MID CONTINENT NAIL CORPORATION, Plaintiff-Appellant v. UNITED STATES, DUBAI WIRE FZE, ITOCHU BUILDING PRODUCTS CO., INC., Defendants PRECISION FASTENERS, LLC, Defendant-Appellee
In 2012, the Department of Commerce issued a final determination in an antidumping investigation of certain steel nails from the United Arab Emirates (“UAE”) finding that Precision Fasteners, LLC had engaged in targeted dumping and imposed a duty. In calculating Precision's dumping margin, Commerce declined to apply a regulation limiting the use of the average-to-transaction methodology to non-targeted sales because the agency asserted that the regulation had been withdrawn in 2008. See 19 C.F.R. § 351.414(f)(2) (2008).
The Court of International Trade (“Trade Court”) held that Commerce had violated the Administrative Procedure Act (“APA”) by withdrawing the regulation without providing notice and opportunity for comment. On remand, Commerce redetermined Precision's duty by applying the withdrawn regulation and found that no duty was owing. The Trade Court affirmed. We hold that Commerce violated the requirements of the APA in withdrawing the regulation, leaving the regulation in force; that its violation of the APA was not harmless; and that the agency did not err in applying the regulation on remand. We therefore affirm the final judgment of the Trade Court.
In 2011, appellant Mid Continent Nail Corp. filed a petition with Commerce alleging that “imports of certain steel nails from the UAE ․ [were being] sold in the United States at less than fair value, ․ and that such imports [were] materially injuring, or threatening material injury to, an industry in the United States.” Certain Steel Nails From the United Arab Emirates: Initiation of Antidumping Duty Investigation, 76 Fed. Reg. 23,559, 23,560 (Apr. 27, 2011). Commerce initiated an antidumping investigation during which it determined that appellee Precision was among the mandatory respondents, i.e., an importer whose dumping rate would be individually determined in the course of the investigation.1 See Certain Steel Nails From the United Arab Emirates: Preliminary Determination, 76 Fed. Reg. 68,129 (Nov. 3, 2011). In 2012, Commerce issued an antidumping duty order imposing a 2.51 percent duty on Precision. See Certain Steel Nails From the United Arab Emirates: Final Determination, 77 Fed. Reg. 17,029, 17,031–32 (Mar. 23, 2012); Certain Steel Nails from the United Arab Emirates: Amended Final Determination, 77 Fed. Reg. 27,421, 27,422 (May 10, 2012).
Commerce found that Precision had engaged in “targeted dumping” because Precision's sales reflected a “pattern of export prices ․ that differ[ed] significantly among certain customers, regions, and time periods.” 77 Fed. Reg. at 17,031; see also 19 U.S.C. § 1677f-1(d)(1)(B)(i); U.S. Steel Corp. v. United States, 621 F.3d 1351, 1359 (Fed. Cir. 2010). And, central to this appeal, the agency proceeded to calculate Precision's dumping margin by applying the average-to-transaction methodology to all U.S. sales reported by Precision, irrespective of whether the agency had deemed a sale to be targeted or not. See 77 Fed. Reg. at 17,031.
The average-to-transaction methodology is one of the three methods that Commerce may use in an investigation to calculate dumping margins in accordance with the Tariff Act of 1930, as amended by the Uruguay Round Agreements Act (URAA), Pub. L. No. 103–465, 108 Stat. 4809 (1994). The statute provides that, in general, Commerce “shall determine whether ․ subject merchandise is being sold in the United States at less than fair value” by either: (1) “comparing the weighted average of the normal values to the weighted average of the export prices (and constructed export prices) for comparable merchandise”; or (2) “comparing the normal values of individual transactions to the export prices (or constructed export prices) of individual transactions for comparable merchandise.” 19 U.S.C. § 1677f-1(d)(1)(A)(i)–(ii). These two methods are respectively known as the “average-to-average” and “transaction-to-transaction” methodologies.
The statute permits Commerce to use a third method—the average-to-transaction methodology—if certain conditions are met. The average-to-transaction methodology “compar[es] the weighted average of the normal values to the export prices (or constructed export prices) of individual transactions for comparable merchandise.” Id. § 1677f-1(d)(1)(B). To calculate dumping margins using the average-transaction methodology, however, Commerce must find “a pattern of export prices (or constructed export prices) for comparable merchandise that differ significantly among purchasers, regions, or periods of time,” (i.e., targeted dumping) and explain “why such differences cannot be taken into account using” the first two methods. Id. § 1677f-1(d)(1)(B)(i)–(ii). In other words, Commerce must first conclude that a respondent is engaged in targeted dumping and explain why the other two statutory methodologies fail to sufficiently account for it. See U.S. Steel, 621 F.3d at 1358–59.
In calculating dumping margins using the average-to-transaction methodology, Commerce has “historically” used a practice known as “zeroing” in which “negative dumping margins (i.e., margins of sales of merchandise sold at nondumped prices) are given a value of zero and only positive dumping margins (i.e., margins for sales of merchandise sold at dumped prices) are aggregated.” Union Steel v. United States, 713 F.3d 1101, 1104 (Fed. Cir. 2013). As a result, “dumping margins for sales below normal value are not offset by ‘negative dumping margins' for those sales made above normal value.” Corus Staal BV v. United States, 502 F.3d 1370, 1372 (Fed. Cir. 2007). This lack of offsetting leads to higher dumping margins when the average-to-transaction methodology is used, which has made calculation of margins using this methodology “controversial.” See Union Steel, 713 F.3d at 1104.
Shortly after the enactment of the URAA, Commerce promulgated a regulation through notice-and-comment rulemaking restricting the agency's use of the average-to-transaction methodology. This regulation—known as the “Limiting Regulation”—provided that even in cases meeting the statutory criteria for applying the average-to-transaction methodology, the agency would “normally ․ limit [its] application ․ to those sales that constitute targeted dumping,” as opposed to applying the average-to-transaction methodology to all of a respondent's sales. See 19 C.F.R. § 351.414(f)(2) (2008); see also Antidumping Duties; Countervailing Duties, Final Rule, 62 Fed. Reg. 27,296, 27,375 (May 19, 1997).
In 2008, however, Commerce withdrew the Limiting Regulation, along with several other regulations governing the agency's handling of targeted dumping allegations. See Withdrawal of the Regulatory Provisions Governing Targeted Dumping in Antidumping Duty Investigations, Interim Final Rule, 73 Fed. Reg. 74,930, 74,931 (Dec. 10, 2008) [hereinafter Withdrawal Notice]. The agency stated that it had originally promulgated the regulations “without the benefit of any experience on the issue of targeted dumping,” and that the regulations “may have established thresholds or other criteria that ․ prevented the use of [the average-to-transaction] methodology to unmask dumping, contrary to the [c]ongressional intent.” Id. Commerce noted that withdrawal would allow the agency to gain “additional experience” with targeted dumping through “case-by-case adjudication.” Id.
Commerce acknowledged in Withdrawal Notice that repeal of the targeted dumping regulations was subject to “the requirement to provide prior notice and opportunity for public comment, pursuant to ․ 5 U.S.C. § 553(b)(B),” but expressly “waive[d] the requirement” by invoking the APA's “good cause” exception to notice-and-comment rulemaking. 73 Fed. Reg. at 74,931.
In finding good cause, Commerce explained that notice-and-comment rulemaking was “impracticable and contrary to the public interest” because the rescinded regulations were “applicable to ongoing antidumping investigations” and that “immediate revocation [was] necessary to ensure the proper and efficient operation of the antidumping law[s].” Id. At no point in Withdrawal Notice did Commerce refer to any prior notices proposing to withdraw the Limiting Regulation, or otherwise suggest that the agency had provided adequate notice and opportunity for comment under the APA.
In calculating Precision's dumping margin three years later in this proceeding, Commerce applied the average-to-transaction methodology, having found both “a pattern of export prices ․ that differ[ed] significantly among customers, regions, or by time-period,” and that applying the “average-to-average methodology mask[ed] differences in the patterns of prices between the targeted and non-targeted groups.” 77 Fed. Reg. at 17,031. In this appeal, no party has challenged Commerce's determination that the statutory criteria for applying the average-to-transaction methodology were met. What the parties dispute is the agency's decision to apply the average-to-transaction methodology not just to “those sales that constitute[d] targeted dumping,” as the Limiting Regulation had previously provided, but “to all U.S. sales reported by ․ Precision.” See id. (emphasis added).
Precision challenged Commerce's final determination in the Trade Court. See Mid Continent Nail Corp. v. United States (Mid Continent I), 999 F. Supp. 2d 1307, 1309–10 (Ct. Int'l Trade 2014). In relevant part, Precision argued that Commerce was required to apply the Limiting Regulation to calculate Precision's dumping margin because the agency's repeal of the regulation in “Withdrawal Notice was ineffective and contrary to law,” as it had “occurred outside the basic procedural framework required by Congress under the [APA].” Id. at 1319–20. According to Precision, had the agency applied the Limiting Regulation, application of the average-to-transaction methodology to all of Precision's domestic sales would not have been “justif[ied]” because the agency had “only found evidence of targeting for less than one percent” of Precision's U.S. sales, the exact scenario that had concerned Commerce when it adopted the Limiting Regulation in the first place. Id. at 1319.2
The Trade Court agreed that Commerce's withdrawal of the Limiting Regulation violated the APA. After concluding that withdrawal of the regulation was subject to notice-and-comment rulemaking, the court rejected the argument that the agency had provided adequate notice and opportunity for comment through two earlier Federal Register notices because those notices had not proposed to repeal the regulation. See id. at 1322. The court also rejected Commerce's invocation of good cause and found that the agency's procedural default was not excusable as harmless error. See id. Accordingly, the Trade Court remanded Commerce's final determination and instructed the agency to “redetermine [Precision's] dumping margin[ ] by applying the Limiting Regulation.” Id. at 1323.
On remand, Commerce applied the Limiting Regulation as ordered by the Trade Court. As the regulation provided that Commerce would “normally” not apply the average-to-transaction methodology to all sales, see 19 C.F.R. § 351.414(f)(2) (2008), the agency concluded that application of the average-to-transaction methodology to all of Precision's sales was unwarranted because “the record does not contain evidence to suggest that this normal limitation should not be applied.” J.A. 89. As a consequence of limiting the average-to-transaction methodology to only targeted sales, Commerce found that Precision's dumping margin was “de minimis,” and therefore imposed a duty of 0.00 percent. Id.
Mid Continent appealed Commerce's remand redetermination to the Trade Court, arguing that the agency had misapplied the Limiting Regulation. See Mid Continent Nail Corp. v. United States (Mid Continent II), 113 F. Supp. 3d 1318, 1326 (Ct. Int'l Trade 2015). The court rejected Mid Continent's argument and affirmed Commerce's remand redetermination. See id. at 1327–28, 1331. Mid Continent then filed this appeal, which Commerce has not joined. We have jurisdiction under 28 U.S.C. § 1295(a)(5).
During the pendency of the Trade Court proceedings, and in light of the court's ruling that Withdrawal Notice was ineffective to repeal the Limiting Regulation,3 Commerce in 2013 initiated a new proceeding to accomplish the repeal. The agency published a Notice of Proposed Rulemaking (“NPRM”) in which it sought comments on a proposal “not to apply ․ the previously withdrawn regulatory provisions governing targeted dumping.” Non-Application of Previously Withdrawn Regulatory Provisions Governing Targeted Dumping in Antidumping Duty Investigations, Proposed Rule, 78 Fed. Reg. 60,240, 60,240 (Oct. 1, 2013). In 2014, Commerce issued a final rule making withdrawal of the regulations effective May 22, 2014. See 79 Fed. Reg. 22,371 (Apr. 22, 2014). No party to this appeal has challenged the 2014 withdrawal, or contended that it should be applied retroactively. Accordingly, this case solely addresses whether the withdrawn regulations were in effect during the period between December 10, 2008, and May 22, 2014.
We review the Trade Court's decision to uphold Commerce's remand redetermination de novo. See U.S. Steel, 621 F.3d at 1357. We will affirm the agency unless its decision “is unsupported by substantial evidence on the record, or otherwise not in accordance with law.” 19 U.S.C. § 1516a(b)(1)(B)(i). “Commerce's decision will [also] be set aside if it is arbitrary and capricious.” Changzhou Wujin Fine Chem. Factory Co. v. United States, 701 F.3d 1367, 1374 (Fed. Cir. 2012).
We do not defer to an agency's interpretation of the APA's statutory requirements, although the statute itself presumes that review of agency action under the arbitrary-and-capricious standard is “highly deferential.” Nat'l Org. of Veterans' Advocates, Inc. v. Sec'y of Veterans Affairs, 260 F.3d 1365, 1372 (Fed Cir. 2001); see also Collins v. Nat'l Transp. Safety Bd., 351 F.3d 1246, 1253 (D.C. Cir. 2003) (“For generic statutes like the APA, ․ the broadly sprawling applicability undermines any basis for deference, and courts must therefore review interpretative questions de novo.”); Mobil Oil Corp. v. Dep't of Energy, 728 F.2d 1477, 1486–87 (Temp. Emer. Ct. App. 1983) (“We are free to make an independent determination of the legal question as to whether the agency has made a showing of good cause.”).4
We first address Mid Continent's contention that Commerce provided adequate notice for the repeal of the Limiting Regulation through two Federal Register notices issued in 2007 and 2008: (1) Targeted Dumping in Antidumping Investigations; Request for Comment, 72 Fed. Reg. 60,651 (Oct. 25, 2007) [hereinafter Request for Comment]; and (2) Proposed Methodology for Identifying and Analyzing Targeted Dumping in Antidumping Investigations; Request for Comment, 73 Fed. Reg. 26,371 (May 9, 2008) [hereinafter Proposed Methodology].
The requirement that an agency provide adequate notice before altering its regulations is rooted in the APA's provisions governing the administrative rulemaking process. Under the APA, whenever an agency decides to “formulat[e], amend[ ], or repeal[ ] a rule,” it must first publish an NPRM setting forth “either the terms or substance of the proposed rule[,] or a description of the subjects and issues involved.” 5 U.S.C. §§ 553(b), 551(5). For the purposes of notice and comment, withdrawal or repeal of an existing regulation is treated the same as promulgation of a new regulation. See Tunik v. MSPB, 407 F.3d 1326, 1342 (Fed. Cir. 2005). Although the notice “need not specify every precise proposal which [the agency] may ultimately adopt,” it “must be sufficient to fairly apprise interested parties of the issues involved.” Nuvio Corp. v. FCC, 473 F.3d 302, 310 (D.C. Cir. 2006) (quoting Action for Children's Television v. FCC, 564 F.2d 458, 470 (D.C. Cir. 1977)). Adequate notice “is crucial to ‘ensure that agency regulations are tested via exposure to diverse public comment, ․ to ensure fairness to affected parties, and ․ to give affected parties an opportunity to develop evidence in the record to support their objections to the rule and thereby enhance the quality of judicial review.’ ” Int'l Union, United Mine Workers of Am. v. Mine Safety & Health Admin., 626 F.3d 84, 95 (D.C. Cir. 2010) (quoting Int'l Union, United Mine Workers of Am. v. Mine Safety & Health Admin., 407 F.3d 1250, 1259 (D.C . Cir. 2005)).
The dispositive question in assessing the adequacy of notice under the APA is whether an agency's final rule is a “logical outgrowth” of an earlier request for comment. Long Island Care at Home, Ltd. v. Coke, 551 U.S. 158, 174 (2007); Veteran's Justice Grp., LLC v. Sec'y of Veterans Affairs, 818 F.3d 1336, 1344 (Fed. Cir. 2016).
The logical outgrowth doctrine recognizes that a certain degree of change between an NPRM and a final rule is inherent to the APA's scheme of rulemaking through notice and comment. See Int'l Harvester Co. v. Ruckelshaus, 478 F.2d 615, 632 n.51 (D.C. Cir. 1973). Accordingly, judicial formulations of the doctrine have sought to “balance” the values served by adequate notice, see Int'l Union, 626 F.3d at 94–95, with “the public interest in expedition and finality.” Small Refiner Lead Phase-Down Task Force v. EPA, 705 F.2d 506, 547 (D.C. Cir. 1983). We recently stated, for instance, that “[a] final rule is a logical outgrowth of a proposed rule only if interested parties should have anticipated that the change was possible, and thus reasonably should have filed their comments on the subject during the notice-and-comment period.” Veteran's Justice, 818 F.3d at 1344 (alterations and internal quotation marks omitted).5
Courts have consistently upheld final rules as logical outgrowths “where the NPRM expressly asked for comments on a particular issue or otherwise made clear that the agency was contemplating a particular change.” CSX Transp. Inc. v. Surface Transp. Bd., 584 F.3d 1076, 1081 (D.C. Cir. 2009) (citing Owner–Operator Indep. Drivers Ass'n v. Fed. Motor Carrier Safety Admin., 494 F.3d 188, 209–10 (D.C. Cir. 2007); and City of Portland v. EPA, 507 F.3d 706, 715 (D.C. Cir. 2007)); see also, e.g., Alto Dairy v. Veneman, 336 F.3d 560, 570 (7th Cir. 2003) (upholding final rule prohibiting “paper pooling” of milk producers with “distant supply plants” because agency's notice raised the issue of “pool” eligibility); Public Service Commission v. FCC, 906 F.2d 713, 715 (D.C. Cir. 1990) (final rule was logical outgrowth because board affiliated with the agency asked for comments on the proposal that was finally adopted, even though the agency itself did not).
Courts applying the logical outgrowth doctrine have also permitted agencies to drop critical elements of proposed rules even if a resulting final rule effectively abandons an agency's initial proposal. In Long Island Care, for example, the Department of Labor proposed a rule that would have rendered certain “companionship workers” outside the exemption to wage and hour restrictions under the Fair Labor Standards Act (“FLSA”). See 551 U.S. at 174–75. The rule the agency eventually adopted, however, left these workers within the FLSA's exemption. The Court sustained the agency's final rule, observing that “[s]ince the proposed rule was simply a proposal, its presence meant that the Department was considering the matter; after that consideration the Department might choose to adopt the proposal or to withdraw it.” Id. at 175. Because this result was “reasonably foreseeable,” the Court held that the agency had complied with notice-and-comment rulemaking. Id.6
Nonetheless, there are limits to how far a notice of proposed rulemaking may be stretched under the logical outgrowth doctrine. In some cases, these limits may be difficult to discern, Kooritzky v. Reich, 17 F.3d 1509, 1513 (D.C. Cir. 1994), but certain clear lines have been drawn. “The logical outgrowth doctrine does not extend to a final rule that finds no roots in the agency's proposal because something is not a logical outgrowth of nothing, ․ [or] where interested parties would have had to divine the agency's unspoken thoughts, because the final rule was surprisingly distant from the [a]gency's proposal.” Envtl. Integrity Project v. EPA, 425 F.3d 992, 996 (D.C. Cir. 2005) (internal quotation marks and citations omitted).
Having summarized the principles animating the logical outgrowth doctrine, we turn to whether Commerce's repeal of the Limiting Regulation in Withdrawal Notice was a logical outgrowth of Request for Comment and Proposed Methodology. The Trade Court determined that the notices were insufficient because neither notice made “obvious to an interested observer that ․ rule making [to withdraw the rule was] intended” by the agency. Mid Continent I, 999 F. Supp. 2d at 1322. We agree.
We begin with the statute. The Tariff Act as amended by the URAA obligates Commerce to make two findings before the agency may use the average-to-transaction methodology to assess targeted dumping in an investigation. First, the agency must find “a pattern of export prices (or constructed export prices) for comparable merchandise that differ significantly among purchasers, regions, or periods of time.” 19 U.S.C. § 1677f-1(d)(1)(B)(i). Second, Commerce must “explain[ ] why such differences cannot be taken into account using” the other two statutory methods. Id. § 1677f-1(d)(1)(B)(ii).
Once these criteria are met, however, the statute leaves undefined the precise scope of Commerce's application of the average-to-transaction methodology; this led to concerns that if a respondent had been found to be engaged in targeted dumping, but only in some limited fashion, application of the methodology to “all of [the respondent's] sales ․ would be unreasonable and unduly punitive.” See Antidumping Duties; Countervailing Duties, 61 Fed. Reg. 7308, 7350 (Feb. 27, 1997). Commerce responded to these concerns by promulgating the Limiting Regulation, which provided that the agency would “normally limit the application of the average-to-transaction method[ology] to those sales that constitute targeted dumping.” 19 C.F.R. § 351.414(f)(2) (2008). Thus, even if the agency had found a respondent to have engaged in targeted dumping—a condition precedent to the agency's use of the average-to-transaction methodology—under the Limiting Regulation, Commerce would “normally” limit the scope of the average-to-transaction methodology to the respondent's targeted sales—instead of all sales.
Ten years after promulgating the Limiting Regulation, Commerce published Request for Comment, in which the agency sought guidance regarding an appropriate test to determine the existence of targeted dumping. In this notice, Commerce admitted that it had accrued only “limited experience with targeted dumping” despite the intervening years; that it had yet to develop a standard targeted dumping test; and that its “experience with regard to the use of the [average-to-transaction] method ha[d] been very limited.” 72 Fed. Reg. at 60,651. By publishing Request for Comment, Commerce hoped to solicit the public's views on “its development of a methodology for determining whether targeted dumping is occurring in antidumping investigations,” and “input on standards and tests that may be appropriate in a targeted dumping analysis.” Id. Specifically, the agency sought guidance on: (1) how to determine the existence of a “pattern of export prices ․ among purchasers, regions, or periods of time”; (2) how to determine if such a pattern “differ[s] significantly”; and (3) the “appropriate statistical techniques” to assess targeted dumping. Id.
Despite raising these concerns, Request for Comment was not published in the Federal Register as an NPRM, meaning that the notice on its face did not indicate that Commerce was considering a rulemaking. More problematically, Request for Comment did not propose any kind of rule or raise any question about the scope of the average-to-transaction methodology, much less the conditions under which the agency should depart from its “normal” practice of not applying the methodology to all sales. Request for Comment did not even include a citation to the Limiting Regulation. Instead, in Request for Comment, Commerce simply sought information on the broad issue of how the agency should determine the existence of targeted dumping—a distinct, predicate issue to the problem addressed by the Limiting Regulation (i.e., the scope of the average-to-transaction methodology).
The consequence of these deficiencies is that Request for Comment falls short of satisfying the APA's requirements for notice and opportunity for comment. We find the D.C. Circuit decision in Kooritzky to be instructive on this point. At issue in Kooritzky was a “no-substitution” rule promulgated by the Department of Labor that prohibited employers from substituting one alien for another with respect to certifications necessary for obtaining employment-based visas. See 17 F.3d at 1512. In a notice of proposed rulemaking to implement then-recent statutory amendments, the agency made no mention of substitution. Id. at 1513. In rejecting the agency's NPRM as inadequate, the D.C. Circuit observed that the “notice ․ contain[ed] nothing, not the merest hint, to suggest that the [agency] might tighten its existing practice of allowing substitution,” and that the preamble to the agency's notice in the Federal Register “offered no clues” to a “nonexpert reader ․ of what was to come.” Id.
Like the notice at issue in Kooritzky, Request for Comment gave no indication that Commerce was contemplating a potential change in the Limiting Regulation. Nor did commentators responding to Request for Comment perceive the agency to be raising the issue of the regulation's repeal or revision, or suggest such repeal or revision themselves.7 We therefore have no doubt that Commerce's repeal of the Limiting Regulation was not a logical outgrowth of Request for Comment because, as in Kooritzky, “[s]omething is not a logical outgrowth of nothing.” 17 F.3d at 1513.
Six months after Request for Comment, Commerce—still concerned with the appropriate test for determining the existence of targeted dumping—proposed a new two-part test addressing the problem in Proposed Methodology.8 This second notice acknowledged the responses that Commerce had received following Request for Comment, but did not offer the agency's response thereto. See Proposed Methodology, 73 Fed. Reg. at 26,371. In addition to seeking new comments on its proposed test for determining the existence of targeted dumping, Commerce also raised several “related issues.” Id. In particular, the agency “request[ed] comment on the application of the [average-to-transaction methodology] and the conditions, if any, under which the [average-to-transaction] methodology should apply to all sales to the target, even if some sales of a control number do not pass the targeted dumping test.” Id. at 26,372 (emphasis added).
Proposed Methodology thus presents a closer question under the logical outgrowth doctrine than Request for Comment. The Limiting Regulation had provided that Commerce would “normally” apply the average-to-transaction methodology only to “those sales” found to “constitute targeted dumping.” 19 C.F.R. § 351.414(f)(2) (2008). Therefore, by seeking public comment on “the conditions, if any,” under which the average-to-transaction methodology should be applied to all sales made by a respondent—instead of just the respondent's targeted sales—Commerce effectively raised the general subject of the Limiting Regulation, perhaps suggesting wholesale elimination of the agency's discretion to apply the average-to-transaction methodology to all sales.
Although courts have found logical outgrowths when an NPRM “expressly asked for comments on a particular issue or otherwise made clear that the agency was contemplating a particular change,” CSX, 584 F.3d at 1081, we do not think that this principle supports holding Proposed Methodology to have provided the “necessary predicate” for Withdrawal Notice. Kooritzky, 17 F.3d at 1513. For starters, like Request for Comment, Proposed Methodology on its face did not indicate that further action to withdraw the Limiting Regulation was being considered. Instead, Proposed Methodology merely sought public views on how to interpret the regulation itself—which provided that the agency would “normally” not apply the average-to-transaction methodology to all sales—that is, how exactly Commerce should apply the “normally” limitation. Because the agency had left the circumstances in which it would have applied the average-to-transaction methodology to all sales largely undefined, “interested persons” would have perceived the question regarding the Limiting Regulation posed in Proposed Methodology as simply Commerce's first step in clarifying the scope of its own regulation. Indeed, comments that the agency received in response to Proposed Methodology did not understand Commerce to be raising a broader question, i.e., whether to repeal the Limiting Regulation. See note 10, infra.
Posing such a general “scope” question does not suffice to provide the requisite “fair notice” for an agency rule to be upheld as a logical outgrowth. See Long Island Care, 551 U.S. at 174. In CSX, the D.C. Circuit confronted a similar problem in addressing a rule promulgated by the Surface Transportation Board (“STB”) to resolve railroad rate disputes. The STB had originally proposed a rule allowing such disputes to be resolved using “comparison groups drawn from the most recent year of waybill sampling.” 584 F.3d at 1078. In the rule finally adopted, however, the agency “switch[ed] from one year to four years' worth of data.” Id. The STB argued that the final rule was a logical outgrowth because “mention[ing] ․ the release of one-year data ․ gave notice that the amount of data available ․ might change.” Id. at 1082.
The D.C. Circuit rejected this argument for two reasons. First, the court observed that although the STB's notice had proposed a number of related regulatory changes, “it nowhere even hinted that [the agency] might consider expanding the number of years from which comparison groups could be derived.” Id. Second, permitting the “mere mention” of the one-year timeframe for drawing comparison groups to provide adequate notice would allow the agency “to justify any final rule it might be able to devise by whimsically picking and choosing within the four corners of a lengthy ‘notice.’ ” Id. (quoting Envtl. Integrity Project, 425 F.3d at 998). “Such a rule would hardly promote the purposes of the APA's notice requirement.” Id.
The same reasoning applies to Proposed Methodology. Despite mentioning the subject matter of the Limiting Regulation, Commerce's primary purpose in the Proposed Methodology was to propose a new test for determining whether a respondent was engaged in targeted dumping and to seek public comment on this proposal. As a “related issue” the agency posed a general question of when to apply the average-to-transaction methodology to all sales, not just targeted sales. But this question did not raise the “particular issue” of withdrawing the Limiting Regulation; it sought only to clarify the meaning of the Limiting Regulation's recitation of the word “normally.” And, as in CSX, allowing Commerce's question in Proposed Methodology to provide adequate notice for Withdrawal Notice would permit the agency to adopt a final rule from a limitless continuum of regulatory actions. Given this range of possibilities, we cannot say that Commerce's repeal of the Limiting Regulation was “reasonably foreseeable.” Long Island Care, 551 U.S. at 175. It follows that neither Request for Comment nor Proposed Methodology provided adequate notice and opportunity for comment necessary for compliance with the APA.
Mid Continent argues that even if Commerce did not itself provide the required notice, comments made in response to Request for Comment and Proposed Methodology urged Commerce to apply the average-to-transaction methodology to “all sales” and thereby effectively raised the issue of repealing the Limiting Regulation.
Although responses by commentators may be relevant to the court's inquiry under the logical outgrowth doctrine, as a general matter, an agency “cannot bootstrap notice from a comment.” Fertilizer Inst. v. EPA, 935 F.2d 1303, 1312 (D.C. Cir. 1991).9 Here, the comments relied on by Mid Continent never urged Commerce to repeal the Limiting Regulation; commentators simply asked the agency to construe the regulation more or less broadly.10 Many of these comments simply urged Commerce to follow the approach the agency had set forth when it first promulgated the regulation in 1997, viz., that “in some instances, it may be necessary to apply the average-to-transaction methodology to all sales to the targeted area, ․ or even to all sales of a particular respondent,” 62 Fed. Reg. at 27,375 (noting that such cases could encompass respondents engaged in “widespread” or “extensive[ ]” targeted dumping). See note 10, supra. And, the fact that comments responding to Request for Comment and Proposed Methodology were entirely silent on the issue of repealing the Limiting Regulation supports the conclusion that these notices were insufficient to render the agency's actions in Withdrawal Notice a “logical outgrowth.” See, e.g., Council Tree Commc'ns, Inc. v. FCC, 619 F.3d 235, 256 (3d Cir. 2010).
Finally, our conclusion that Withdrawal Notice is not a logical outgrowth of either Request for Comment or Proposed Methodology is further bolstered by four other considerations. First, Commerce never referred to Request for Comment or Proposed Methodology in Withdrawal Notice, nor responded to the comments it had received in response to the two earlier notices.11 Second, in Withdrawal Notice the agency did not adopt any of the proposals made by commentators, choosing instead to resolve the scope of the average-to-transaction methodology through “case-by-case adjudication.” 73 Fed. Reg. at 74,931. Third, Commerce curiously requested further comments regarding its repeal of the Limiting Regulation in Withdrawal Notice, which suggests that the agency believed itself to not have secured adequate comments on the issue. In contrast, Commerce did not make a similar request for additional comments in its 2014 rulemaking to withdraw the Limiting Regulation. See 79 Fed. Reg. at 22,378. Last but not least, Commerce did not suggest in Withdrawal Notice that it had in fact complied with the APA by issuing the earlier notices. To the contrary, Commerce thought it necessary to invoke the APA's good cause exception, which implies that the agency did not consider its prior notices to have satisfied the statute's procedural requirements. Although the inconsistency of simultaneously invoking good cause and arguing post hoc compliance with the APA is not dispositive, the tension between these conflicting positions strongly supports our view that Commerce's (and now, Mid Continent's) assertion that the agency had complied with notice-and-comment rulemaking is not supportable.
In summary, we hold that Commerce's repeal of the Limiting Regulation in Withdrawal Notice was not a logical outgrowth of Request for Comment and Proposed Methodology, and that agency failed to provide adequate notice under the APA.
We must now consider whether Commerce's failure to provide adequate notice may be excused for good cause, the sole ground Commerce cited for dispensing with notice and comment in Withdrawal Notice. An agency may forgo notice-and-comment rulemaking for good cause if it “finds (and incorporates the finding and a brief statement of reasons therefor in the rules issued) that notice and public procedure thereon are impracticable, unnecessary, or contrary to the public interest.” 5 U.S.C. § 553(b)(3)(B).
As a general matter, exceptions to notice-and-comment rulemaking under the APA are “narrowly construed and only reluctantly countenanced.” Mobil Oil, 728 F.2d at 1490 (quoting New Jersey v. EPA, 626 F.2d 1038, 1045 (D.C. Cir. 1980)).12 In Mobil Oil, we stated that an invocation of good cause requires an agency to show that delaying the rule at issue would create “a significant threat of serious damage to important public interests” as the exception would otherwise become an “all purpose escape-clause” to the APA's rulemaking provisions. Id. at 1492. Such “significant threat[s]” encompassed situations where the announcement of a proposed rule itself would “precipitate activity by affected parties that would harm the public welfare,” for example, price controls subject to predatory regulatory arbitrage or other market dislocations. Id. (citing Nader v. Sawhill, 514 F.2d 1064, 1068–69 (Temp. Emer. Ct. App. 1975)). Other courts have emphasized the need to find similarly serious threats in order to invoke the good cause exception. See, e.g., Mack Trucks, Inc. v. EPA, 682 F.3d 87, 93 (D.C. Cir. 2012) (citing “possible imminent hazard to aircraft, persons, and property” and rules of “life-saving importance” necessary to “stave off any imminent threat to the environment or safety or national security”); Haw. Helicopter Operators Ass'n v. FAA, 51 F.3d 212, 214 (9th Cir. 1995) (citing a “recent escalation of fatal air tour accidents”).
The requirement that an agency “incorporate[ ] the finding and a brief statement of reasons” for good cause “in the rules issued” means that we are limited to examining the reasons Commerce cited in Withdrawal Notice to justify its invocation of good cause. See Mobil Oil Corp. v. Dep't of Energy, 610 F.2d 796, 802–03 (Temp. Emer. Ct. App. 1979); see also N.C. Growers' Ass'n, Inc. v. United Farm Workers, 702 F.3d 755, 766–67 (4th Cir. 2012). Commerce cited two of the three available statutory grounds for invoking the good cause exception. First, the agency stated that notice and comment were “impracticable” because the Limiting Regulation was applicable to ongoing dumping investigations, and “immediate revocation [was] necessary to ensure the proper and efficient operation of the antidumping law and to provide the relief intended by Congress.” 73 Fed. Reg. at 74,931. Mid Continent relatedly asserts that dumping investigations are subject to statutory deadlines that cannot be extended at the agency's discretion. See 19 U.S.C. §§ 1673b(b)(1)(A), 1673b(c), 1673d(a).
“Notice and comment on a rule may be found to be ‘impracticable’ when ‘the due and required execution of the agency functions would be unavoidably prevented by its undertaking public rulemaking proceedings.’ ” N.C. Growers, 702 F.3d at 766 (quoting Nat'l Nutritional Foods Ass'n v. Kennedy, 572 F.2d 377, 384–85 (2d Cir. 1978)). Critically, we along with several other courts have held that statutory deadlines in and of themselves do not generally provide a basis for invoking good cause on the ground of impracticability. See, e.g., Shell Oil Co. v. Fed. Emer. Admin., 527 F.2d 1243, 1248 (Temp. Emer. Ct. App. 1975).13 But see Phila. Citizens in Action v. Schweiker, 669 F.2d 877, 885–86 (3d. Cir. 1982) (upholding good cause where Congress gave the agency only 49 days to promulgate regulations implementing a complex scheme of federally funded state benefits). A contrary rule would encourage administrative gamesmanship because “an agency unwilling to provide notice or an opportunity to comment could simply wait until the eve of a statutory, judicial, or administrative deadline, then raise up the ‘good cause’ banner and promulgate rules without following APA procedures.” Council of S. Mountains, Inc. v. Donovan, 653 F.2d 573, 581 (D.C. Cir. 1981). In fact, the temporal exigency implied by Withdrawal Notice appears more theoretical than actual—as Mid Continent observes, Commerce did not have occasion to apply the Limiting Regulation's withdrawal until eight months after Withdrawal Notice, and did not issue a final determination relying on the withdrawal until fifteen months later.14 Thus, the fact that Commerce would have had to apply the Limiting Regulation to ongoing investigations cannot constitute a basis for good cause excusing its failure to go through notice and comment.
Second, Commerce invoked the good cause exception on the ground that notice was “contrary to the public interest” because the agency's application of the Limiting Regulation “may have ․ prevented the use of [the average-to-transaction] methodology to unmask dumping.” 73 Fed. Reg. at 74,931. This argument, however, is again foreclosed by precedent because an assertion of mere pocketbook (or balance-sheet) harm to regulated entities is generally not sufficient to establish good cause as nearly every agency rule imposes some kind of economic cost.15 See Mack Trucks, 682 F.3d at 95 (contrasting such economic harms with a situation “in which an entire industry and its customers [are] imperiled”). As the Trade Court observed, the denial of regulatory relief in this case is not the sort of “pressing urgency of a type that does not always exist in the trade context.” Mid Continent I, 999 F. Supp. 2d at 1323 (citing Gold East Paper (Jiangsu) Co. v. United States, 918 F. Supp. 2d 1317, 1327 (Ct. Int'l Trade 2013)). Thus, Commerce did not show a public interest consideration sufficient to support the agency's invocation of good cause.
On appeal, Mid Continent offers a new justification for good cause that Commerce did not adopt in Withdrawal Notice. Citing Commerce's statement that the “effect” of the agency's targeted dumping regulations was “to deny relief to domestic industries,” and that this effect was “inconsistent with the [agency's] statutory mandate to provide [such] relief,” 73 Fed. Reg. at 74,931, Mid Continent argues that Commerce “had determined the withdrawal was necessary because the existing regulations were contrary to law,” and thus “immediate withdrawal was ․ fully justified.” In connection with this argument, Mid Continent cites the doctrine of deference to agency statutory interpretations under Chevron, U.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 842–45 (1984), to assert that the Limiting Regulation was contrary to statute because it was inconsistent with Commerce's view of the statute in Withdrawal Notice. Mid Continent's theory, therefore, is that there was no need for Commerce to undergo notice-and-comment rulemaking because the Limiting Regulation was contrary to the statutory provisions of the Tariff Act.
This theory of good cause did not appear in Withdrawal Notice and therefore cannot support a finding of good cause. See N.C. Growers, 702 F.3d at 767. In any case, we do not agree with Mid Continent's premise that the agency had determined the Limiting Regulation to be “contrary to law.” Commerce did not state in Withdrawal Notice that the Limiting Regulation was contrary to an unambiguous statutory provision—and, to our knowledge, no party has ever challenged the validity of the Limiting Regulation under the Tariff Act. What Commerce actually stated was that the “effect” of the regulations was “inconsistent ․ with [its] statutory mandate,” which the agency broadly framed as “provid[ing] relief to domestic industries materially injured by unfairly traded imports.” 73 Fed. Reg. at 73,931. These statements are not tantamount to a determination that a regulation is contrary to an unambiguous provision of statutory law.16
Nor do we agree that the inconsistency of a regulation adopted under an agency's previous statutory interpretation with the agency's present statutory interpretation ipso facto renders the regulation “contrary to law.” By definition, an agency's ability to alter its statutory interpretation requires statutory ambiguity, and, under Chevron, an agency can only reject a prior interpretation of an ambiguous statute if it explains why it is doing so. See, e.g., Nat'l Cable & Telecomms. Ass'n v. Brand X Internet Servs., 545 U.S. 967, 981–82 (2005); Smiley v. Citibank (South Dakota), N.A., 517 U.S. 735, 742 (1996); Rust v. Sullivan, 500 U.S. 173, 186–87 (1991). In this situation, notice-and-comment rulemaking under the APA is more—rather than less—important to lay the groundwork for the agency's exercise of its Chevron authority.
Thus, we agree with the Trade Court that Commerce's invocation of the good-cause exception did not support its decision to dispense with notice-and-comment rulemaking under the APA.
Mid Continent argues that even if Commerce's repeal of the Limiting Regulation violated the APA, the agency's actions may nonetheless be affirmed on the ground of harmless error. The APA directs reviewing courts to take “due account ․ of the rule of prejudicial error” in deciding whether to “hold unlawful and set aside agency action.” See 5 U.S.C § 706(2). The Supreme Court has dedescribed this provision as an “administrative law ․ harmless error rule.” Shinseki v. Sanders, 556 U.S. 396, 406 (2009). We must therefore determine whether Commerce's failure to comply with notice-and-comment rulemaking may be excused as harmless error.
Mid Continent contends that Commerce's procedural error was harmless because Precision cannot show prejudice of a sort cognizable under the statute. Relying on our decision in Intercargo Insurance Co. v. United States, Mid Continent argues that “[p]rejudice ․ means injury to an interest that the statute, regulation, or rule in question was designed to protect,” and that the only injury Precision can show—that Commerce reached an adverse decision in its dumping investigation—is not an interest protected by notice and comment. 83 F.3d 391, 396 (Fed. Cir. 1996). Precision counters that it was required to show only that Commerce's procedural error had some “bearing on ․ the substance of [the] decision reached,” and that given the magnitude of the agency's error and its inability to participate in a rulemaking, this standard is satisfied. Riverbend Farms, Inc. v. Madigan, 958 F.2d 1479, 1487 (9th Cir. 1992).
In determining whether a procedural error committed in the course of rulemaking was harmless under the APA, courts have distinguished between an agency's “technical failure” or substantial compliance with the APA's procedural requirements on one hand (which may constitute harmless error), and its “complete failure” to do so on the other (which may prevent the error from being harmless). United States v. Reynolds, 710 F.3d 498, 516–19 (3d Cir. 2013).17 “In the first category, the agency has provided some notification and method for commenting but some technical failure in that process violates statutory requirements. In these ‘technical failure’ cases, the party challenging the agency rule ‘may be required to demonstrate that, had proper notice been provided, they would have submitted additional, different comments that could have invalidated the rationale’ of the rule.” Id. at 516 (internal citation omitted) (quoting City of Waukesha v. EPA, 320 F.3d 228, 246 (D.C. Cir. 2003)).
To illustrate this first “technical failure” category of cases, in Riverbend Farms, the Ninth Circuit concluded that the Secretary of Agriculture's failure to publish notice in the Federal Register and refusal to accept written comments were harmless because the parties challenging the rule were given actual notice—albeit not published in the Federal Register—and had the opportunity to give oral comments at meetings conducted by the agency. See 958 F.2d at 1488. Similarly, in Friends of Iwo Jima v. National Capital Planning Commission, the Fourth Circuit held that the National Capital Planning Commission's failure to provide notice for two meetings in a “protracted process” was harmless because the challengers had notice of other opportunities to submit comments, and the substance of the comments they allegedly would have submitted was the “main focus of each stage in the approval process.” 176 F.3d 768, 774 (4th Cir. 1999); see also, e.g., Air Transport Ass'n of Am. v. Civil Aeronautics Bd., 732 F.2d 219, 224 n.11 (D.C. Cir. 1984) (agency's failure to timely provide internal staff studies was harmless in the absence of petitioner “explain[ing] what it would have said had it been given earlier access”).
Our decision in Intercargo is consistent with these “technical failure” cases. In Intercargo, after concluding that the Customs Service was required to recite a statutory basis when issuing an extension of time to liquidate import entries, we considered whether the Service's failure to do so was harmless. See 83 F.3d at 392, 394. We noted that the “omission of the requisite language ․ had no effect on [the] right to challenge the extension” and that the importer had not alleged the absence of a statutory basis—the agency had simply failed to identify the basis in its notice. Id. at 396. In rejecting the imposition of additional duties as a source of prejudice, we observed that “[a] party is not ‘prejudiced’ by a technical defect simply because that party will lose its case if the defect is disregarded.” Id. (emphasis added).
In the second, “complete failure” category of cases, the total absence of notice-and-comment rulemaking and the resulting thin or nonexistent record make it difficult for a reviewing court to conclude with certainty that no prejudice has ensued. See Reynolds, 710 F.3d at 518. In such cases, even a minimal showing of prejudice may suffice to defeat a claim of harmless error because “an utter failure to comply with notice and comment cannot be considered harmless if there is any uncertainty at all as to the effect of that failure.” Sugarcane Growers Coop. of Fla. v. Veneman, 289 F.3d 89, 96 (D.C. Cir. 2002); see also, e.g., Sprint Corp. v. FCC, 315 F.3d 369, 376 (D.C. Cir. 2003); Paulsen v. Daniels, 413 F.3d 999, 1007 (9th Cir. 2005).
Commerce's failure to comply with the APA was not a mere technical defect, but amounted to a complete failure to provide the adequate notice and opportunity for comment that the APA requires. There is considerable uncertainty as to the effect of this failure. We find it significant that during Commerce's subsequent rulemaking to withdraw the Limiting Regulation, the agency relied on its post-2008 experience to justify the repeal. See 79 Fed. Reg. at 22,375 (noting Commerce's development of “differential pricing analysis”). Moreover, Commerce did not in Withdrawal Notice address any substantive objections to withdrawing the Limiting Regulation. Cf. United States v. Johnson, 532 F.3d 912, 931 (5th Cir. 2011) (finding harmless error where the agency “thoroughly engage[d] the issues and challenges inherent in the regulation” and “was able to address objections in the interim final rule”). The agency in fact did not address those objections until its 2014 rulemaking. See 79 Fed. Reg. at 22,374–75. All this suggests that Commerce's failure to go through notice and comment could well have affected the result reached in Withdrawal Notice.18
Accordingly, we hold that Commerce's failure to comply with notice-and-comment rulemaking cannot be excused as harmless error.
We finally address Mid Continent's argument that Commerce erred in applying the Limiting Regulation on remand from the Trade Court. To recap, Commerce's remand redetermination applied the Limiting Regulation and concluded that application of the average-to-transaction methodology to all of Precision's sales was unwarranted because “the record does not contain evidence to suggest that this normal limitation should not be applied.” J.A. 89. On appeal, Mid Continent argues that Commerce misapplied the Limiting Regulation by failing to reinterpret the regulation to be consistent with the agency's post-2008 interpretation of the statute, which assertedly requires broader application of the average-to-transaction methodology. In connection with this argument, Mid Continent suggests that Commerce misinterpreted the Trade Court's remand instructions as prohibiting the agency from reinterpreting the regulation, or that the court erred by depriving the agency of such discretion and then deferring to Commerce's application of the Limiting Regulation on remand.
Having examined Commerce's remand redetermination, we find Mid Continent's arguments unavailing. The Trade Court's instructions did not compel Commerce to apply the average-to-transaction methodology only to targeted sales, and on remand, the agency did not misinterpret the court's instructions. See Mid Continent II, 113 F. Supp. 3d at 1327 (“To the extent that [Mid Continent] argues that the government adopted an inappropriately narrow view of its authority ․ and inaccurately construed the remand order as cover for doing so, [Mid Continent] is mistaken.”).
As for Mid Continent's argument that Commerce erred by not reinterpreting the Limiting Regulation, this argument misses the mark. There is no serious contention that Commerce's application of the Limiting Regulation contravened an unambiguous provision of statutory law, or was otherwise “plainly erroneous or inconsistent with the [Limiting] [R]egulation” itself. Auer v. Robbins, 519 U.S. 452, 462 (1997); Bowles v. Seminole Rock & Sand Co., 325 U.S. 410, 414 (1945). Nor has Mid Continent argued that the agency's application of the regulation was arbitrary, capricious, or unsupported by substantial evidence. In the absence of these contentions, a court is not free to displace an agency's reasoned application of its own rule. Mid Continent's argument that Commerce misapplied the Limiting Regulation in the agency's remand redetermination is without merit.
For the stated reasons, we hold that Commerce failed to comply with notice-and-comment rulemaking under the APA by repealing the Limiting Regulation in Withdrawal Notice, that its failure cannot be excused for good cause or harmless error, and that the agency did not err in applying the Limiting Regulation on remand. The judgment of the Court of International Trade is
Costs to appellee.
1. Another mandatory respondent identified by Commerce, Dubai Wire FZE (“Dubai Wire”), participated in the agency's dumping investigation and intervened in the Trade Court, but did not file a brief in this appeal. We have limited our recitation of the facts to those pertinent to Precision, but note that the relief sought by Mid Continent could impact Dubai Wire as well.
2. Commerce stated at the time that “it would be ‘unreasonable and unduly punitive’ to apply the [average-to-transaction methodology] to all sales where, for example, targeted dumping accounted for only one percent of a firm's total sales.” 62 Fed. Reg. at 27,375.
3. The Trade Court first reached this conclusion in an earlier case, Gold East Paper (Jiangsu) Co. v. United States, 918 F. Supp. 2d 1317, 1328 (Ct. Int'l Trade 2013).
4. Accord Sorenson Commc'ns Inc. v. FCC, 755 F.3d 702, 706 (D.C. Cir. 2014); Iowa League of Cities v. EPA, 711 F.3d 844, 872 (8th Cir. 2013); Meister v. U.S. Dep't of Agric., 623 F.3d 363, 370–71 (6th Cir. 2010); Reno-Sparks Indian Colony v. EPA, 336 F.3d 899, 909 n.11 (9th Cir. 2003); Warder v. Shalala, 149 F.3d 73, 79 (1st Cir. 1998); Weyerhaeuser Co. v. Shalala, 590 F.2d 1011, 1027 (D.C. Cir. 1978). See generally United States v. Reynolds, 710 F.3d 498, 507–09 (3d Cir. 2013); Jared P. Cole, Cong. Research Serv., R44356, The Good Cause Exception to Notice and Comment Rulemaking: Judicial Review of Agency Action 13–14 (2016).
5. See also Am. Water Works Ass'n v. EPA, 40 F.3d 1266, 1274 (D.C .Cir. 1994) (“We apply [the logical outgrowth] standard functionally by asking ․ whether a new round of notice and comment would provide the first opportunity for interested parties to offer comments that could persuade the agency to modify its rule.”).
6. See also, e.g., Veterans Justice, 818 F.3d at 1345 (upholding a final rule because “[o]ne logical outgrowth of a proposal is surely ․ to refrain from taking the proposed step”); Ariz. Pub. Serv. Co. v. EPA, 211 F.3d 1280, 1299 (D.C. Cir. 2000) (“In first proposing that tribes would have to meet the ‘same requirements' [for judicial review under the Clean Air Act] as states, EPA effectively raised the question as to whether this made sense.”).
7. At best, commenting parties understood the agency to be open to suggestions on how to apply the Limiting Regulation. To illustrate, as one commentator stated: “[T]he Department should clarify when it will apply the average-to-transaction methodology to all sales, rather than only targeted sales. We think it would be appropriate ․ to apply the average-to-transaction method to all sales ․ where the targeted quantity exceeds twenty percent of the U.S. sales database.” Letter from David A. Hartquist, Executive Director, Committee to Support U.S. Trade Laws to David Spooner, Assistant Secretary for Import Administration 3 (Dec. 10, 2007) (emphasis added), available at https://perma.cc/5FJR-WKZD.
8. Under this test—also known as the Nails test, see JBF RAK LLC v. United States, 790 F.3d 1358, 1367 & n.5 (Fed. Cir. 2015)—Commerce first “determine[s], on an exporter-specific basis, the share of the allegedly targeted customer's purchases of subject merchandise, by sales value, that are at prices more than one standard deviation below the weighted-average price to all customers of that exporter, targeted and non-targeted.” 73 Fed. Reg. at 26,372. “If that share exceeds 33 percent of the total value of the exporter's sales of subject merchandise to the allegedly targeted customer, then the pattern requirement is met.” Id. In the second part of the Nails test, Commerce “determine[s] the total sales value for which the difference between (i) the sales-weighted average price to the allegedly targeted customer and (ii) the next higher sales-weighted average price to a non-targeted customer exceeds the average price gap ․ for the non-targeted group.” Id. If the share of sales satisfying these criteria “exceeds 5 percent of the total value of sales of subject merchandise to the allegedly targeted customer,” then the pattern of price differences is deemed “significant,” and the exporter will be found to have engaged in targeted dumping. Id.
9. See also Shell Oil Co. v. EPA, 950 F.2d 741, 751 (D.C. Cir. 1991) (noting that under NRDC v. Thomas, 838 F.2d 1224, 1243 (D.C. Cir. 1998), comments to an agency proposal are a relevant factor if they raise a “foreseeable possibility of agency action”); Int'l Union, 407 F.3d at 1261 (underscoring that NRDC v. Thomas represents “the outer limits of the ‘logical outgrowth’ doctrine” and that the agency in that case gave notice and opportunity for comment two weeks before promulgating the final rule).
10. Responses to Proposed Methodology, for example, suggested a number of ways to apply the Limiting Regulation, including the establishment of numerical thresholds that if satisfied would result in applying the average-transaction methodology to all sales. See Letter from King & Spalding LLP to Hon. David Spooner, Assistant Secretary for Import Administration 12 (June 23, 2008), available at https://perma.cc/Q7T6-5RH3 (proposing a twenty percent threshold based on U.S. sales); Letter from Skadden, Arps, Slate, Meagher & Flom LLP to David Spooner, Assistant Secretary for Import Administration 20 (June 23, 2008), available at https://perma.cc/HUJ4-UXPE (proposing a twenty percent threshold or “when the Department cannot identify the full scope of the respondent's targeted dumping”).Mid Continent identifies a number of specific comments responding to Proposed Methodology that it contends addressed “possible modification” of the Limiting Regulation. We disagree. These comments addressed Commerce's interpretation of the Limiting Regulation's “normally” limitation and did not suggest revision or repeal. To illustrate, one comment cited by Mid Continent stated that Commerce “should apply the [average-to-transaction] methodology to all of the sales to the target” because “[o]nce a customer or region has been identified as being targeted by a respondent ․ [Commerce] should consider that all sales to that target are subject to the same pricing practices and are, therefore, targeted sales.” Letter from David A. Hartquist, Kelley Drye & Warren LLP to Secretary of Commerce 30 (June 23, 2008) (emphasis added), available at https://perma.cc/D34C-VU94. The emphasized portions of this comment underscore the comment's consistency with the Limiting Regulation, i.e., that Commerce should “normally” limit the average-to-transaction methodology to “sales that constitute targeted dumping.” 19 C.F.R. § 351.414(f)(2) (2008).
11. See, e.g., Motor Vehicle Mfrs. Ass'n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 48 (1983); Disabled Am. Veterans v. Gober, 234 F.3d 682, 692 (Fed. Cir. 2000) (“[I]nextricably intertwined with ․ 5 U.S.C. § 553(c) is the agency's need to respond, in a reasoned manner, to any comments received by the agency that raise significant issues with respect to a proposed rule.”).
12. Accord NRDC v. Abraham, 355 F.3d 179, 204 (2d Cir. 2004); United States v. Reynolds, 710 F.3d 498, 507–08 (3d Cir. 2013); United States v. Gould, 568 F.3d 459, 469 (4th Cir. 2009); United States v. Johnson, 632 F.3d 912, 928 (5th Cir. 2011); United States v. Cain, 583 F.3d 408, 420–21 (6th Cir. 2009); Nw. Airlines, Inc. v. Goldschmidt, 645 F.2d 1309, 1321 (8th Cir. 1981); Alcaraz v. Block, 746 F.2d 593, 612 (9th Cir. 1984); N. Am. Coal Corp. v. Dep't of Labor, 854 F.2d 386, 388 (10th Cir. 1988); United States v. Dean, 604 F.3d 1275, 1279 (11th Cir. 2010); Utility Solid Waste Activities Grp. v. EPA, 236 F.3d 749, 754 (D.C. Cir. 2001).
13. See also Council of S. Mountains, Inc. v. Donovan, 653 F.2d 573, 581 (D.C. Cir. 1981); U.S. Steel Corp. v. EPA, 595 F.2d 207, 213 (5th Cir. 1979); Am. Iron & Steel Inst. v. EPA, 568 F.2d 284, 292 (3d Cir. 1977).
14. See Polyethylene Retail Carrier Bags From Taiwan, Preliminary Determination, 74 Fed. Reg. 55,183, 55,187–88 (Oct. 27, 2009); Polyethylene Retail Carrier Bags From Taiwan, Final Determination, 75 Fed. Reg. 14,569, 14,569 (Mar. 26, 2010).
15. See generally Maeve P. Carey, Cong. Research Serv., R41974, Cost-Benefit and Other Analysis Requirements in the Rulemaking Process (Dec. 9, 2014) (summarizing presidential and congressional actions requiring agencies to conduct economic cost-benefit analysis).
16. See Nat'l Customs Brokers & Forwarders Ass'n v. United States, 59 F.3d 1219, 1223–24 (Fed. Cir. 1995) (upholding agency's invocation of good cause where regulation was amended without notice or comment to exactly parallel intervening statutory amendment).
17. See also, e.g., Sprint Corp. v. FCC, 315 F.3d 369, 376 (D.C. Cir. 2003) (“[A]n utter failure to comply with notice and comment cannot be considered harmless if there is any uncertainty at all as to the effect of that failure.”); Shell Oil Co. v. EPA, 950 F.2d 741, 752 (D.C. Cir. 1991) (“While petitioners must show that they would have submitted new arguments to invalidate rules in the case of certain procedural defaults, such as an agency's failure to provide access to supplemental studies, petitioners need not do so here, where the agency has entirely failed to comply with notice-and-comment requirements, and the agency has offered no persuasive evidence that possible objections to its final rules have been given sufficient consideration.” (citations omitted)); compare Int'l Union, 407 F.3d at 1259–61 (final rule setting maximum air velocity cap where proposed rule only set minimum cap was not a logical outgrowth, not harmless), with Int'l Union, 626 F.3d at 95–96 (D.C. Cir. 2010) (challenger failed to demonstrate prejudice because it was able to participate in the agency's rulemaking proceedings and raised issues on appeal that were already “encompassed in its comments”).
18. We also do not think that Commerce's subsequent decision to formally withdraw the Limiting Regulation changes the calculus of our decision; agency attempts to cure procedural defects ex post are not generally accepted as validating prior missteps. See, e.g., Mack Trucks, 682 F.3d at 95; U.S. Steel Corp., 595 F.2d at 214–15.
DYK, Circuit Judge.