CIT Says Use of Full Population Doesn't Resolve CAFC Concerns in Use of 'd' Test

The judge also denied Koehler's challenge to Commerce's rejection of its exhibits pertaining to the normal distribution of data as part of the Cohen's d test for "harmless error."

Katzmann remanded the coding of the static sensitivity product characteristic, classification of Koehler's accrued interest expenses as a cost of production and the use of the Cohen's d test to root out "masked" dumping, staying the case until the U.S. Court of Appeals for the Federal Circuit issues a decision in Stupp Corp. v. U.S.

Cohen's d Test

As is its practice, Commerce used Cohen's d test in the thermal paper investigation to detect masked dumping by creating a test group and comparison groups from Koehler's sales and measuring the extent of the difference in the mean prices between the groups.

The Federal Circuit has previously called Commerce's general use of the test into question where the "data groups being compared are small, are not normally distributed, and have disparate variances." In response to the ruling in Stupp v. U.S., Commerce explained that its data need not meet these requirements when it uses the entire population of data. The Federal Circuit is now considering that explanation and, in the meantime, the U.S. said CIT doesn't have to address the test's use in the thermal paper investigation because the Federal Circuit case is part of ongoing litigation, and said Commerce's use of the test without these basic statistical assumptions is justified, given its use of the full population of data.

Katzmann rejected both points out of hand. On the former, the judge said that while the opinion is part of ongoing litigation, Commerce didn't account for the fact that it is a "published, precedential decision with holdings that clarify (and indeed shape) the background law against which Commerce's actions are to be found either reasonable or unreasonable."

As for the "full population" defense, Katzmann held that the Federal Circuit explicitly said Commerce's use of the test raises concerns where the "test groups and the comparison groups" are not "normally distributed, of sufficient size, and of roughly equal variances." Whether the sales used represent the entirety of an exporter's U.S. sales is seemingly "irrelevant to this inquiry," the judge added.

“The logical link between these two propositions is not so reasonably discernable as to obviate the need for explanation," the opinion said. Katzmann stayed the case pending the appellate court's ruling.

Related to the test, Koehler argued that Commerce unlawfully rejected four exhibits the company attempted to submit that purportedly show that the d test violated the precondition of normality. Katzmann said Koehler failed to show that the rejection of the exhibits was prejudicial because the company claimed the lack of normality in Commerce's data sets can be shown by other evidence. The principle of "harmless error" applied, the court said.

Koehler's Blue4est Product

Koehler also contested Commerce's clarification that the company's Blue4est paper fits within the scope of the investigation. The AD order's scope covers thermal active coating, typically made of "sensitizer, dye, and coreactant, and/or like materials," on one or both sides of the paper.

Kohler argued that the Blue4est paper's thermal active coating isn't made exclusively of sensitizer, dye and coreactant and should be excluded. The court noted Koehler's own submissions showed that the paper is "coated with a layer of bubbles that collapse when exposed to heat," meaning it has a "thermal active" coating, the judge said. Koehler failed to show that its layer of bubbles does not constitute a "like material" to "sensitzers, dyes, and co-reactants."

Kohler said the "concept" of thermal active coating is limited to mechanisms through which a coating undergoes a chemical reaction to make a color, and that this concept does not include Blue4est's mechanism. Katzmann said Koehler only succeeded in showing that its product differs from ordinary thermal paper via its "bubble-collapsing mechanism," though this distinction does not control the question of what defines a thermal active coating. Only Commerce has the power to clarify this, the court said in deferring to the agency.

Kohler argued Commerce's final scope interpretation is irreconcilable with the original petition and other sources describing the order's scope. Katzmann found that Commerce has "broad discretion to define and clarify the scope of an" AD investigation in a way that reflects the petition's intent. This means that the agency can "freely depart from the Petition's initial language and the ITC Description" in setting the scope.

Finally, Kohler claimed that inclusion of the Blue4est paper unlawfully expanded the scope of the investigation. The court found Commerce's determination that "thermal active coating" applies to coatings that evidence shows undergo a physical reaction upon exposure to heat is in "no way" an expansion of the original scope language. Commerce found that the original scope language applies to the Blue4est paper based on record evidence, the court noted.

Dynamic Sensitivity Test Data and Static Sensitivity Data

The petitioners in the investigation, Appvion and Domtar Corp., challenged Commerce's decision not to assign the same control number (CONNUM) to two of Koehler's thermal paper products, dubbed Products "A" and "B" in the opinion, due to the exporter's differing sensitivity characteristics for the goods. The petitioners claimed that Koehler manipulated its data to get Commerce to classify the products under two different CONNUMs.

Katzmann said the claim would be "more persuasive" if Commerce were reviewing the data for "publication in a scientific journal." But here, the inquiry is whether the agency's decision was backed by substantial evidence, not subject to the "highest possible standard of reliability."

While the petitioners said their claim was not a challenge to how Commerce weighed the evidence because the agency didn't actually give a reason for how it weighed the evidence, the court disagreed. Appvion and Domtar didn't agree with "Commerce's choice to look to the frequency of a test result as a means of determining whether to rely on that result," but that disagreement "does not negate Commerce's reasoned explanation for why it chose to rely on Koehler's submission," the opinion said.

Appvion and Domtar also contested Commerce's conclusion that Koehler's submissions of static sensitivity data were complete, alleging that the exporter used a flawed testing methodology. The specific alleged flaw was that Koehler didn't conduct tests at "sufficiently high temperatures" to find the point at which the tested goods reached their maximum optical density unit according to Koehler's own testing device.

While Katzmann said that the court will not reweigh evidence, he said he does look at whether the decision of completeness is backed by reasonable evidence. Unable "to discern such evidence on the record" at this stage, the court remanded the finding that "Koehler's submissions of static sensitivity data were complete."

Price Adjustments for Koehler's Rebates and Accrued Interest as a COP

In the investigation, Commerce made downward adjustments to Koehler's home market prices to reflect rebates the company applied to certain home market sales. This adjustment lowered Koheler's AD rate, prompting a challenge from the petitioners, who said the adjustment reflected a break from the agency's established practice. Appvion and Domtar said, to be in line with past practice, the agency "must show at the very least that the terms and conditions of the price adjustment were established and known to the customer at the time of sale."

However, Katzmann said this misstates the legal background of the decision. The petitioners' framing of the issue puts "Commerce’s enumeration of the five factors that Commerce may consult in determining entitlement to a price adjustment on the same footing as Commerce’s apparent statement that the agency generally won't consider a price adjustment absent prior customer knowledge of a rebate.” None of the agency's regulations directly say Commerce considers customer knowledge of a rebate to be a necessary condition for "entitlement to a price adjustment."

The agency "thoroughly explained" its rationale for giving Koehler a price adjustment, the court said. Even if this were an established practice, Commerce adequately explained why a departure from it is warranted, the opinion said.

Lastly, the petitioners challenged Commerce's decision to add Koehler's interest expenses on unpaid AD incurred during the period of investigation to the cost of production calculation. Due to this addition, the agency excluded certain home market sales that were made at lower prices than this cost of production from the company's normal value calculation.

The petitioners said Commerce should've included Koehler's interest expenses as indirect selling expenses to be deducted from the constructed export price. Katzmann said that the agency's only response, which said that Koehler's expenses linked with commercial activity in the U.S. don't include any financial interest expenses, "assumes a conclusion as to the proper classification of the expenses."

Instead of rebutting the petitioners' argument, "Commerce instead made a flat declaration to the effect of 'not so,'" the judge said. "This is not the kernel of an argument from which 'the agency's decisional path' is 'reasonably discernible'" and is in fact "no argument at all," the court said in remanding the issue.

(Matra Americas v. United States, Slip Op. 24-14, CIT # 21-00632, dated 02/08/24; Judge: Gary Katzmann; Attorneys: R. Will Planert of Morris Manning for plaintiffs Matra Americas and Matra Atlantic GmbH; Thomas Trendl of Steptoe & Johnson for plaintiff-intervenors Koehler Paper SE and Koehler Oberkirch GmbH; Emma Bond for defendant U.S. government; Daniel Schneiderman of King & Spalding for defendant-intervenors Domtar Corp. and Appvion)