Dillinger France S.A. v. United States, 981 F.3d 1318

Background

Commerce investigated antidumping for certain carbon and alloy steel cut‑to‑length plate from France and selected Dillinger France S.A. as a mandatory respondent.
Dillinger produces “prime” (warranty) and “non‑prime” (no warranty) plate; its accounting allocates non‑prime cost based on likely selling price, offsetting prime cost.
Commerce adjusted Dillinger’s reported costs by reducing non‑prime cost (using recorded selling values) and reallocating the difference to prime plate; Commerce assigned a 6.15% antidumping margin.
The Trade Court largely sustained Commerce’s determination and remand results; Dillinger appealed to the Federal Circuit challenging (1) Commerce’s cost allocation reliance on Dillinger’s books, (2) use of the average‑to‑transaction method (Cohen’s d + ratio test), and (3) decision denying a level‑of‑trade adjustment.
The Federal Circuit: vacated and remanded Commerce’s reliance on Dillinger’s books for cost allocation (records did not reasonably reflect production costs), affirmed Commerce’s use of the Cohen’s d/ratio tests and the average‑to‑transaction method, and affirmed Commerce’s denial of a level‑of‑trade adjustment.

Issues

Issue	Plaintiff's Argument (Dillinger)	Defendant's Argument (Commerce)	Held
Whether Commerce properly relied on Dillinger’s books and records to allocate costs between prime and non‑prime plate	Commerce improperly relied on records that allocate costs by likely selling price rather than actual production costs	§1677b(f) requires reliance on exporter records if kept per GAAP and they reasonably reflect production/sales costs	Vacated and remanded — Dillinger’s records did not reasonably reflect production costs; Commerce must determine actual costs on remand
Whether Commerce permissibly used the average‑to‑transaction method using Cohen’s d and a ratio test to find a “pattern” of price differences	The Cohen’s d/ratio methodology improperly aggregates across categories and ignores the statutory word “pattern”; products are custom so average‑to‑average should apply	Statute is silent on how to identify a pattern; Commerce’s interpretation and tests are a reasonable construction under Chevron; WTO Appellate Body not binding	Affirmed — Commerce’s use of Cohen’s d and ratio test was reasonable and supported application of average‑to‑transaction
Whether Commerce erred in treating factory and affiliated service center sales as the same level of trade (denying an adjustment)	Inventory maintenance and service‑center processing (cutting/sawing) produce substantial selling‑activity differences warranting a separate level of trade	Inventory maintenance alone is not a substantial selling‑activity difference; cutting/sawing are processing (production) costs, not selling functions	Affirmed — substantial evidence supports Commerce’s finding that selling functions did not differ enough to warrant an adjustment

Key Cases Cited

IPSCO, Inc. v. United States, 965 F.2d 1056 (Fed. Cir. 1992) (rejects allocating costs based on relative prices of products rather than production costs)
Thai Pineapple Pub. Co. v. United States, 187 F.3d 1362 (Fed. Cir. 1999) (Commerce may rely on producer records that reasonably reflect production costs)
Thai Plastic Bags Indus. Co. v. United States, 746 F.3d 1358 (Fed. Cir. 2014) (explains §1677b(f) two‑part test for reliance on exporter/producer records)
PSC VSMPO‑Avisma Corp. v. United States, 688 F.3d 751 (Fed. Cir. 2012) (upholds use of purchase price to determine input cost where appropriate)
U.S. Steel Corp. v. United States, 621 F.3d 1351 (Fed. Cir. 2010) (standard of review for Trade Court/Commerce decisions)
Micron Tech., Inc. v. United States, 243 F.3d 1301 (Fed. Cir. 2001) (purpose and mechanics of level‑of‑trade adjustment)
Corus Staal BV v. Dep’t of Commerce, 395 F.3d 1343 (Fed. Cir. 2005) (WTO decisions are not binding on U.S. courts)
Timken Co. v. United States, 354 F.3d 1334 (Fed. Cir. 2004) (addresses deference to agency interpretations and related precedent)