US Duty Calculations on Canadian Lumber Are Flawed (Part 2)
Visualizing the Effect of “Zeroing” on Anti-Dumping Calculations
Clarification
(Original article follows)
Since I published the article below, Kelly McCloskey, editor of Tree Frog News, referenced it an op-ed. Subsequently, the U.S. Lumber Coalition published a rebuttal.
In response to Mr. McCloskey’s arguments about countervailing and anti-dumping duties, the Coalition simply says:
“As is often the case for those trying to defend Canada’s softwood lumber trade practices, they must rely on arguments that are irrelevant to the current round of antidumping and countervailing duty cases against unfairly traded Canadian lumber imports. In this instance, Mr. McCloskey’s criticism of the zeroing methodology is beside the point, because the US Department of Commerce did not use zeroing when calculating the antidumping duty rates that are currently in place.”
Whoa, wait just a moment there. If the Coalition says the US Department of Commerce (USDoC) didn’t use zeroing this year, aren’t they are in effect admitting to using this patently unfair practice in other years? And in not commenting on any of the other questionable methodological practices identified in Mr. McCloskey’s article (and mine), are they admitting to those too?
Admittedly, I erred in suggesting that the practice of zeroing may have been one of the reasons behind the current extra-high anti-dumping duty rates on Canadian softwood lumber exports to the US (35.47% for Canfor, 9.56% for West Fraser, and 20.52% for all other companies). As it turns out, for the current anti-dumping duties, the USDoC’s analysis determined that zeroing was unnecessary. This decision is outlined in the following comment from the July 24, 2025 memo Issues and Decision Memorandum for the Final Results of the 2023 Administrative Review of the Antidumping Duty Order on Certain Softwood Lumber Products from Canada:
“… Additionally, in the instant review, the results of the meaningful difference test for both mandatory respondents indicate that use of the A-A methodology sufficiently accounts for identified patterns of price differences. Therefore, since we are using the A-A methodology to calculate both mandatory respondents’ weighted-average dumping margins in the final results of this review, arguments from Canfor, the Canadian Parties, and Central Canada regarding the A-T methodology, and the practice of “zeroing” are moot.”
Let’s unpack the above statement. “Dumping” is the practice of exporting at a lower price than the normal price in the exporter’s home country; if a company is found guilty of this, they can be charged an anti-dumping duty.
The US also looks for examples of targeted dumping, that is, selling at a lower price to specific customers or regions, or for specific time periods. The USDoC uses its differential pricing methodology to find and remedy targeting dumping. If the differential pricing methodology’s statistics find evidence that this has occurred, the USDoC can then switch from using its usual “average-to-average methodology” for calculating the dumping margin to using an “average-to-transaction methodology." They may then also opt to use zeroing. (My article below describes these practices in detail).
The USDoC’s analysis found that, for the 2023 Administrative Review, the statistical tests did not meet the criteria for using the average-to-transaction methodology (a prerequisite for using zeroing), so the USDoC used the average-to-average methodology when it calculated the anti-dumping margin.
Nevertheless, the fact that the USDoC did not use zeroing this time still does not mean they do not ever use it, nor does it mean the rest of the calculation process was fair and transparent. Practices such as using constructed values in place of actual sales prices still exist. Unfortunately, it is nearly impossible for the average person to find out exactly how the USDoC does its calculations, because the numbers are not released publicly.
Original Article
On October 14, the US began charging a Section 232 (“national security”) tariff of 10% on imports of Canadian softwood lumber, on top of the duties that were already being charged. Canadian exporters are therefore now paying three surcharges:
The Section 232 tariff of 10%, justified by the argument that imported Canadian softwood lumber and derivative products represent a threat to US national security;
The countervailing duty of 14.63% (for most companies), justified by the argument that Canadian lumber producers buy logs from the government and thus are subsidized; and
The anti-dumping duty of 20.53% (for most companies), justified by the argument that Canadian companies are selling their products into the US at unfairly low prices (= “dumping”).
Regarding the Section 232 tariff, the premise that imports of Canadian 2x4s, sofas and bathroom vanities are somehow a threat to America’s national security is so ludicrous it hardly deserves rebuttal (although you can read a good analysis here). “National security”, as an excuse for tariffs, is as transparent as you can get: the US wishes to protect its own industries and will make up any reason it wants to rationalise it.
The duties, in contrast, have been promoted as being carefully calculated responses to Canadian wrongdoing. On its website, the U.S. Lumber Coalition outlines how the duty investigations by the US Department of Commerce take over a year to complete. Even the duty rates, calculated to the hundred of a percent, give off an aura of precision and accuracy.
Nevertheless, the duty rates are every bit as ridiculous as the new tariffs; this ridiculousness is just more cleverly hidden. For example, the argument that Canadian companies pay less for their logs than American companies do has been shown to be inaccurate: cost comparison studies by analysts such as Forest Economic Advisors show that Canadian mills’ log costs are often higher than those of their US neighbours.
Similarly, the argument that Canadian logs are “dumped” into US markets is based on biased calculations, due to the US Commerce Department’s use of zeroing in its calculations. The following article provides a visual representation of how zeroing creates biased results.
How US Anti-Dumping Calculations Manipulate the Numbers
Dumping, as defined by the World Trade Organization (WTO), is the practice of exporting products at lower prices than the prices in the exporting country.
While the definition of dumping is simple, its calculation is much more difficult. Prices in both the “exporter” and the “exportee” countries fluctuate daily. However, allowing for currency and transportation cost differences, these prices over time tend to average out. For example, in the following hypothetical scenario, some US sales are less expensive than Canadian sales made on the same day, and some are more expensive. Still, the weighted average price of both the US and the Canadian sales is the same: $100. In other words, there is no difference between the average prices, and therefore the dumping margin is zero.
However, if the calculations omit the higher-priced US sales, they can be manipulated to show a positive dumping margin. The following three examples show the difference between dumping margins calculated first without and then with zeroing.
Example 1: Average-to-Average Calculations
Using the “average-to-average calculation method”, the dumping margin is the difference between the weighted average Canadian price and the weighted average US price, divided by the average US price.
In the following chart1, we compare the two averages (highlighted in green) to calculate the total dumping and the dumping margin (highlighted in yellow). As the average Canadian and US prices end up being the same, the dumping ratio is thus zero.
Example 2: Average-to-Transaction Calculations
Anther way to compare prices is to compare individual export transaction prices to the average price in the importing country. This is known as the “average-to-transaction method”.
In the following example, US transaction (individual sale) prices (highlighted in green) are compared to the weighted average Canadian price (also in green). The sum of the price differences (in yellow) is the amount of dumping. As the US sales that are priced lower than the Canadian average are balanced by the sales that are priced higher, the total amount of dumping is still zero.
Looking at a graphical interpretation, we can see how the US transaction prices that are higher than the average Canadian price balance out the US transaction prices that are lower than the Canadian average price.
Example 3: Average-to-Transaction Calculations with Zeroing
In the “average-to-transaction method with zeroing”, US transaction prices are compared to the weighted average Canadian price, and the US sales that are more expensive than the Canadian average price are set to zero (shown in orange). With the higher-priced US sales omitted from the total, the total amount of dumping in the example is now $55, and the dumping margin is 6.1%.
Graphically, we can see that the higher-priced US transactions no longer balance out the lower-priced transactions, because the higher-priced transactions have been “zeroed out” (ignored in the calculations). There is an obvious bias to this calculation method!
In summary, it is clear that the practice of “zeroing” – selectively including some export sales while excluding others from the calculations – biases the results against importers and yields an imposed competitive advantage to the US domestic mills.
“Dumped” Canadian SPF Lumber is NOT the Problem
As the self-declared voice of its lumber and timber company members, the U.S. Lumber Coalition is clear about its intentions: it believes the volume of lumber for sale exceeds the demand, and it is therefore lobbying the US Government to push Canadian imports out of the US marketplace to make more room for supply produced in the US.
However, as analysts such as Russ Taylor and David Elstone have pointed out, the US does not actually have an oversupply of Canadian SPF (spruce-pine-fir) lumber. Instead, it has a growing oversupply of SYP (southern yellow pine) lumber from the US Southeast. While Canadian lumber production has steadily decreased over the past decade, the US has added 9 billion board feet of annual production capacity since 2016. (Nine billion board feet is enough lumber to build roughly 600,000 new single-family homes.) The oversupply problem is with SYP lumber and not with imported SPF – that should be obvious!
SPF and SYP have different properties and differing end uses. Light, structurally stable SPF lumber is favoured for house framing. Dense, strong SYP is good for truss manufacturing and treated products, such as decking. As the supply of SPF in the US has decreased and the supply of SYP has increased, SPF prices have risen to huge premiums relative to SYP prices. Simply put, American builders and consumers like using SPF and want to continue buying it, even though SYP is now cheaper and more plentiful.
What Import Taxes Are Really About
Does the Trump Administration really think that Canadian softwood is a threat to national security? I doubt it. Instead, protectionism is about vote-buying. Lumber is manufactured in rural areas, which tend to vote Republican.
It may appear easier to lobby the government for protection than to invest in market development and innovation. Why grow the market pie when you can simply take your neighbour’s slice? The problem with that reasoning is that SPF and SYP are actually two different types of pie. Forcing builders who enjoy maple-syrup pie (i.e., SFP) to eat pecan pie (i.e., SYP) won’t work if the maple-syrup pie eaters have a nut allergy.
A better approach would be to grow the market for wood products. This will be the topic of future Sustainable Forests articles.
Tables adapted from “A One-Two Punch on Zeroing: US–Zeroing (EC) and US–Zeroing (Japan)” by Thomas J. Prusa and Edwin Vermulst, World Trade Review (2009), 8: 1, 187–241.