Past tariffs point to U.S. importers, but Canada's new 50% tariff burden is not yet observable

By Cumulant Research

The evidence available on July 20 does not show that Canadian suppliers will absorb less than half of the new tarifftariffA tariff is a tax imposed on a product when it crosses an international border. [CBP](https://www.cbp.gov/travel/international-visitors/know-before-you-visit/customs-duty-information) during its first 90 days. It cannot show that yet: the proclamations apply only to covered goods entered for consumption from August 19, so no post-effective Canadian import observations existed when the policy was announced. [Motor-vehicle proclamation](https://www.whitehouse.gov/presidential-actions/2026/07/imposing-additional-duties-to-offset-canadian-discrimination-against-the-commerce-of-the-united-states-with-respect-to-motor-vehicles/) [Alcohol proclamation](https://www.whitehouse.gov/presidential-actions/2026/07/imposing-additional-duties-to-offset-canadian-discrimination-against-the-commerce-of-the-united-states-with-respect-to-alcoholic-beverages/) [Dairy proclamation](https://www.whitehouse.gov/presidential-actions/2026/07/imposing-additional-duties-to-offset-canadian-discrimination-against-the-commerce-of-the-united-states-with-respect-to-dairy/)

History supplies a strong prior, not an observation. Two studies of the broader 2025 tariff episode found that foreign exporters absorbed a minority of the border-price burden. The New York Fed estimated foreign shares of 6% through August, 8% in September and October, and 14% in November. A Brookings conference paper estimated about 10% in its quarterly baseline and pass-throughpass-throughPass-through is the portion of a tariff reflected in a later price, such as the duty-inclusive price paid at the border. [New York Fed](https://libertystreeteconomics.newyorkfed.org/2026/02/who-is-paying-for-the-2025-u-s-tariffs/) between 80% and 100% across controls and time horizons, implying foreign shares between about 0% and 20%. [New York Fed](https://libertystreeteconomics.newyorkfed.org/2026/02/who-is-paying-for-the-2025-u-s-tariffs/) [Brookings](https://www.brookings.edu/wp-content/uploads/2026/03/1_Fajgelbaum-Khandelwal_unembargoed.pdf)

President Donald Trump signed three proclamations under Section 338Section 338Section 338 of the Tariff Act of 1930 authorizes duties of up to 50% when the president finds that another country disadvantages U.S. commerce through unequal or discriminatory treatment. [USTR](https://ustr.gov/about/policy-offices/press-office/press-releases/2026/july/ambassador-greer-issues-statement-president-trump-imposing-section-338-tariffs-canada) of the Tariff Act of 1930. Each adds a 50% dutydutyA duty is a tax owed when covered merchandise is imported. [CBP](https://www.cbp.gov/travel/international-visitors/know-before-you-visit/customs-duty-information) to a specified list of Canadian products beginning at 12:01 a.m. Eastern time on August 19. The White House said covered products include goods ranging from wine to hockey sticks and cement, and USTRUSTRUSTR is the Office of the United States Trade Representative, the federal agency responsible for U.S. trade policy and negotiations. [USTR](https://ustr.gov/about/policy-offices/press-office/press-releases/2026/july/ambassador-greer-issues-statement-president-trump-imposing-section-338-tariffs-canada) said the combined actions cover nearly $20 billion in imports. USTR's release did not identify that figure as annual, so this article does not describe it that way. [White House fact sheet](https://www.whitehouse.gov/fact-sheets/2026/07/fact-sheet-president-donald-j-trump-imposes-additional-tariffs-on-canada/) [USTR](https://ustr.gov/about/policy-offices/press-office/press-releases/2026/july/ambassador-greer-issues-statement-president-trump-imposing-section-338-tariffs-canada)

The duties apply to listed goods even when they qualify for USMCA preferences. The proclamations exclude goods already subject to Section 232Section 232Section 232 is a separate U.S. authority for restricting imports found to threaten national security, and products subject to its duties are excluded from these Section 338 duties. [White House](https://www.whitehouse.gov/presidential-actions/2026/07/imposing-additional-duties-to-offset-canadian-discrimination-against-the-commerce-of-the-united-states-with-respect-to-motor-vehicles/) duties and aircraft covered by the civil-aircraft trade agreement, while the White House fact sheet also identifies exclusions including energy, potash, fish and critical minerals. [White House fact sheet](https://www.whitehouse.gov/fact-sheets/2026/07/fact-sheet-president-donald-j-trump-imposes-additional-tariffs-on-canada/) [Motor-vehicle proclamation](https://www.whitehouse.gov/presidential-actions/2026/07/imposing-additional-duties-to-offset-canadian-discrimination-against-the-commerce-of-the-united-states-with-respect-to-motor-vehicles/)

The two governments dispute the measures' legal and political meaning. The U.S. administration describes Canada's treatment of American autos, alcohol and dairy as discriminatory. Canada's prime minister describes the Canadian auto measures as matching earlier U.S. tariffs and argues that the U.S. actions violate CUSMA. These are opposing government positions, not findings reached by this incidence analysis. [USTR](https://ustr.gov/about/policy-offices/press-office/press-releases/2026/july/ambassador-greer-issues-statement-president-trump-imposing-section-338-tariffs-canada) [Prime Minister of Canada](https://www.pm.gc.ca/en/news/statements/2026/07/20/statement-prime-minister-carney-united-states-administrations-intention)

Official Canadian records confirm that Canada has imposed 25% auto counter-tariffs since April 9, 2025 and ties tariff relief for automakers to production and investment commitments. Separate Canadian notices show that retailers are ineligible for the USMCA cheese quota but eligible under the European Union agreement's cheese quota. Those records establish that the underlying measures exist; they do not settle whether Section 338's legal test is satisfied. [Finance Canada](https://www.canada.ca/en/department-finance/programs/international-trade-finance-policy/canadas-tariff-responses/canadas-tariffs-automobiles.html) [USMCA cheese notice](https://www.international.gc.ca/trade-commerce/controls-controles/notices-avis/1079.aspx?lang=eng) [European Union cheese notice](https://www.international.gc.ca/trade-commerce/controls-controles/notices-avis/993_2.aspx?lang=eng)

The importer of recordimporter of recordThe importer of record is the party legally responsible for an import entry and payment of the duties and taxes due. [CBP](https://www.cbp.gov/sites/default/files/documents/icp073_3.pdf) is legally responsible for depositing estimated duties with CBPCBPCBP is U.S. Customs and Border Protection, the agency that processes import entries and collects duties. [CBP](https://www.help.cbp.gov/s/article/Article-1643?language=en_US). That answers who writes the check to the government, not who ends up poorer. [CBP entry process](https://www.help.cbp.gov/s/article/Article-1643?language=en_US) [CBP importer guide](https://www.cbp.gov/sites/default/files/documents/icp073_3.pdf)

Incidence asks where the economic loss lands after prices and quantities adjust. A Canadian supplier can absorb part of the tariff by cutting its before-duty invoice. A U.S. importer can absorb part by accepting a lower margin. A downstream business can raise its price, passing some cost toward consumers. The party at the Customs window and the party whose wallet ends thinner need not be the same. [New York Fed](https://libertystreeteconomics.newyorkfed.org/2026/02/who-is-paying-for-the-2025-u-s-tariffs/)

The arithmetic is easy to misread because an ad valorem dutyad valorem dutyAn ad valorem duty is a tax calculated as a percentage of an imported product's customs value. [White House](https://www.whitehouse.gov/presidential-actions/2026/07/imposing-additional-duties-to-offset-canadian-discrimination-against-the-commerce-of-the-united-states-with-respect-to-motor-vehicles/) is charged on the new invoice value. If an invoice falls from $100 to $80, the additional 50% duty is $40. The seller gives up $20 and the importer's duty-inclusive cost rises from $100 to $120, also a $20 loss. That is the 50-50 border split.

The New York Fed studied monthly U.S. import data at the HTS10HTS10HTS10 is the ten-digit U.S. product classification used for tariff administration and detailed import statistics. [CBP](https://www.help.cbp.gov/s/article/Article-1015?language=en_US)-country level and used changes in unit values as a proxy for before-duty export prices. Its estimated U.S. importer share fell from 94% in January-August 2025 to 92% in September-October and 86% in November. Put plainly, foreign suppliers cut their measured prices somewhat more later in the year, but U.S. importers still bore most of the estimated border-price increase. [New York Fed](https://libertystreeteconomics.newyorkfed.org/2026/02/who-is-paying-for-the-2025-u-s-tariffs/)

Pablo Fajgelbaum and Amit Khandelwal reached a similar central estimate using quarterly data. Their 2025 baseline put pass-through to duty-inclusive import prices at 90%, meaning foreign exporters absorbed about 10%. Depending on controls and measurement horizon, their estimates ranged from 80% to 100% pass-through. [Brookings](https://www.brookings.edu/wp-content/uploads/2026/03/1_Fajgelbaum-Khandelwal_unembargoed.pdf)

Neither result is a Canadian forecast. The 2025 tariff changes covered many products, countries and rates. The new action targets selected Canadian goods at an additional 50%, and suppliers may differ in contracts, margins, market power, production costs and ability to redirect sales. Applying a historical average directly to Canada would turn evidence into false precision.

A currency or stock-price move after the announcement would be a market reaction, not a measurement of the tariff's economic incidence. This analysis therefore does not use market moves to decide who will bear the duty.

A February 2026 working paper by Sharat Ganapati and Colin Hottman challenges a common reading of customs unit values. A unit valueunit valueA unit value is reported customs value divided by reported quantity, so it can change when the mixture or size of shipments changes. [Census Bureau](https://www.census.gov/foreign-trade/guide/sec2.html) is an average: total customs valuecustoms valueCustoms value is generally the price paid for imported merchandise before U.S. duties and international freight, insurance and related transport charges. [Census Bureau](https://www.census.gov/foreign-trade/guide/sec2.html) divided by quantity. It can rise because the price of the same item rose, but also because buyers ordered a different mixture of products or placed smaller orders that lost quantity discounts. It is like watching the average price of a grocery basket rise after shoppers replace store-brand items with premium ones: the average changed even if no shelf label did. [Census working paper](https://www2.census.gov/library/working-papers/2026/adrm/ces/CES-WP-26-17.pdf) [Census Bureau](https://www.census.gov/foreign-trade/guide/sec2.html)

For continuing relationships from June 2018 to June 2019, the paper reports conventional pass-through of 0.871 and scale-corrected estimates of 0.603 and 0.572. Its argument is that smaller and less frequent orders pushed up per-unit costs, masking price concessions made through lower markups. [Census working paper](https://www2.census.gov/library/working-papers/2026/adrm/ces/CES-WP-26-17.pdf)

The paper then asks a broader question than the New York Fed study. Its baseline welfare calculation combines changes in trade value, quantity, costs, markups and government revenue. It estimates a $22.9 billion loss of foreign-producer surplus and a $16.6 billion loss for domestic buyers, both annualized in 2017 dollars, making the foreign share of the combined seller-buyer loss 58%. A transaction-only specification produces 30%, while a calculation approximating entry and exit produces 54%. [Census working paper](https://www2.census.gov/library/working-papers/2026/adrm/ces/CES-WP-26-17.pdf)

Those percentages are not competing estimates of the same object. The New York Fed and Brookings estimates ask how the border price moved relative to the tariff. The Census working paper asks how economic surpluseconomic surplusEconomic surplus is the benefit a buyer or seller receives from a transaction beyond what was needed for the transaction to occur. changed after quantities, production costs and trading relationships moved too. Comparing them without that distinction is like comparing the change on a receipt with the change in a business's annual profit.

The clean test begins with transaction-level entries for the tariffed Canadian product codes after August 19. Each observation needs the before-duty customs value, quantity, duty collected, supplier, importer, product code, entry date and related-party status. Census explains that customs value generally excludes U.S. duties and international transport charges, while affiliated transactions require care because the relationship should not influence the declared value. [Census Bureau](https://www.census.gov/foreign-trade/guide/sec2.html)

• Track identical or very narrowly defined products so a switch from expensive goods to cheaper goods is not mistaken for a price cut.
• Separate continuing supplier-importer relationships from suppliers that enter or leave, because the intensive and extensive margins answer different questions.
• Control for quantity per shipment so the loss of bulk discounts is not mislabeled as a change in supplier pricing.
• Compare exposed Canadian products with credible controls, such as unaffected Canadian products or the same products from other origins, while accounting for product and date effects.
• Report related-party and unrelated transactions separately because affiliated companies can set internal invoice prices differently.
• Calculate both border-price incidence and the broader changes in quantities and surplus rather than treating one measure as the whole economic effect.

The pre-period also needs special handling. Importers can front-load shipments before a tariff begins, and an IMF study documents such acceleration before tariff increases during the 2024-2025 U.S.-China episode. A rush of Canadian goods before August 19 could alter the products and order sizes observed immediately afterward. [IMF](https://www.elibrary.imf.org/view/journals/001/2026/013/001.2026.issue-013-en.xml)

Finally, the result should carry an uncertainty range. A point estimate just below 50% is not meaningful evidence of less than half if its confidence intervalconfidence intervalA confidence interval is a range produced by a statistical method to express uncertainty around an estimate. [NIST](https://www.itl.nist.gov/div898/handbook/prc/section1/prc14.htm) crosses 50%. NIST explains that a confidence level describes the long-run performance of the interval-producing method, not a probability assigned directly to one completed interval. [NIST](https://www.itl.nist.gov/div898/handbook/prc/section1/prc14.htm)

The most relevant historical evidence points toward U.S. importers bearing more than half of comparable tariff shocks at the border. That is useful for setting expectations and rejecting the assumption that foreign suppliers automatically pay the whole tariff. [New York Fed](https://libertystreeteconomics.newyorkfed.org/2026/02/who-is-paying-for-the-2025-u-s-tariffs/) [Brookings](https://www.brookings.edu/wp-content/uploads/2026/03/1_Fajgelbaum-Khandelwal_unembargoed.pdf)

But the newer welfare analysis shows why the measurement choice matters, and none of the cited studies observes the Canadian products, contracts and supplier responses created by the July 20 proclamations. [Census working paper](https://www2.census.gov/library/working-papers/2026/adrm/ces/CES-WP-26-17.pdf)

The evidence leans below half for Canadian suppliers at the border. It does not yet prove it.