Commerce Says 'Publicly Available' SV Data Means Data on the Public Record

Commerce said it was proper to use the Freightos data since it was on the public record and that it defines the phrase "publicly available information" to mean information that's on the public record. In the review, the agency refused to use data from the International Air Transport Association since parts of the data weren't on the public record. The trade court questioned this decision and the general policy treating business proprietary information as being publicly available.

On remand, Commerce said the information it uses to value factors of production (FOP) in a particular segment of a proceeding "must be on the record of that proceeding segment." Simply listing the source of the data but treating the data as business proprietary information, "even if the data are from a public source, does not allow the public to see exactly what Commerce saw on its record and relied upon to make its determination," the remand results said.

The agency offered various defenses of this policy, including the fact that it boosts accuracy "if all interested parties, not just parties with access to BPI, could comment on the excerpts and provide rebuttal factual information." Using confidential information "shrinks the pool of parties that are able to file comments on the merits of the prospective" surrogate values (SVs), the remand said. In addition, using the confidential information limits the ability of parties to fully engage in debate on the best surrogates, since parties not represented by counsel or represented by counsel without access to the private record "could not fully evaluate the publicly available information for valuing FOPs that was placed on the confidential record."

Using the business proprietary information also doesn't promote predictability because counsel to parties in other AD cases involving non-market economies can't access this information, the agency said. As a result, Commerce continued using the Freightos data to value air freight.

CIT also sent back Commerce's use of partial AFA, finding that the agency manipulated data to use its adverse inference. In the review, Commerce used partial AFA since Risen's unaffiliated solar cell supplier failed to provide their factors of production data. In using AFA, Commerce derived a multiplier from the reported factors of production data, which it multiplied by all the reported quantities.

The court asked Commerce to explain how its formula works and to explain its averaging groups. The agency did just that on remand, explaining that for each input with missing data, it calculated a ratio for each monocrystaline solar module control number by dividing the per-unit consumption quantity by the highest per-unit consumption quantity reported for the input or any solar module. The agency then averaged all the control number-specific ratios it calculated for that input to get a single ratio for all the inputs that were used to make solar modules. Two more simple averages were then calculated -- one of all the input ratios for all the inputs used to make solar cells and another for all the material inputs used in packing solar modules. Commerce then multiplied each reported per-unit consumption quantity by the multiplicative inverse of the simple average input ratio.

The agency said it used three separate average input ratios to calculate separate adverse inference adjustments for the solar cells, solar modules and packing instead of using input-specific ratios to get a separate adverse inference adjustment for each input. It did so to get a more adverse rate for the exporters to get "a greater incentive for cooperation," Commerce said.

Commerce defended its decision not to use a simple average of all input ratios to get one adverse adjustment multiplier since "it was more accurate to calculate an adverse inference multiplier that bore some relationship to the inputs to which it applied the multiplier." The agency added that it wasn't "appropriate to include input ratios for packing together with input ratios for solar cells and solar modules in the pool of ratios from which it calculated a simple average ratio" and that its methodology boosts accuracy since it "only applied the adverse inference multipliers with respect to the missing information.”

Lastly, Commerce defended its valuation of solar glass using Romanian data despite picking Malaysia as the main surrogate country. The agency said the Malaysian data was unusable since it was reported in a different unit and neither Risen nor exporter Jinko Solar Import and Export Co. provided an adequate means to convert the units of the Malaysian data. Both conversion methods offered by the companies failed to show that the thickness of the glass consumed by the companies matched the thickness of the glass imported into Malaysia.