In Oral Arguments, Government, Exporter Argue Standard for Challenging Commerce Data Selection

“I believe as we go through the questions, there’s an underlying theme and challenge that we’re all trying to grapple with,” Court of International Trade Judge Timothy Reif said. “And that is that it’s difficult in general for Commerce to come up with an exact match, right, for a data set related to imports.”

He said that exporter Jiangsu Zhongji Lamination Materials' preferred input price benchmark was a composite of [Global Trade Atlas] data and data from Commodities Research Unit (CRU) that didn’t actually include any imports of the input in question, aluminum alloy of grade 1050. Commerce, meanwhile, decided to rely on Trade Data Monitor data, though Zhongji argued that it was overly broad because it included many other aluminum alloy grades not used in the manufacture of aluminum foil (see 2309190058).

Given that both of Commerce’s options are imperfect and that it tried its best to product-match, Reif asked why the court would hold that Commerce must replace its data with Zhongji’s.

Zhongji’s counsel, Cleo Lee, argued that Commerce has a duty to calculate its antidumping duty rates as accurately as possible. The aluminum inputs covered by Commerce’s data are destined for use in everything from airplanes to ships, she said.

In turn, DOJ attorney Sosun Bae agreed with Lee regarding Commerce’s duty to reach accurate margins. But she said that the ultimate question is whether Commerce made a reasonable decision after it “considered the relative strengths and weaknesses,” arguing it had done so “even if some evidence might lean toward using the CRU combined with the GTA data.”

Reif said Zhongji was making the argument that the regulation governing AD reviews calls for Commerce to use the “better and closer match” when selecting data. He asked the DOJ whether the court should adopt the concept that “a better and closer match” fulfills the regulation’s purpose.

Without conceding that the CRU data was better, Bae said that the court’s role was to review Commerce’s decision and determine whether or not it was supported by substantial evidence. In this case, it was, she said.

Bae called the CRU data “cherry-picked,” as it came from only a few countries. Those countries didn’t contain all the major exporters of 1050 alloy, according to an expert, she said.

She also said that Zhongji’s proposed data had significant flaws. For example, Bay said, it was the result of a “complicated formula” that only created an estimate of world market prices, whereas Commerce’s data actually reflected those prices. Zhongji also proposed using CRU data alone, without combining it with GTA data; but the CRU information wasn’t tier-two data, making it too narrow, she said.

And Zhongji didn’t actually purchase 1050 alloy during the period of inquiry, she noted, so Commerce’s data “actually covered the actual purchases made by respondents.”

Lee disagreed that the mandatory respondent’s data contained only estimates, saying that it was instead composed of “actual prices.” She also said that the CRU data could be tier-two data, as “Commerce has no practice of creating a bright line of the minimum countries you need to cover in a tier-two benchmark.”

And she argued that two pieces of evidence showed that the 1050 alloy was the closest chemical match to Zhongji’s inputs, contesting the government’s point that the 1050 alloy contained a “quite different aluminum content” than Zhongji’s.

Reif pointed out that these were the sorts of points Zhongji had already raised in administrative proceedings. He agreed that plaintiffs face a “high bar” when challenging Commerce’s selection of data.

The judge then moved on to discuss “the infamous 2010 CBRE [Coldwell Banker Richard Ellis] report” that Commerce put on the record prior to knowing when Zhongji purchased land in China.

Bae said that she wasn’t “quite sure” about the relevance of Commerce’s timing. The department only subsequently made the decision that the 2010 report was the most contemporaneous to Zhongji’s land purchases, she said.

In response, Lee argued that Commerce couldn’t have known the dates of those land purchases, meaning the inclusion of the report in the record “indicate[d] a foregone conclusion” on the part of Commerce to use that report.

She also said that other cases show Commerce doesn’t have a “common practice” of relying on contemporaneity when selecting such benchmarks, as the department claimed.

Bae said they acknowledged that Commerce’s practice had not been “100% consistent,” but claimed that it had done so regularly since 2008 and that it specifically found that it should use contemporaneous land purchase dates in another case involving aluminum foil.