CIT Says Commerce Still Creating New Non-Record Information to Calculate AFA in Solar Cell Case

CIT Judge Claire Kelly again remanded the department’s selection of its adverse facts available rate, which she had previously found was the result of Commerce “manipulating evidence on the record” (see 2405090045).

The department refused, after an initial remand issued May 1, 2024, to change up its selection method (see 2408300020). Kelly explained that Commerce “calculated a consumption ratio for three categories of inputs using simple averages of the highest ratio for each item in each category, and multiplied each input ratio by the multiplicative inverse of the simple average input ratio.” It did so to develop a formula for choosing an AFA rate after concluding that simply using "the highest consumption quantity” wasn’t “sufficiently adverse,” she said.

Despite the first remand, Commerce has still failed to explain how its formula “is logically related to record information,” Kelly said.

She said that the first few steps of the department’s calculation -- deriving a consumption rate for each of mandatory respondent Risen’s inputs, then separating these inputs into three categories and averaging each -- might be unreasonable, but still at least were based on information on the record.

But Commerce then “divided the number 1 by the simple average input ratio to create a multiplicative inverse and then multiplied each input ratio by the multiplicative inverse of the simple average input ratio,” she said. By creating this “multiplicative inverse,” the department was in essence creating new facts that weren’t on the record or derived from it, she ruled, which was contrary to law.

“It is possible that Commerce’s use of the multiplicative inverse may be its attempt at adding a ‘built-in increase intended as a deterrent to non-compliance,’” she said, citing the 2000 U.S. Court of Appeals for the Federal Circuit case F.lli De Cecco v. United States. “If Commerce’s use of the multiplicative inverse is meant to build in an increase as contemplated by De Cecco, Commerce must explain that it is doing so and further explain why that built in increase should be considered a derivation from record evidence, and therefore in accordance with the statute.”

If the department wants to rely on new information it has derived from the record, it must provide a reasonable explanation for doing so, she said.

It also must still ensure its resultant determination is supported by substantial evidence, which it has failed to do up to this point, she said. She held that Commerce hasn’t explained why a 4% AFA rate based on record information was inadequately adverse or why the 15% AFA rate it calculated instead was reasonable.

She also remanded for the department’s use of Romanian rather than Malaysian surrogate data to value respondents’ solar glass input.

For the review, the department selected Malaysia as its primary surrogate. But it valued solar glass using Romanian data, citing a concern that the Malaysian data, which was reported in square meters, wasn’t as specific to the mandatory respondents’ products, which Commerce had respondents report in kilograms. In her first remand order, Kelly told the department that it hadn’t properly addressed contradictory arguments raised by exporters led by Jinko Solar that the Romanian data itself had flaws.

Kelly said in her recent second remand order that Commerce still hadn’t addressed these contradictory arguments. The department still needs to properly consider the fact that the Romanian tariff schedule excludes glass made with an “absorbent layer,” among other exceptions, from products covered by its heading, she said.

The term “absorbent layer” isn’t defined on the record, she noted. Commerce argued that, based on plain meaning and the other exceptions provided in the chapter heading, glass with an absorbent layer “takes in light without releasing it.” But the department’s argument only looks to some of those other exceptions, ignoring the heading’s exceptions for glass used products such as in motor vehicles or aircraft, she said.

“It is unclear how Commerce can contend that all of the exemplars support the interpretation of glass with an absorbent layer as light limiting, when it does not address all of the exemplars,” Kelly said.

She also said Commerce had failed to deal with the exporters’ arguments that the Malaysian data, which reported solar glass purchases in square meters, was usable. The department instructed the review’s mandatory respondents to report their purchases of solar glass by weight, even though they themselves recorded their purchases by piece in their books, she said. It didn’t explain why it couldn’t convert the exporters’ raw per-piece data to square meters and use the Malaysian data instead of resorting to a separate surrogate.

She did hold that Commerce had adequately explained why it couldn’t convert the exporters’ reported per-kilogram numbers to square meters.

(Jinko Solar Import and Export Co. v. United States, Slip Op. 25-67, CIT Consol. # 22-00219, dated 05/30/25; Judge: Claire Kelly; Attorneys: Ned Marshak of Grunfeld Desiderio for plaintiffs led by Jinko Solar Import and Export; Robert Gosselink of Trade Pacific for consolidated plaintiffs led by Trina Solar; Jeffrey Grimson of Mowry & Grimson for consolidated plaintiffs led by JA Solar Technology Yangzhou Co.; Craig Lewis of Hogan Lovells for plaintiff-intervenor BYD (Shangluo) Industrial Co.; Gregory Menegaz of deKieffer & Horgan for consolidated plaintiff-intervenor Risen Energy; Kara Westercamp for defendant U.S. government; Timothy Brightbill of Wiley Rein for defendant-intervenor American Alliance for Solar Manufacturing)