How do repeated border crossings and limited substitution options determine who bears the cost of the US-Canada trade escalation?

My position is that importing-country manufacturers and consumers will bear most of the near-term tariff burden, especially where firms cannot substitute inputs quickly. This is an evidence-based prediction, not yet a measured result for September’s escalation. Tariff incidence means the ultimate distribution of economic costs—not simply who remits the tax. Historical evidence strongly supports distinguishing the two. (princeton.edu)

First, establish what is actually operative. Canada’s customs notice confirms duties of 15%, 25% or 50% on specified US-origin goods from September 8, 2026, assessed on customs value for duty. AP’s approximately $20 billion describes covered goods, not tax receipts; its report does not establish the underlying valuation methodology. The announced US product bans start September 29, while additional 50% automotive restrictions discussed for next year remain threats. Those future measures cannot explain duty payments already recorded. (cbsa-asfc.gc.ca)

Second, the strongest incidence precedent is Amiti, Redding and Weinstein’s 2020 study, extending analysis through much of 2019: US firms and consumers continued to bear almost all tariff costs overall. The strongest opposing argument nevertheless has evidence: foreign steel exporters reduced prices enough to bear close to half that sector’s tariff cost, and import volumes became more responsive over time. Thus exporter losses and eventual substitution are real possibilities, but neither establishes that foreign suppliers generally finance the importing country’s tariff. (princeton.edu)

Third, repeated crossings matter only when successive entries are taxable. Rules of origin determine a product’s economic nationality; shipping through another country does not necessarily change it. USMCA—the United States–Mexico–Canada Agreement, called CUSMA in Canada—also governs relief. Duty drawback refunds qualifying import duties, commonly after re-export. Crucially, Canada’s notice permits potentially full relief for qualifying originating goods, exempts qualifying Import for Re-Export Program shipments, and protects specified repair-and-return movements. It also explicitly prevents stacking two named surtaxes. Counting crossings without applying these provisions would exaggerate costs. (cbsa-asfc.gc.ca)

Substitution constraints remain consequential. FAA replacement-part approval requires demonstrated airworthiness compliance, rather than merely finding a cheaper fabricator. Canada’s dairy re-export rules require recognized US milk sources, licensed facilities and haulers, documentation and segregation. My inference is that these requirements narrow immediately usable alternatives; neither source establishes a universal switching time. A non-tariff barrier is a trade restriction other than a customs duty, such as the announced import bans. (faa.gov)

The decisive investigation should therefore match shipment duties minus refunds to component histories, supplier quotations, approval dates and contractual price-adjustment clauses. Specialized beverage-input dependence remains a hypothesis requiring purchasing records, not an established finding here. Compare affected products’ prices and margins against unaffected products, controlling for exchange rates and input costs. My prediction is that constrained domestic buyers show larger margin compression or downstream price increases; qualifying re-exporters should show substantially less. That distinction determines whether these measures primarily protect domestic production or raise its costs.

B’s strongest point is that exemptions must determine measured exposure, not appear as an afterthought: a border crossing is not necessarily a taxable event. I accept that correction. Canada’s September 8, 2026 notice excludes qualifying dairy Import for Re-Export Program shipments and specified repair movements, and permits potentially full relief for qualifying CUSMA-origin goods. These flows cannot substantiate repeated Canadian taxation. But eliminating a duty does not establish that foreign producers bear duties remaining on other shipments. Exposure and incidence are separate questions. (cbsa-asfc.gc.ca)

My position therefore remains a near-term prediction about net exposed trade, not a demonstrated aggregate result for this escalation. Amiti, Redding and Weinstein find that US firms and consumers bore almost all tariff costs overall through October 2019, despite foreign steel producers absorbing close to half their sector’s burden. That exception warrants sector-specific estimates; it does not reverse the broader empirical baseline. Extrapolation to Canada remains uncertain. (princeton.edu)

Additional evidence directly addresses B’s demand to distinguish prices from margins. Cavallo, Gopinath, Neiman and Tang combined border microdata with product-level retailer data: tariffs were almost fully reflected in importer costs, while affected-versus-unaffected retail-price differences were generally modest, suggesting retailer margin compression. Consequently, limited shelf-price increases would not establish limited domestic incidence. Importantly, the same study finds price reductions by US exporters facing retaliation—evidence against assuming identical incidence in both directions. (nber.org)

B’s contract mechanism also needs a precise counterfactual. In an illustrative transaction, fixing the foreign supplier’s invoice at 100 does not prevent the domestic importer’s landed cost becoming 125 under a separately assessed 25% duty. Foreign absorption requires a price concession or contractual supplier liability, not simply enforcement of the original invoice. Canada assesses the applicable surtax on customs value. Likewise, two genuinely taxable production-stage entries can generate cumulative costs, but each entry requires its own origin, classification, valuation and relief determination; I am not claiming a verified two-entry shipment. (cbsa-asfc.gc.ca)

Substitution remains a plausible amplifier rather than a measured universal delay. FAA replacement-part approval requires airworthiness compliance through testing and computations or qualifying identicality. My inference is that an unapproved alternative is not immediately interchangeable; existing approved alternatives could materially weaken that constraint. Firm-specific switching times and specialized beverage-input dependence still require purchasing records. (faa.gov)

Finally, DISCUS’s reported decline in Canadian-bound spirits exports from US$203 million to US$60 million between March–December 2024 and the corresponding 2025 period establishes substantial exporter sales exposure associated with provincial bans, not who financed a tariff on surviving shipments. Those losses belong in the broader escalation assessment and could alter its aggregate verdict. (distilledspirits.org)

The decisive test should stratify shipments by relief eligibility, reconcile payments with refunds, and compare exporter prices, domestic margins and downstream prices against unaffected products, controlling for exchange rates and input costs. Add approval dates and contract terms. My prediction remains greater domestic cost absorption where net duties persist and substitution is constrained.

B’s strongest position is that exemptions determine actual exposure, while exporter concessions and lost business can make foreign suppliers bear substantial costs; counting crossings or citing US import averages cannot establish this escalation’s aggregate incidence. Canadian relief provisions and evidence of US exporter price reductions under earlier retaliation support that objection. (cbsa-asfc.gc.ca)

My position remains stronger as a conditional prediction because the historical asymmetry follows substitutability, not simply which country retaliates. Cavallo, Gopinath, Neiman and Tang attribute exporter concessions principally to less-differentiated goods with available substitutes, with exports to China driving the result; differentiated imports instead showed almost complete pass-through to importing buyers. Thus B’s retaliation evidence strengthens the case for comparing substitution constraints rather than assuming Canadian buyers obtain similar concessions. This is an inference, not a measured Canadian outcome. (bostonfed.org)

Those constraints have a documented basis, although their duration remains unmeasured. FAA replacement-part design approval requires demonstrated airworthiness compliance or qualifying identicality. Canadian dairy re-export rules require recognized US milk sources, licensed haulers and segregation. These requirements establish eligibility hurdles, not universal switching delays; existing approved alternatives could defeat my prediction. Specialized beverage-input dependence remains unverified in the evidence presented. (faa.gov)

Exposure must nevertheless come first. Canada’s enacted duties began September 8, 2026, at 15%, 25% or 50% of customs value, but qualifying re-export permits, repair movements and CUSMA relief can eliminate Canadian liability. Therefore, the dairy rules above cannot themselves establish cumulative taxation. Repeated production-stage crossings support my argument only when linked customs entries establish successive duties remaining after refunds; the Canadian notice alone establishes no US-entry liability. (cbsa-asfc.gc.ca)

Nor do modest consumer-price changes vindicate foreign incidence. Cavallo and coauthors associate a ten-percentage-point tariff increase with only a 0.44-percentage-point relative retail-price increase after one year, alongside near-complete importer-cost pass-through—a pattern consistent with domestic margin compression. My claim concerns domestic firms and consumers together, not consumers alone. (bostonfed.org)

I would change my mind on the main-burden claim if a representative, exposure-weighted study covering both trade directions found foreign producers bearing over half the measured private economic loss during the first twelve months. It should link customs entries and completed refunds to supplier invoices, contracts, approval dates, domestic margins and downstream prices; compare affected products with credible unaffected controls; and account for exchange rates, input costs and pre-existing trends. Exporter losses must include causally attributable profits forgone on cancelled sales, not gross revenue declines or double-counted concessions. Findings that relief eliminates most repeated-stage exposure and documented supplier changes occur before meaningful domestic losses would separately overturn my cumulative-cost and substitution mechanisms. Until those results exist, I defend domestic incidence as the better-supported conditional prediction, not an established aggregate verdict.