President Trump’s Canada tariffs are headed to court. But the judiciary isn’t the only institution that should scrutinize his claims of foreign discrimination. Congress assigned that responsibility to an independent trade commission for a reason.

The International Trade Commission must act as an independent investigator, not a source of institutional legitimacy for conclusions the administration has already reached.

The commission is seeking public comments on how it should implement Section 338 of the Tariff Act of 1930, the long-dormant statute Trump invoked to impose tariffs of up to 50 percent. What qualifies as discrimination? What evidence is required? Should the commission’s findings be public? The answers to these questions will test whether the International Trade Commission’s independence means anything.

When Congress drafted Section 338’s predecessor, Section 317 of the Tariff Act of 1922, it could have left investigations to an executive agency that the president preferred. Instead, it assigned the job to an independent Tariff Commission, today’s International Trade Commission. That was a deliberate institutional choice. The point was to separate the facts from the politics.

Start with discrimination. Trump increasingly treats any disadvantage facing American exporters as evidence of foreign discrimination. The commission should distinguish unequal commercial outcomes from discrimination in fact. Consider Canada, where European cheese exporters enjoy preferences negotiated under the Canada-European Union Comprehensive Economic and Trade Agreement. American exporters receive different treatment under the U.S.-Mexico-Canada Agreement.

The administration calls that discrimination. But trade agreements result from bargaining. The EU negotiated its preferences in exchange for other concessions. The U.S. negotiated a different deal. Just because European exporters secured more favorable treatment doesn’t automatically mean American exporters are being discriminated against.

Congress anticipated this distinction. In 1922, the Senate struck a clause that would have made any failure to give American commerce equal treatment grounds for action and kept a test of discrimination instead. As Barry Appleton of New York Law School explains, Congress did not enact the broader equal treatment provision it considered.

Trade scholar Simon Lester has similarly argued that unequal effects alone should not establish discrimination under Section 338. The International Trade Commission should examine a measure’s design, structure and application to determine whether it discriminates in fact against American commerce, placing it at a demonstrable disadvantage relative to comparable foreign commerce.

Otherwise, Section 338 becomes a most-favored-nation clause that nobody signed, allowing Washington to claim concessions that others negotiated without offering anything in return. The commission should ask whether these differences satisfy the statute’s definition of discrimination, not simply whether American exporters would prefer a better deal.

The same rigor is also needed when countries retaliate against American tariffs. If a foreign government targets U.S. products in response to Trump’s tariffs, is that discrimination within the statute’s meaning? That conclusion requires an investigation, not simply a change in terminology from “retaliation” to “discrimination.”What about evidence? The commission should require more than mere allegations of unfair treatment. It should identify the foreign measure, establish how American commerce is treated differently from that of other countries and demonstrate the resulting commercial disadvantage.

It should also consider alternative explanations. Market share can decline for reasons unrelated to discrimination. Regulations can fetter American exporters without being discriminatory per commitments rather than hostility toward American commerce.

Most importantly, the International Trade Commission should assess the magnitude of any demonstrated disadvantage. This is where Section 338’s language becomes especially important.

Congress authorized duties of up to 50 percent to “offset” the burden or disadvantage of foreign discrimination. Trump’s initial Section 338 action against Canada imposed the statutory maximum on roughly $20 billion in imports, citing alleged discrimination involving automobiles, dairy and alcohol. Yet the tariffs covered a much broader assortment of Canadian products. How does a 50 percent tariff on those goods offset the particular disadvantage allegedly suffered by American exporters? Why wouldn’t 15 or 25 percent suffice?

Those are precisely the questions an independent Commission should ask. There is a difference between compensating for lost export opportunities, protecting American industries from import competition and exerting economic pressure to secure concessions. The president may prefer the last objective. But Section 338’s requirement to offset a disadvantage cannot simply be ignored.

Finally, the International Trade Commission’s findings and recommendations should be public. The commission can use established procedures for trade investigations, including published standards, opportunities for public comment and hearings. The justification for imposing tariffs on Americans does not deserve secrecy.

The courts will have to decide whether the commission’s findings must precede presidential action. But the point is that its independent fact-finding isn’t an obstacle to effective trade enforcement. It’s what gives its findings credibility.

Congress gave the International Trade Commission the responsibility to establish the facts, not to validate the president’s conclusions. A century later, the commission should demonstrate why that distinction matters.

Marc L. Busch is the Karl F. Landegger Professor of International Business Diplomacy at the Walsh School of Foreign Service, Georgetown University.