AD Petitioner Says Commerce Doesn't Need to Pick 'Perfect' Methodology to ID Masked Dumping

The Commerce Department doesn't have to "undertake the impossible task of choosing a perfect or near-perfect methodology" when rooting out "masked" dumping, petitioner Welspun Tubular said in a Nov. 6 reply brief at the U.S. Court of Appeals for the Federal Circuit. Welspun added that Commerce didn't have to "follow the teachings of statisticians on the Cohen's d test simply because Commerce has decided in its expert opinion to rely on the Cohen's d formula and the effect size coefficient corresponding to a grossly perceptible difference to identify whether price differences between customers, regions, or time periods are significant" (Stupp Corp. v. United States, Fed. Cir. # 23-1663).

Welspun added that Commerce's remand "fully addressed the concerns" previously raised by the Federal Circuit in July 2021 regarding the use of the d test. The appellate court took issue with the fact that Commerce used the test with data sets that didn't adhere to basic statistical assumptions, including normal distribution, equal variances and large sample sizes (see 2107150032).

Welspun said that on remand the agency explained why its use of the test is reasonable despite not satisfying these statistical assumptions. "Applying the Cohen’s d test in this context does not produce an upward bias with more passing sales and higher dumping margins," the brief said. "As the CIT recognized, the differential pricing analysis operates as an integrated whole which looks at the frequency and impact of effect size to detect targeted dumping -- not the effect size alone."

The brief echoed the U.S.'s arguments, which it made earlier this month (see 2311060058). For instance, Welspun argued that the fact that sales may pass the d test since "a small price difference is given greater significance when the variances are small does not render the result inaccurate or erroneous." The d test "assesses the significance of price differences based on the actual variances in the respondent’s sales prices," the brief said. The petitioner added that there is no evidence showing that Commerce's use of the test was "distortive or unreasonable."