Every so often, a technical story reveals a simple truth—showing how easily numbers, once baked into the system, can become policy. A recent essay by analyst Alice Palmer in Sustainable Forests, Resilient Industry—“US Duty Calculations on Canadian Lumber Are Flawed”—does exactly that. Using clear examples and straightforward arithmetic, she shows how the US Department of Commerce’s anti-dumping calculations turn fair trade into a numerical fiction—and why, even as markets shift, the duties stay high.

Of course, there is a broader backdrop to these numbers—one shaped by lobbying pressure, political alliances, and the United States’ wider turn toward protectionism. But it remains worth examining the calculations themselves, since the Department of Commerce maintains that duty rates are objective and data-driven. As Palmer’s analysis shows, the arithmetic tells a different story.

Her illustrations walk readers through a hypothetical set of lumber transactions in which average prices are identical in both countries. When Commerce uses the average-to-average method, the dumping margin is zero. But when it applies zeroing—counting only the lower-priced US sales and setting the higher-priced ones to zero—a nonexistent gap suddenly appears. The result is a calculated dumping margin of 6.1% where none actually exists.

Palmer’s finding naturally raises a broader question: if one methodological choice can create a margin from nothing, are other elements in the system doing similar work? Building on her analysis, Tree Frog reached out to Palmer and looked further into how the duty calculations are made—not to relitigate old debates, but to understand why the numbers stay high even when markets normalize.

First, the anti-dumping side of the equation. The issue is not just that the math can be wrong—as Palmer demonstrated; it’s that the methodology tends to produce an inflated result. Zeroing, long questioned by the World Trade Organization, has been found to inflate anti-dumping margins[i] by removing the very data that would show no unfair pricing. What looks like precision—duty rates carried to two decimal places—is the result of a methodology that excludes contrary data. That helps explain why anti-dumping duties on Canadian softwood exports often remain elevated, even when current price data between the two markets show little difference.

In parallel with anti-dumping duties, the United States also applies countervailing duties (CVDs)—meant to offset what it considers to be government subsidies to Canadian producers. Whereas anti-dumping duties target pricing behavior, countervailing duties address alleged policy advantages such as below-market stumpage fees or other forms of support. The next question, then, was whether the methods used to identify these “subsidies” might introduce distortions of their own.

Under US trade-law practice, Commerce identifies whether Canadian producers receive a “benefit” from provincial stumpage or other measures. In many instances, Commerce has substituted private-land stumpage prices from the US South as the benchmark for comparison, rather than relying exclusively on Canadian stumpage sales or auctions. More recently, it has incorporated private-market data from Nova Scotia into this analysis—arguing that these prices better reflect market-based rates. In both cases, the result is the same: two fundamentally different systems are treated as if they were comparable. In Canada, where 90% of forestland is public, stumpage reflects a managed-resource model with built-in obligations for regeneration, roads, and environmental protection. By contrast, in the US South, most timber comes from private land sold through competitive auctions with minimal public-resource obligations—an approach that reflects market dynamics but overlooks structural differences in Canada’s tenure system, as well as the volatility that auction pricing introduces in a cyclical commodity like lumber. The WTO has repeatedly questioned these benchmarks as non-comparable[ii], yet Commerce has reaffirmed its approach in later reviews, citing such prices “in the absence of usable in-country benchmarks.”[iii]

While stumpage remains the centerpiece of Washington’s subsidy claim, Commerce’s determinations also list dozens of other federal and provincial programs—from road-building offsets and reforestation credits to energy pricing, tax incentives, and R&D grants—reflecting the broad mix of supports typical in resource industries on both sides of the border. In practice, these contribute little to the overall duty calculation—typically less than a percentage point each. WTO panels have generally found such measures to be either broadly available or too indirect to constitute a trade-distorting benefit.[iv] Similarly, state-level incentives for mill modernization or job creation in the US South—while common—generally fall outside the scope of what US trade-remedy law defines as countervailable subsidies. 

Finally, timing. Although Commerce does not cherry-pick review windows—investigations follow statutory annual and administrative timelines—the periods of investigation typically cover about one year of data. In a volatile commodity like lumber, that fixed window can have unintended effects.[v] When an investigation period happens to coincide with a market trough, export prices appear abnormally low and stumpage—automatically adjusted to market prices—appears to confer a benefit. When prices rebound, the “negative dumping” that would offset earlier distortions is excluded because the zeroing method discards higher-priced transactions. The combination of short review periods and zeroing can therefore turn ordinary price cycles into apparent evidence of unfair trade. A more neutral approach could average data across multiple years or align comparison windows so both sides reflect similar market conditions. Instead, data from a temporary downturn can influence duty rates long after prices recover.

Taken together, the effects point to a consistent pattern: much of the duty burden reflects method and timing rather than market reality. Reviews by independent analysts—including WTO panel findings and Palmer’s own calculations—suggest that methodological effects account for a substantial share of those anti-dumping and countervailing duties, potentially in the range of a third to a half of the total.[vi] Of these, a significant share—perhaps half—stems from timing and calculation choices rather than actual cost or price differences.

The Section 232 tariffs stand apart entirely. Most analysts—Palmer among them—see little economic or legal basis for their application to Canadian lumber. The rationale that Canada poses a national-security threat in the lumber market has been widely criticized on both sides of the border as political rather than factual.[vii] These tariffs are therefore best understood as a separate, policy-driven layer atop the trade-law mechanisms that govern the anti-dumping and countervailing duties.

Even at conservative estimates, most of what Canadian exporters now pay stems more from how and when the math is done than from any real competitive advantage. The selective arithmetic of zeroing, the use of non-comparable benchmarks, and the timing distortions of price cycles all point to the same conclusion: the perception of unfair trade persists largely because of how it is calculated.

If the anti-dumping and countervailing duties were recalculated using complete data (no zeroing), domestic benchmarks (no non-comparable price substitution), and up-to-date prices (no cycle lag), their combined rate—now roughly 35%—could fall to minimal levels. In practical terms, that would provide a more accurate measure of competitiveness on both sides of the border and a clearer signal to producers, builders, and consumers alike.

In that wider context, the methodological issues described here are not just statistical—they reflect a system without an effective referee. The WTO Appellate Body remains dormant after the US blocked new appointments, and the Canada–US–Mexico Agreement (known in Canada as CUSMA and in the United States as USMCA) offers no practical remedy. The usual checks on bias have eroded, leaving little recourse for affected industries. Political-risk analyst Robert McKellar—whose earlier Tree Frog News commentary,Trump’s Second Term and the Canadian Forest Sector, examined the institutional risks of US protectionism—argues that this represents “a structural vulnerability: when the rules are written by the same players who benefit from them.” Just recently, the US Lumber Coalition reinforced that trend, urging that any USMCA extension be conditioned on eliminating the Chapter 10 binational panel review process—a move that would effectively eliminate external oversight of US trade-remedy decisions.

At a policy level, nor is relief likely to come from regional trade agreements. The Canada–United States–Mexico Agreement (CUSMA) remains important for overall trade stability, but the softwood lumber dispute operates outside its framework. The upcoming CUSMA review may shape the tone of North American relations, yet it will not resolve the duty methodology itself. Inside the United States, the political economy reinforces the status quo: the US Lumber Coalition continues to lobby for protection, while home builders warn that high tariffs worsen housing affordability. As McKellar has noted, this tension between political power and economic logic typifies today’s protectionist era—a system where duties reflect not only distorted math but also the absence of a functioning arbiter to restore balance.

 

__________________________________

Endnotes:

[i] The World Trade Organization has repeatedly ruled that “zeroing” inflates dumping margins by excluding offsetting data. See, for example, the WTO Analytical Index, Anti-Dumping Agreement, Article 2.4.2; and “Zeroing In: Antidumping’s Flawed Methodology Under Fire,” Cato Institute (2006).

[ii] World Trade Organization, United States – Countervailing Duty Measures on Softwood Lumber from Canada (WT/DS533/R, adopted 26 August 2020), paras 7.131–7.145, finding that the US Department of Commerce’s use of private-land stumpage prices from the US South and Nova Scotia failed to reflect “market conditions prevailing in the country of provision,” and noting that reliance on narrow data windows can capture transitory market conditions and price volatility.”

[iii] US Department of Commerce, Issues and Decision Memorandum for the Final Results of the Countervailing Duty Administrative Review, 2023 (7 August 2025), reaffirming reliance on Nova Scotia private-market stumpage data “in the absence of usable in-country benchmarks.”

[iv] For example, in the Canada–United States Softwood Lumber Dispute, Commerce identified dozens of provincial and federal programs—road offsets, reforestation credits, energy pricing, and others—but panels under NAFTA and the WTO found most to be broadly available or too indirect to constitute a trade-distorting benefit.

[v] See Government of Canada, submission in United States – Softwood Lumber V (WT/DS533/R, paras 7.165–7.171), noting that reliance on a single year of data in a volatile market can distort results; and WTO Panel Report, paras 7.301–7.318, discussing the compounding effect of zeroing on temporary price changes.

[vi] WTO panel findings in Softwood Lumber V (WT/DS533/R, 2020) and subsequent Commerce administrative reviews show that removing zeroing or using in-country benchmarks reduces calculated margins by 30–50%. Palmer’s analysis in Sustainable Forests (2025), together with related assessments from the C.D. Howe Institute (Wyonch, 2022) and UBC Sauder School of Business (Antweiler, 2021), reach broadly similar conclusions.

[vii] For example, the US Chamber of Commerce argued that timber and lumber imports “do not represent a national security risk” in its Comments on Section 232 Investigation of Imports of Timber and Lumber (March 2025).