventories, sourcing shifts, and delayed repricing. That means A should not claim every tariff dollar instantaneously appeared one-for-one on the shelf. The strongest version of B’s point is that some of the $166 billion refund may restore margins that were compressed during the tariff period, not represent a pure second profit on every transaction.

But that concession does not defeat my position; it narrows it. The core question is not whether retail pass-through was literally 100% in every category and month. It is whether the economic burden was shifted substantially downstream and whether the legal refund mechanism returns money to the parties that ultimately bore that burden. On that question, the evidence still favors A. First, the Supreme Court’s Learning Resources decision created a legal remedy for importers, not for downstream consumers or small businesses. The Court itself flagged the precise problem: billions may have to be refunded to importers even though some importers may already have passed costs to consumers or others. BEA then quantified the federal obligation at about $166 billion, excluding interest, as repayments to affected importers and businesses. That is a distributional fact, not a rhetorical claim.

Second, B’s “partial and gradual” retail pass-through evidence still confirms household incidence. The Fed’s 2025 retail analysis found at least 30% consumer pass-through for China-origin goods by December 2025 and an 8.5% year-over-year price rise for those goods; New York Fed evidence found nearly 90% of the overall economic burden fell on U.S. firms and consumers rather than foreign exporters. That is not “foreign countries paid”; it is domestic incidence divided between households and firms. Once refunds flow only to importers, both groups downstream from the importer remain exposed unless there is an enforceable rebate chain.

Third, the corporate-recipient evidence fits the concern. Warren’s letters identify Walmart, Apple, Target, Amazon, and Energizer as receiving or expecting large refunds, while noting no general consumer-refund mechanism. Even if Walmart or another firm voluntarily uses some refund for price rollbacks, voluntarism proves the institutional defect: consumers have no legal claim proportional to what they paid. Finally, Canada reinforces the pattern rather than undermining it. Bank of Canada analysis assumes roughly 75% tariff cost pass-through to consumer prices within six quarters. So B is right about incomplete short-run shelf pass-through, but wrong that this dissolves the windfall problem. Partial pass-through plus importer-only refunds still means businesses capture the refund while households and downstream firms receive relief only at corporate discretion.

B’s strongest argument is that A risks overstating “straight through” pass-through by treating border-price evidence as if it were identical to retail-shelf evidence. The strongest empirical version of B is: Cavallo-style retail studies, the Fed’s 2025 work, Goldman’s incidence estimates, and firm cases like GM show that many companies absorbed meaningful tariff costs through lower margins, inventory timing, or delayed repricing, so refunds may partly restore losses rather than create a pure second profit. B is also right that downward price adjustment can be sticky, so unchanged shelf prices immediately after refunds are not by themselves proof of opportunistic profiteering.

I still think A’s position is stronger because B’s evidence weakens only the literal claim of 100% immediate shelf pass-through; it does not solve the distributional problem created by importer-only refunds. The central question is not whether every retailer passed every tariff dollar to consumers in the same month. It is whether tariff costs were substantially shifted downstream, and whether the refund mechanism is linked to who actually bore those costs. On that question, the evidence favors A. The Supreme Court in Learning Resources held that IEEPA did not authorize broad unilateral tariffs and specifically recognized the pass-on problem: refunds may go to importers even where importers already passed costs to consumers or others. BEA’s accounting confirms the scale: about $166 billion, excluding interest, was treated as repayments to affected importers and businesses, not as consumer compensation.

B’s own incidence evidence supports A’s core point that the costs were domestic, not foreign-paid. Goldman’s reported August 2025 estimate put 37% on consumers and 51% on U.S. businesses, while expecting the consumer share to rise to 55% by year-end. The Federal Reserve’s retail evidence likewise found real consumer pass-through: at least 30% for China-origin goods by December 2025, with those goods up 8.5% year over year. That is enough to make importer-only refunds overinclusive whenever importers raised wholesale or retail prices, because the legal refund does not trace incidence through the supply chain.

Canada reinforces rather than dilutes this conclusion. Bank of Canada staff assumed 75% tariff-cost pass-through to consumer prices within six quarters in tariff scenarios, and estimated roughly 60% pass-through after six quarters from the 2018 experience. That cross-border evidence is not identical to U.S. IEEPA refunds, but it shows the same economic mechanism: tariffs migrate into prices over time unless policy deliberately prevents that. Therefore B proves an important qualification—partial margin absorption occurred—but A remains stronger because partial pass-through plus untraced importer refunds still creates a one-way institutional design: households pay through prices, while refunds return to businesses by legal status rather than economic burden.