dissenting.
The so-called “petition support requirement” of the Byrd Amendment requires that a company publicly express “support of the petition” resulting in an antidump-ing duty order in order to be eligible to receive any funds collected as a result of the order. 19 U.S.C. § 1675c(b)(l)(A) (2000) (repealed 2006). Put simply, under the petition support requirement, if a domestic company publicly expresses the viewpoint that the government should impose a tariff on an importer, then the domestic company is eligible to receive some part of that tariff. If the domestic company either expresses the viewpoint *1361that a tariff should not be imposed or takes no public position, it is not eligible.
The majority concedes that the petition support requirement implicates the First Amendment, but it concludes that the Byrd Amendment satisfies the test for regulation of commercial speech, because “reward[ing] injured parties who assisted government enforcement of the antidump-ing laws by initiating or supporting anti-dumping proceedings” is “similar to commercially contracting with them to assist in the performance of a government function.” Maj. Op. at 1352, 1355. I respectfully disagree.
The Byrd Amendment has nothing to do with rewarding helpfulness during trade investigations as the majority suggests. The majority errs by relying on the statutory construction doctrine of constitutional avoidance to graft its “reward” purpose onto the statute, when that purpose is not apparent in the statutory text or legislative history and has been expressly disclaimed by the government in this case. The majority compounds this error by using its “reward construction” of the petition support requirement to justify evaluating the constitutionality of the requirement under the more lenient commercial speech doctrine, when, in fact, the petition support requirement regulates pure political speech and — by the language of the statute itself — “petitioning].” See U.S. Const, amend. 1 (“Congress shall make no law ... abridging ... the right ... to petition the Government for a redress of grievances”). Because I would conclude that the petition support requirement is an unconstitutional viewpoint discriminatory restriction on political speech and petitioning activity that cannot survive strict scrutiny, I respectfully dissent.1
I
Under 19 U.S.C. § 1673, an antidumping duty can only be imposed if the Department of Commerce (“Commerce”) first determines that “foreign merchandise is being, or is likely to be, sold in the United States at less than its fair value” and the International Trade Commission (“ITC”) then determines that a domestic industry “is materially injured, or ... is threatened with material injury.” The purpose of an antidumping investigation is to determine whether these two criteria have been satisfied. See, e.g., 19 U.S.C. § 1673a(a)(l) (“An antidumping duty investigation shall be initiated whenever the administering authority determines, from information available to it, that a formal investigation is warranted into the question of whether the elements necessary for the imposition of a duty under section 1673 of this title exist”).
Critically, an antidumping duty order is neither required nor even permitted in every case of dumping. As the majority correctly points out, “dumping” is merely “the sale or likely sale of goods at less than fair value.” Id. § 1677(34). But an antidumping duty order requires an additional finding of material injury to the domestic industry. The companies that make up the domestic industry may reasonably disagree as to whether particular dumping has “materially injured” the domestic industry as a whole. Indeed, recognizing the complex and somewhat subjective nature of the material injury requirement, we have held that the ITC “has broad discretion” in determining whether the domestic industry has been materially injured. Nucor Corp. v. United *1362States, 414 F.3d 1331, 1336-37 (Fed.Cir.2005) (concluding that ITC’s methodology in assessing material injury is entitled to Chevron deference). Thus, one member of the domestic industry may honestly believe that the industry is not harmed by particular dumping, while another member may honestly believe that the industry has been harmed. It is the ITC’s obligation to sort out these conflicting views in an anti-dumping duty investigation.
During the course of the ITC’s investigation, the ITC submits questionnaires to domestic producers in the affected industry. As the majority notes, these questionnaires are “extremely detailed, requesting several years of data on a domestic producer’s shipments, employment, sales, finances, pricing, customers, and competitors” and “the costs of responding to such questionnaires are substantial.” Maj. Op. at 1358. Yet all members of the domestic industry who receive such a questionnaire are required by law to complete it. See 19 U.S.C. § 1333(a), (f) (authorizing ITC to request information, issue subpoenas, and demand statements under oath); see also U.S. Int’l Trade Comm’n, Generic U.S. Producer Questionnaire (“Producers’ Questionnaire ”), at 1, available at http:// www.usitc.gov/trade_remedy/731_ad_701_ cvd/investigations/question/USProducer Questionnaire.pdf (“This report is mandatory and failure to reply as directed can result in a subpoena or other order to compel the submission of records or information in your possession.... ”). Moreover, each questionnaire requires that an authorized company official certify the correctness of all responses. Producers’ Questionnaire at 1.
The questionnaire includes various questions related to the harm that the member of the domestic industry has suffered as a result of alleged dumping. Specifically, the questionnaire includes question III — 14, which asks whether the domestic company “experienced any actual negative effects on its return on investment or its growth, investment, ability to raise capital, existing development and production efforts (including efforts to develop a derivative or more advanced version of the product), or the scale of capital investments as a result of imports of’ the allegedly dumped product. Id. at 13. Likewise, question III — 15 asks whether the domestic company “an-tieipate[s] any negative impact of imports of’ the allegedly dumped product. Id. Questions IV-20 and IV-21 also ask for detailed information about any lost revenues or lost sales as a result of dumping. Id. at 24-25.2
In addition to all of these questions about the harm that the alleged dumping has caused each domestic producer, the questionnaire includes, in its “General Information” section, question 1-3, which asks simply “Do you support or oppose the petition?” Id. at 2. Question 1-3 offers three possible choices with corresponding checkboxes: “Support,” “Oppose,” and “Take no position.” Id. at 2.3 It is the domestic producer’s response to this question that the ITC uses to determine wheth*1363er the “petition support requirement” of the Byrd Amendment has been satisfied.
Notably, Commerce also uses responses to question 1-3 to determine whether a petition seeking imposition of an anti-dumping duty is filed “on behalf of the industry” — as is required by 19 U.S.C. § 1673a. For a petition to meet this requirement, “domestic producers or workers who support the petition [must] account for at least 25 percent of the total production of the domestic like product” and “domestic producers or workers who support the petition [must] account for more than 50 percent of the production of the domestic like product produced by that portion of the industry expressing support for or opposition to the petition.” Id. § 1673a(c)(4)(A)(i)-(ii). In other words, question 1-3 is an opportunity for each member of the domestic industry to vote on whether a petition should or should not go forward. To go forward, the petition needs the votes of at least 25% of the domestic industry by production, and no more than 50% in opposition.
Under the Byrd Amendment, United States Customs and Border Protection (“Customs”) disburses duties collected pursuant to antidumping duty orders to “affected domestic producers” who submit a certification claiming that they have incurred certain specified types of expenditures. Id. § 1675c. Though the ordinary meaning of “affected domestic producer” would not seem to require that the producer have taken any particular position in the antidumping duty investigation that resulted in the antidumping duty order, the Byrd Amendment includes a special definition of “affected domestic producer” that imposes just such a requirement:
The term “affected domestic producer” means any manufacturer, producer, farmer, rancher, or worker representative (including associations of such persons) that—
(A) was a petitioner or interested party in support of the petition with respect to which an antidumping duty order, a finding under the Antidump-ing Act of 1921, or a countervailing duty order has been entered, and
(B) remains in operation.
Id. § 1675c(b)(l) (emphasis added).
SKF participated in the investigation that led to the antidumping order at issue in this case, but SKF opposed the petition on the ground that the domestic industry was not being materially injured by dumping. The ITC disagreed with SKF and found material injury. But even though SKF is a member of the injured industry, SKF is precluded from receiving distributions by operation of the petition support requirement, as a result of expressing its view that an antidumping duty order should not be imposed.
II
The majority begins its First Amendment analysis by reciting the well known doctrine of constitutional avoidance, which holds that “[w]here an otherwise acceptable construction of a statute would raise serious constitutional problems, the Court will construe the statute to avoid such problems unless such construction is plainly contrary to the intent of Congress.” Edward J. DeBartolo Corp. v. Fla. Gulf Coast Bldg. & Constr. Trades Council, 485 U.S. 568, 575, 108 S.Ct. 1392, 99 L.Ed.2d 645 (1988) (cited in Maj. Op. at 1349). Applying this doctrine, the majority concludes that the purpose of the Byrd Amendment was not, as the government argues, only “to compensate those who are *1364injured by dumping,” Maj. Op. at 1351, but rather “to reward injured parties who assisted government enforcement of the anti-dumping laws by initiating or supporting antidumping proceedings,” id. at 26. From this conclusion, the majority reasons that evaluating the petition support requirement under the commercial speech doctrine “seems appropriate,” because “[rjewarding parties under the circumstances here is similar to commercially contracting with them to assist in the performance of a government function.” id. at 31. Applying the Central Hudson test for commercial speech, the majority concludes that the petition support requirement survives First Amendment scrutiny, and that the petition support requirement “is not more extensive than is necessary to serve [the government’s] interest” in rewarding injured parties who assist in antidumping investigations. Id. at 31 (quoting Cent. Hudson Gas & Elec. Corp. v. Pub. Serv. Comm’n, 447 U.S. 557, 566, 100 S.Ct. 2343, 65 L.Ed.2d 341 (1980)).
While the majority opinion is well written, thoughtful, and thorough, I respectfully disagree with several aspects of the majority’s analysis. First, the majority focuses on the Byrd Amendment as a whole, rather than on the challenged portion of the Byrd Amendment — namely, the petition support requirement in the definition of “affected domestic producer.” SKF does not challenge the constitutionality of imposing a duty on dumped goods that harm a domestic industry, nor does it challenge the constitutionality of distributing the duties collected as a result of anti-dumping orders to domestic producers. To the contrary, SKF challenges only the petition support requirement of 19 U.S.C. § 1675c(b)(l)(A). That is, SKF challenges only the aspect of the Byrd Amendment that precludes it from receiving duties solely because it answered “Oppose” to question 1-3 of the investigation questionnaire. Thus, the issue is whether the petition support requirement — not the Byrd Amendment as a whole — survives First Amendment scrutiny. See, e.g., Republican Party of Minn. v. White, 536 U.S. 765, 774-75, 122 S.Ct. 2528, 153 L.Ed.2d 694 (2002) (focusing on specific challenged clause of statute and holding that “[u]nder the strict-scrutiny test, respondents have the burden to prove that the [challenged] clause is (1) narrowly tailored, to serve (2) a compelling state interest”). Thus, it is not, as the majority suggests, the government’s “interest in trade law enforcement” that matters. Maj. Op. at 1355. The relevant interest is the government’s more limited interest in conditioning receipt of distributions on public support for an anti-dumping position.
Second, in my view, the majority’s undue focus on “determinfing] the purpose of the Byrd Amendment,” Id. at 23, is inconsistent with the Supreme Court’s First Amendment jurisprudence. The Supreme Court has made clear that it is a statute’s effect on speech that matters, not its intended purpose. See Simon & Schuster, Inc. v. Members of N.Y. State Crime Victims Bd., 502 U.S. 105, 117, 112 S.Ct. 501, 116 L.Ed.2d 476 (1991) (“The Board next argues that discriminatory financial treatment is suspect under the First Amendment only when the legislature intends to suppress certain ideas. This assertion is incorrect; our cases have consistently held that illicit legislative intent is not the sine qua non of a violation of the First Amendment.” (internal quotation marks omitted)). The question is not whether Congress intended the Byrd Amendment to violate the First Amendment. The question is whether it does.
To be sure, determining whether the government interest served by a restric*1365tion on speech is “compelling” — or, in some cases, “important” or “substantial”— is a part of the First Amendment analysis. See, e.g., Boos v. Barry, 485 U.S. 312, 322, 108 S.Ct. 1157, 99 L.Ed.2d 333 (1988) (holding that content-based restrictions on political speech in public forum must be “necessary to serve a compelling state interest and ... narrowly drawn to achieve that end” (quoting Perry Educ. Ass’n v. Perry Local Educators’ Ass’n, 460 U.S. 37, 45, 103 S.Ct. 948, 74 L.Ed.2d 794 (1983))); Central Hudson, 447 U.S. at 566, 100 S.Ct. 2343 (assessing whether “the asserted governmental interest is substantial”); United States v. O’Brien, 391 U.S. 367, 377, 88 S.Ct. 1673, 20 L.Ed.2d 672 (1968) (assessing whether regulation “furthers an important or substantial governmental interest”). But the Byrd Amendment’s purpose plays a much greater role in the majority’s analysis than serving an important government interest. The majority uses its view of the purpose of the Byrd Amendment to shield the petition support clause from strict scrutiny under the First Amendment entirely. Specifically, the majority reasons that because the purpose of the Byrd Amendment was to reward injured parties who assisted government enforcement of the antidumping laws, then the petition support requirement is “similar to commercially contracting with [parties] to assist in the performance of a government function, in this particular context assisting in the enforcement of government policy in litigation.” Maj. Op. at 1355. Thus, the majority uses the purpose of the Byrd Amendment as justification for applying the more lenient Central Hudson test, rather than strict scrutiny. I know of no case- — and the majority has cited none — in which an unambiguous statute that would otherwise be subject to strict constitutional scrutiny receives more lenient scrutiny because of its perceived purpose.
Third, I believe that the majority is incorrect in concluding that that purpose is to reward parties that assist the government in antidumping investigations. Id. at 1352 (“[T]he purpose of the Byrd Amendment’s limitation of eligible recipients was to reward injured parties who assisted government enforcement of the antidump-ing laws by initiating or supporting anti-dumping proceedings.”). There is nothing in the statutory text or legislative history of the Byrd Amendment to suggest that its purpose was to reward assistance or cooperation with the government’s investigation of dumping. To the contrary, the purpose of the Byrd Amendment was to compensate domestic producers injured by dumping. The text and structure of the statute itself makes that clear in specifying that distributions are made to “affected domestic producer[s]” — domestic producers that have been “affected” (i.e., injured) by dumping. 19 U.S.C. § 1675c. The majority relies on general statements in the Congressional findings that “United States unfair trade laws have as their purpose the restoration of conditions of fair trade” and that “injurious dumping is to be condemned.” Maj. Op. at 1352. But neither these statements nor anything else in the statutory text says anything at all about rewarding parties for “assisting] government enforcement” in antidumping proceedings.
Moreover, the legislative history of the Byrd Amendment supports the view that its purpose was to compensate injured domestic producers: “Current law also does not contain a mechanism to help injured U.S. industries recover from the harmful effects of foreign dumping and subsidization.” 145 Cong. Rec. S497, 497 (1999) *1366(statement of Sen. DeWine). The majority’s reliance on general statements in the legislative history — e.g., that the Byrd Amendment is necessary to “deter unfair trade practices” and that “United States trade laws should be strengthened to see that the remedial purpose of those laws is achieved” — is to no avail, because none of these statements says anything about rewarding parties for helping to enforce trade laws. Maj. Op. at 1352.
I note further that the majority’s “reward for assistance” rationale was not argued by either of the two government agencies that are parties to this appeal. To the contrary, the government argued that the purpose of the Byrd Amendment was solely compensation for injury, not reward for assistance: “Simply stated, as a supplement to unfair trade laws already in existence, in the [Byrd Amendment], Congress chose to provide a separate monetary remedy to a subset of domestic producers that were the most seriously injured by foreign unfair trade practices, and it rationally assumed that this subset of most-harmed producers would be those producers that had supported the petition.” Br. of Defendant-Appellant U.S. Customs & Border Protection at 20-21 (emphasis added). At oral argument, the government expressly and repeatedly rejected the court’s suggestion that the Byrd Amendment was intended to reward parties for assisting the government. See Oral Arg. at 14:25-31,15:18-23 available at http://oralarguments.cafc.uscourts.gov/mp 3/2008-1005.mp3 (government responding to question about purpose of Byrd Amendment “to reward people who bring these antidumping petitions and those who support the petition” by stating that “[t]he purpose of this classification should not really be seen as one of rewarding”); id. at 25:15-31 (“There is nothing in that statute, your honor, that indicates any attempt to reward parties as opposed to provide a subsidy to American manufacturers who have been injured.”); id. at 1:03:42-1:04:36 (“The parties seem to be in agreement that this statute and the classification is really not that similar to the situation of relators in qui tarn cases where they are providing a service to the government and receiving some amount by statute as a reward for having brought to the attention of the government fraud, waste, and abuse.”); see also id. at 23:34-24:03 (“It was ... apparent on the face of the findings of Congress that preceded the [Byrd Amendment] and also the floor statements of Senator DeWine and Senator Byrd that this was intended to be a remedial statute that was going to aid members of domestic industry that continue to be injured by dumping....”).
The majority dismisses the government’s repeated statements rejecting the “reward for assistance” rationale.4 Specifically, the majority argues that “the views of the government as litigator are simply not binding on the issue of Congressional intent.” Maj. Op. at 1352. But — as the majority’s own parenthetical summaries *1367make clear — the cases that the majority cites for that proposition all address the views of the government as to the proper interpretation of ambiguous statutory language, not to the asserted purpose of a statute for purposes of constitutional scrutiny. See Id. (“Cherokee Nation of Okla. v. Leavitt, 543 U.S. 631, 646-57, 125 S.Ct. 1172, 161 L.Ed.2d 66 (2005) (recognizing and then rejecting the government’s interpretation of a statute); United States v. Reorganized CF & I Fabricators of Utah, Inc., 518 U.S. 213, 223, 116 S.Ct. 2106, 135 L.Ed.2d 506 (1996) (rejecting the government’s interpretation of a tax statute)” (emphases added)). The Supreme Court has recognized that it is the government’s “asserted” purpose that is relevant in assessing the constitutionality of a statute — • i.e., the purpose that the government as litigator asserts to justify the statute in the face of a constitutional challenge. See, e.g., City of Erie v. Pap’s A.M., 529 U.S. 277, 296, 120 S.Ct. 1382, 146 L.Ed.2d 265 (2000) (“The asserted interests ... are undeniably important”); Texas v. Johnson, 491 U.S. 397, 407, 109 S.Ct. 2533, 105 L.Ed.2d 342 (1989) (“[W]e must decide whether Texas has asserted an interest in support of Johnson’s conviction that is unrelated to the suppression of expression. ...” The State offers two separate interests to justify this conviction....) Thus, the burden is on the government — in litigation' — to identify the interest served by the regulation and to prove that it is “compelling,” “substantial,” or “important.” See, e.g., Boos, 485 U.S. at 321, 108 S.Ct. 1157 (“[W]e have required the State to show that the regulation is necessary to serve a compelling state interest and that it is narrowly drawn to achieve that end.” (internal quotation marks omitted and emphasis added)). In fact, in Central Hudson—the very case that establishes the commercial speech test that the majority applies — the Supreme Court made clear that it is the interest that the government asserts in litigation challenging a regulation that is relevant for the constitutional inquiry:
In commercial speech cases, then, a four-part analysis has developed. At the outset, we must determine whether the expression is protected by the First Amendment. For commercial speech to come within that provision, it at least must concern lawful activity and not be misleading. Next, we ask whether the asserted governmental interest is substantial. If both inquiries yield positive answers, we must determine whether the regulation directly advances the governmental interest asserted, and whether it is not more extensive than is necessary to serve that interest.
Central Hudson, 447 U.S. at 566, 100 S.Ct. 2343 (emphases added). Thus, it is the government’s asserted purpose — not the “reward for assistance” purpose expressly rejected by the government — that is relevant to the First Amendment analysis here. It is not the role of the court to substitute its judgment for that of the government and to decide which interest the government should have “asserted.”
The majority also relies heavily on the doctrine of constitutional avoidance, which it suggests “extends to the ascertainment of a statute’s purpose.” Maj. Op. at 1353. The well-established “canon of constitutional avoidance” holds that “[wjhere a possible construction of a statute would render the statute unconstitutional, courts must construe the statute ‘to avoid such problems unless such construction is plainly contrary to the intent of Congress.’ ” Consolidation Coal Co. v. United States, 528 F.3d 1344, 1347 (Fed.Cir.2008) (quot*1368ing Edward J. DeBartolo Corp., 485 U.S. at 575, 108 S.Ct. 1392). The doctrine of constitutional avoidance is a doctrine of statutory interpretation — that is, it is relevant when the court is construing disputed statutory language. See, e.g., Fisherman’s Harvest, Inc. v. PBS & J, 490 F.3d 1371, 1377 (Fed.Cir.2007) (discussing “canon of constitutional avoidance in statutory interpretation”). In this case, there is no statutory construction to be performed. The parties do not dispute the meaning of the petition support requirement, and the parties do not dispute that, if the petition support requirement is constitutional, it was correctly applied to SKF. There is therefore no statutory construction dispute, and the doctrine of constitutional avoidance is irrelevant.
The majority, however, reasons that the doctrine of constitutional avoidance “extends to the ascertainment of a statute’s purpose.” Maj. Op. at 1353. That is, in the majority’s view, when evaluating whether a statute serves a compelling, substantial, or important government interest, the court should look not to the interest that is clear from the statutory text or legislative history, nor to the interest that the government actually puts forward during litigation, but rather to any interest that “would make the statute constitutional.” Id. at 1354. I respectfully disagree. While it is proper under rational basis review to evaluate whether any hypothetical interest would render a statute constitutional, under the heightened scrutiny required by the First Amendment, we evaluate only the government’s actual, asserted interest. See, e.g., Thompson v. W. States Med. Ctr., 535 U.S. 357, 373-74, 122 S.Ct. 1497, 152 L.Ed.2d 563 (2002) (“The dissent describes another governmental interest.... Nowhere in its briefs, however, does the Government argue that this interest motivated the advertising ban. Although, for the reasons given by the dissent, Congress conceivably could have enacted the advertising ban to advance this interest, we have generally only sustained statutes on the basis of hypothesized justifications when reviewing statutes merely to determine whether they are rational. The Central Hudson test is significantly stricter than the rational basis test ....” (citations omitted)); Edenfield v. Fane, 507 U.S. 761, 768, 113 S.Ct. 1792, 123 L.Ed.2d 543 (1993) (“Unlike rational-basis review, the Central Hudson standard does not permit us to supplant the precise interests put forward by the State with other suppositions.”). I cannot agree that the doctrine of constitutional avoidance allows us to ignore the government’s asserted purpose and substitute our own when heightened First Amendment scrutiny applies.
Moreover, though the majority cites two cases for its theory that the doctrine of constitutional avoidance extends to the “ascertainment of a statute’s purpose,” Maj. Op. at 1353, those cases actually involve the interpretation of statutory language— not the government interest served by the statute. In the pre-Lochner Delaware & Hudson case on which the majority principally relies, the Supreme Court did apply the principle of constitutional avoidance and make reference to the government’s view concerning the “result intended to be accomplished” by the statutory provision at issue, but it did so solely for the purpose of construing disputed statutory language. See U.S. ex rel Attorney Gen. v. Del. & Hudson Co, 213 U.S. 366, 404-05, 29 S.Ct. 527, 53 L.Ed. 836 (1909) (“Let us, as a prelude to an analysis of the [statutory] clause, for the purpose of fixing its true construction, and determining the constitutional power to enact it when its signifi-*1369canee shall have been rightly defined, point out the questions of constitutional power which will require to be decided if the construction relied upon by the government is a correct one.”). Likewise, the Zadvydas case on which the majority relies considered the doctrine of constitutional avoidance solely for the purpose of statutory construction. See Zadvydas v. Davis, 533 U.S. 678, 689-90, 121 S.Ct. 2491, 150 L.Ed.2d 653 (2001) (“[W]e read an implicit limitation into the statute before us. In our view, the statute, read in light of the Constitution’s demands, limits an alien’s postremoval-period detention to a period reasonably necessary to bring about that alien’s removal from the United States.... A statute permitting indefinite detention of an alien would raise a serious constitutional problem.”). I am aware of no case in which the Supreme Court has applied the doctrine of constitutional avoidance — as the majority does here — to determine the asserted purpose of an unambiguous statute in a constitutional challenge.
It also seems to me that the “reward for assistance” rationale for the Byrd Amendment makes little sense in light of the regulations governing the conduct of anti-dumping investigations. All members of the domestic industry who receive a questionnaire — whether they support the petition or not — are required to complete the questionnaire and to certify to its accuracy. Moreover, the ITC has the authority to subpoena any additional information that it needs from otherwise unwilling companies. See 19 U.S.C. § 1333(a), (f); Producers’ Questionnaire at 1. This explains why the government admitted at oral argument that petition supporters and petition opponents provide exactly the same assistance to the government in anti-dumping investigations. See Oral Arg. 22:41-23:07 (“[WJhat the government obtains form the questionnaire responses is the same for those who supported and thereby are eligible under the classification the [Byrd Amendment] to receive these funds and for those who opposed or took no position. So, they are also aiding the government in a government function in that respect. Certainly that is true.” (emphasis added)); see also id. at 18:50-19:10 (“[Companies that do not support the petition] are required by law to respond to the questionnaire in the same way that those who have answered the question checking support are required to do so.”); id. at 20:10-28 (“[The Court:] Is there something different that parties who support the petition provide to the government as compared to parties that don’t support the petition? [The government:] Not that I’m aware of. I don’t think that that is an important distinction....”).
The facts of this case illustrate why the purpose of the Byrd Amendment cannot have been the reward for assistance rationale that the majority suggests. The majority details the submissions that petitioner Torrington and petition supporters made during the investigation that led to the antidumping order in this case. See Maj. Op. at 1342 (noting that “the petition was over 200 pages in length”); id. at 1343-44 (“The questionnaire responses of these petition supporters were hundreds of pages long, and several of the supporters prepared responses exceeding 300 pages.”); id. at 11 (“Petitioner Torring-ton’s pre-hearing brief was over 200 pages long....”). The majority also notes that “SKF also responded to the ITC’s questionnaire, but stated that it opposed the antidumping petition,” id. at 1343, but what the majority fails to point out is that SKF’s questionnaire responses also totaled more than 200 pages. See Preliminary and Final Responses of SKF USA, Inc. *1370(242 pages). In fact, if, as the majority’s analysis suggests, assistance in an anti-dumping investigation can be measured in part by the page length of questionnaire responses, SKF was actually more helpful that several supporters of the petition that have received distributions under the Byrd Amendment. See, e.g., Response of Emerson Power Transmission Co. (122 pages). Likewise, while the majority claims that Torrington’s briefing “assisted in the investigation,” Maj. Op. at 1343, it fails to acknowledge that SKF also submitted briefing — totaling 163 pages — during the investigation.
To the extent that the majority recognizes SKF’s participation in the antidump-ing investigation, the majority sees it as evidence against SKF, going so far as to suggest that SKF “impede[d] the investigation” by opposing it. See Id. at 39-40. However, SKF did nothing to impede, and merely expressed its view that the domestic industry was not being or about to be materially injured by the alleged dumping. SKF, like other petition opponents, submitted expert analysis and briefing supporting that view to the ITC. The majority does not suggest that SKF withheld any information or submitted any evidence or argument in bad faith. To the contrary, SKF’s only “fault” was that the ITC ultimately disagreed with it and concluded that the domestic industry was, in fact, harmed — a decision that the ITC had not yet made at the time SKF opposed the petition, and a decision that we have held is firmly committed to the ITC’s discretion. See Nucor, 414 F.3d at 1336 (noting ITC’s “broad discretion” in assessing material injury). If taking an opposing view in a proceeding were tantamount to “impeding” an investigation, then every losing party in every action to which the government is a party (not to mention every criminal defense attorney) would be guilty of obstruction. SKF, acting in good faith, assisted in the antidumping investigation by complying with its obligation to submit detailed questionnaire responses, by submitting expert evidence and briefing, and by providing its honest viewpoint to the ITC. The only difference between SKF and the petition supporters was that SKF thought that the ITC should have come to a different conclusion. This illustrates precisely why rewarding petition supporters for their assistance in an investigation cannot have been the Byrd Amendment’s purpose.
My fourth disagreement with the majority concerns its conclusion that the First Amendment test for commercial speech “seems appropriate” in this case. Citing Central Hudson, the majority concludes that the Supreme Court has “broadly defined ‘commercial speech’ as ‘expression related solely to the economic interest of the speaker and its audience.’ ” Maj. Op. at 1355. But Central Hudson did not concern whether the speech at issue — advertising by an electric company — was or was not commercial. The parties agreed that the speech was commercial. Central Hudson, 447 U.S. at 560-61, 100 S.Ct. 2343. The case in which the Supreme Court actually considered the definition of commercial speech came three years later. In Bolger v. Youngs Drug Products Corp., the Supreme Court considered whether informational pamphlets distributed by a contraceptive manufacturer and promoting the use of prophylactics were commercial speech. 463 U.S. 60, 62, 65-66, 103 S.Ct. 2875, 77 L.Ed.2d 469 (1983). The Supreme Court recognized that “the core notion of commercial speech [is] speech which does no more than propose a commercial transaction.” Id. at 66, 103 S.Ct. 2875 (internal quotation marks omitted). *1371Nevertheless, the Supreme Court concluded that the pamphlets were commercial speech:
The mere fact that these pamphlets are conceded to be advertisements clearly does not compel the conclusion that they are commercial speech. Similarly, the reference to a specific product does not by itself render the pamphlets commercial speech. Finally, the fact that [the manufacturer] has an economic motivation for mailing the pamphlets would clearly be insufficient by itself to turn the materials into commercial speech.
The combination of all these characteristics, however, provides strong support for the District Court’s conclusion that the informational pamphlets are properly characterized as commercial speech.
Id. at 66-67, 103 S.Ct. 2875 (citations omitted). The Court went on to say that “[a] company has the full panoply of protections available to its direct comments on public issues, so there is no reason for providing similar constitutional protection when such statements are made in the context of commercial transactions.” Id. at 68, 103 S.Ct. 2875. Thus, as one commentator has put it, speech is commercial under Bolger if: “(1) [i]t is an advertisement of some form, (2) it refers to a specific product, and (3) the speaker has an economic motivation for the speech.” Erwin Chemerinsky, Constitutional Law: Principles and Policies § 11.3.7.2.
The speech affected by the petition support clause is not commercial speech under Bolger. A statement compelled in response to an ITC questionnaire is not an advertisement, nor does it refer to a specific product. SKF may have had “an economic motivation” for answering the questionnaire, but, as Bolger makes clear, “an economic motivation ... would clearly be insufficient by itself to turn [speech] into commercial speech.” Bolger, 463 U.S. at 67, 103 S.Ct. 2875; see also Bigelow v. Virginia, 421 U.S. 809, 818, 95 S.Ct. 2222, 44 L.Ed.2d 600 (1975) (“The State was not free of constitutional restraint merely ... because appellant’s motive or the motive of the advertiser may have involved financial gain. The existence of commercial activity, in itself, is no justification for narrowing the protection of expression secured by the First Amendment.” (citations and internal quotation marks omitted)). To the contrary, SKF’s response to the question “Do you support or oppose the petition?” is precisely the kind of “direct comment[ ] on public issues” for which it has “the full panoply of protections available” under the First Amendment.5
Moreover, even if the majority were correct that the test for commercial speech is whether the regulated “expression re*1372late[s] solely to the economic interests of the speaker and its audience,” Maj. Op. at 1355, I cannot agree that this test is satisfied here. The majority reasons that “[Rewarding parties under the circumstances here is similar to commercially contracting with them to assist in the performance of a government function, in this particular context assisting in the enforcement of government policy in litigation.” Id. The majority’s analysis, however, does not actually address the speech at issue. The petition support clause conditions receipt of funds on expressing support for an antidumping petition. The question is whether that regulated expression — namely, expressing support for an antidumping petition — “relatefs] solely to the economic interests of the speaker and its audience.” The majority’s view that companies who support a petition are more likely to provide assistance to the government and therefore enter into a quasi-contractual relationship is immaterial. The issue is simply whether the expression of support itself relates solely to the economic interests of the company and the audience. Even setting aside whether a statement of support for a petition reflects the economic interests of the company making a statement, it cannot be said that the petition support requirement relates solely to the economic interest of the audience — here, the ITC. The ITC had no economic interest in whether SKF expressed support for the petition or did not. Thus, I cannot agree that the majority’s “commercial contract” analogy, even if correct, would support application of the commercial speech doctrine under Central Hudson.
Further, it is noteworthy that the majority’s view that the commercial speech test “seems appropriate” is not a view shared by any party to this case. Nowhere in any of its briefing does either the government or the ITC argue that the commercial speech doctrine is applicable. Moreover, appellant Timken expressly argues that the commercial speech doctrine is not applicable. See Response-Reply Br. of Defendanfi-Appellant Timken U.S. Corporation at 41 n. 48 (“Providing factual information to the ITC bears no resemblance to the concept of commercial speech, and, by definition, the [Byrd Amendment] does not involve the regulation of commercial speech, which has generally been defined as ‘speech proposing a commercial transaction.’ ”). I agree with the parties that the commercial speech doctrine is inapplicable.
Fifth, even if the majority were correct that Central Hudson’s, test for the constitutionality of commercial speech were the correct test, I cannot agree with the majority that the petition support requirement would survive that test. In Central Hudson, the Supreme Court held that:
At the outset, we must determine whether the expression is protected by the First Amendment. For commercial speech to come within that provision, it at least must concern lawful activity and not be misleading. Next, we ask whether the asserted governmental interest is substantial. If both inquiries yield positive answers, we must determine whether the regulation directly advances the governmental interest asserted, and whether it is not more extensive than is necessary to serve that interest.
Central Hudson, 447 U.S. at 566, 100 S.Ct. 2343. Even assuming that the other elements of the Central Hudson test could be met, the petition support requirement cannot satisfy the final element, because it cannot be said to be “not more extensive than is necessary to serve [the asserted] interest.” Id.
*1373If, as the majority reasons, the government interest furthered by the petition support requirement is “to reward injured parties who assisted government enforcement” in antidumping investigations, Maj. Op. at 1352, then the petition support requirement is far more extensive than is necessary to serve that interest. A much more straightforward method of ensuring the cooperation of private parties in anti-dumping investigations would be simply for the ITC to compel the cooperation of uncooperative parties through the subpoena process — as it already has the authority to do. See 19 U.S.C. § 1333(a), (f). Moreover, to the extent that Congress intended to compensate parties for the expense of preparing petitions or questionnaire responses as the majority suggests, see Maj. Op. at 1358-59, it could simply authorize reimbursement of reasonably incurred expenses to parties that cooperate willingly — a far less restrictive measure than precluding petition opponents from receiving any remedial duties. Reasoning that Central Hvidson does not require a perfect fit between means and ends, the majority argues that “[t]hose who support antidump-ing petitions typically fill out questionnaires from the ITC.” Id. at 1358. But the majority ignores that all recipients of questionnaires are required to complete them — whether they support or oppose the petition. Indeed, by definition, a party excluded from receiving disbursements as a result of checking the “Oppose” box in response to questionnaire question 1-3 has necessarily filled out the questionnaire. While Central Hudson may not require a perfect correspondence of means and ends, I cannot agree that the petition support requirement places is “not more extensive than necessary” to the furtherance of an alleged interest in rewarding cooperation in an antidumping investigation.6
Even if the interest served by the petition support requirement were the interest identified by the government — namely, “providing] a separate monetary remedy to a subset of domestic producers that were the most seriously injured by foreign unfair trade practices,” Br. of Defendant-Appellant U.S. Customs & Border Protection at 20-21 — I would still conclude that it fails to satisfy the final element of the Central Hudson test. It may be true that a party that is more seriously injured by dumping is more likely to check the “Support” box in response to question 1-3 than a party that is less seriously injured. But Central Hudson requires more: that the regulation be not more extensive than necessary.
If the government interest is to compensate the most seriously injured domestic producers, the ITC could look to the detailed financial data provided in response to the rest of the questionnaire, determine for itself which producers are most seriously injured, and distribute collected duties accordingly. Because these better proxies exist (and, in fact, are already part of the questionnaire), the government is wrong to assert that a company’s response to question 1-3 places a burden on speech that is not more extensive than necessary to accomplish its goal of compensating the most seriously injured producers.
*1374Not only is the petition support requirement not the best proxy for injury, it is not even a particularly good one. As the Court of International Trade found, “there are a multitude of reasons why an entity might decide to support, oppose, or take no position in an antidumping investigation” that are unrelated to the seriousness of its injury. SKF USA Inc. v. United States, 451 F.Supp.2d 1355, 1362 (Ct. Int’l Trade 2006). A domestic company, like SKF, that is a subsidiary of a foreign importer, might oppose a petition because it concludes that the worldwide injury caused to its parent by an antidumping duty order would be greater that the injury that the subsidiary suffers domestically as a result of dumping. More altruistically, a company might simply have the ideological view that any restrictions on trade — including restrictions on dumping — are bad. It might be willing to endure serious injury resulting from dumping, rather than support a petition that would, in its view, restrict free trade. Finally, as was the case with amicus Giorgio Foods, Inc., a domestic producer might oppose a petition to protect business relationships in foreign countries having nothing to do with the domestic market, or it might decline to support a petition for fear of retaliation in export markets. See Br. of Amicus Curiae Giorgio Foods, Inc. & PS Chez Sidney LLC at 2, 9-10. To conclude summarily, as the government does, that the petition support requirement identifies the most seriously injured domestic producers evinces a naive view of the economics of international trade. Thus, I cannot conclude that the final element of the Central Hudson test would be met, applying either the majority’s or the government’s asserted government interest.7
Sixth and finally, the majority’s analogy to qui tam actions is simply inapposite. The primary federal qui tam statute is the False Claims Act, 31 U.S.C. §§ 3729-33. Under § 3730(b) of the False Claims Act, a private person may bring an action in the name of the government against a defendant believed to have knowingly presented a false or fraudulent claim for payment to the government. If the defendant is proven to have presented a false claim, the defendant is liable to the government, and the private party who initiated the action may receive up to a thirty percent share of the proceeds of the action or settlement, and reasonable expenses, costs, and attorneys fees. Id. § 3730(d). See generally Cook County, Ill. v. United States ex rel. Chandler, 538 U.S. 119, 122-23, 123 S.Ct. 1239, 155 L.Ed.2d 247 (2003) (describing qui tam provisions of False Claims Act). Qui tam actions for false patent marking work in the same way. See 35 U.S.C. § 292(b) (“Any person may sue for the penalty [of $500 per offense for falsely marking an article as patented or ‘patent pending’], in which event one-half shall go to the person suing and the other to the use of the United States.”).
The majority analogizes antidumping proceedings to qui tam actions, reasoning *1375that the operation of the Byrd Amendment, like a qui tam proceeding, “reward[s] private parties for successfully bringing suit on behalf of the government.” Maj. Op. at 1356. It carries this analogy further, arguing that “SKF here undertook a role that was nearly indistinguishable from that played by a defendant in a qui tam or attorney’s fees award case” by opposing the antidumping petition. Id. at 1358-59. To be sure, Commerce and the ITC rely on private companies — namely, members of the domestic industry that have been harmed by dumping — to bring dumping to its attention through the petition process. But beyond that, the analogy to qui tam proceedings fails. In a qui tam case, the defendant has committed a violation of the law that causes harm to the government, and the government shares its recovery with the plaintiff — an uninjured third party. In other words, a qui tam action is an action by a representative (the plaintiff) against a wrongdoer (the defendant), on behalf of a victim (the government).
By contrast, when a foreign company improperly dumps goods in the domestic market, the foreign company is the wrongdoer, and its victims are the members of the domestic industry. A petitioner brings an action on behalf of the injured domestic industry. Thus, an antidumping action by petition is an action by a representative (the petitioner) against a wrongdoer (the foreign company), on behalf of victims (the members of the domestic industry). In summary form:
[[Image here]]
Two obvious distinctions are apparent. First, in qui tam proceedings it is the government — the victim — that willingly elects to share a portion of its compensation with the uninjured plaintiff representative, essentially as a bounty for bringing the action. But in an antidumping investigation, the government is not the injured party. It is the members of the domestic industry- — not only including the petitioner and petition supporters, but also including all other domestic producers — that are injured and entitled to compensation for its injury.
Second, the majority is wrong to equate SKF to a defendant in a qui tam action. The defendant in a qui tam action is the wrongdoer — the company that violates the law. In an antidumping investigation, the role of the qui tam defendant is played by the foreign company that does the dumping. SKF was a victim of that dumping as one of the injured members of the domestic industry. In sum, the majority’s analogy to qui tam proceedings is simply inap-posite, and it cannot shield the petition support requirement from First Amendment scrutiny. The government recognized as much at oral argument, remarking that “[t]he parties seem to be in agreement that this statute and the classification is really not that similar to the situation of [plaintiffs] in qui tam cases.” Oral Arg. at 1:03:42-1:04:36.
Ill
In my view, the petition support requirement should be subjected to strict scrutiny as a content-based restriction on political speech in a public forum. The principles *1376that control the outcome of this case are beyond serious dispute. First, “a content-based, restriction on political speech in a public forum ... must be subjected to the most exacting scrutiny. Thus, [the government must] show that the ‘regulation is necessary to serve a compelling state interest and that it is narrowly drawn to achieve that end.’ ” Boos v. Barry, 485 U.S. 312, 321, 108 S.Ct. 1157, 99 L.Ed.2d 333 (1988) (quoting Perry Educ. Ass’n v. Perry Local Educators’ Ass’n, 460 U.S. 37, 45, 103 S.Ct. 948, 74 L.Ed.2d 794 (1983)). Second, “[w]hen the government targets not subject matter, but particular views taken by speakers on a subject, the violation of the First Amendment is all the more blatant. Viewpoint discrimination is thus an egregious form of content discrimination.” Rosenberger v. Rector of Univ. of Va., 515 U.S. 819, 829, 115 S.Ct. 2510, 132 L.Ed.2d 700 (1995) (citation omitted). Third, under the so-called “unconstitutional conditions” doctrine, the government “may not deny a benefit to a person on a basis that infringes his constitutionally protected interests — especially, his interest in freedom of speech.” Perry v. Sindermann, 408 U.S. 593, 597, 92 S.Ct. 2694, 33 L.Ed.2d 570 (1972).
Taken together, these principles establish — at an absolute minimum — that a regulation is subject to strict scrutiny if it denies a benefit on the basis of expression of a specific viewpoint on a political matter in a public forum. Cf. Lac Vieux Desert Band of Lake Superior Chippewa Indians v. Mich. Gaming Control Bd., 172 F.3d 397, 409-10 (6th Cir.1999) (holding that ordinance that “grants benefits and imposes burdens according to whether an individual or entity sufficiently supported a particular political issue” was subject to strict scrutiny). I would conclude that the petition support requirement denies a benefit on the basis of expression of a viewpoint on a political matter in a public forum, and is therefore subject to strict scrutiny.
First, in my view, the petition support requirement is viewpoint discriminatory. Under the petition support requirement, a domestic company is ineligible for a distribution unless the company was in “support of the petition,” as indicated by its response to question 1-3, “Do you support or oppose the petition?” 19 U.S.C. § 1675c(b)(l)(A); Producers’ Questionnaire at 2. Domestic companies who express the viewpoint that an antidumping order should issue are eligible; companies who do not express that viewpoint are not. This is classic viewpoint discrimination. As discussed in detail above, see supra at 9, it is immaterial that the government does not intend to suppress a particular viewpoint. It is the viewpoint-discriminatory effect of the statute that offends the First Amendment.
Second, the petition support requirement affects political speech. “Political speech, of course, is ‘at the core of what the First Amendment is designed to protect.’ ” Morse v. Frederick, 551 U.S. 393, 127 S.Ct. 2618, 2626, 168 L.Ed.2d 290 (2007) (quoting Virginia v. Black, 538 U.S. 343, 365, 123 S.Ct. 1536, 155 L.Ed.2d 535 (2003)). Moreover, political speech is not merely advocacy on behalf of a particular candidate. Rather, it encompasses “the free discussion of governmental affairs. This of course includes discussions of candidates, structures and forms of government, the manner in which government is operated or should be operated, and all such matters relating to political processes.” Mills v. Alabama, 384 U.S. 214, 218-19, 86 S.Ct. 1434, 16 L.Ed.2d 484 (1966). In this case, the petition support require*1377ment requires that a domestic company-have expressed the view that a duty should be imposed on a specific class of foreign goods, based in part on whether the domestic industry has been or will be “materially injured.” See 19 U.S.C. § 1673. Notably, not only does the ITC consider the views of opponents when deciding whether the material injury requirement has been met, but § 1673 actually precludes a petition from going forward unless it has support from 25% of the domestic industry by production, and no more than 50% in opposition. See 19 U.S.C. § 1673 a(c)(4)(A)(i)-(ii). Taking a position on this question before the ITC — the body charged with determining whether there has been or will be material injury — is therefore not only political speech on an issue of public concern, but effectively a vote on whether the petition should go forward. It is therefore plainly political speech at the core of the First Amendment’s protection.
For the same reasons that it affects political speech, the petition support requirement implicates the First Amendment’s Petition Clause. As the Supreme Court has explained:
The right to petition is cut from the same cloth as the other guarantees of [the First] Amendment, and is an assurance of a particular freedom of expression. In United States v. Cruikshank, 2 Otto 542, 92 U.S. 542, 23 L.Ed. 588 (1876), the Court declared that this right is implicit in “[t]he very idea of government, republican in form.” Id., at 552. And James Madison made clear in the congressional debate on the proposed amendment that people “may communicate their will” through direct petitions to the legislature and government officials. 1 Annals of Cong. 738 (1789).
McDonald v. Smith, 472 U.S. 479, 482, 105 S.Ct. 2787, 86 L.Ed.2d 384 (1985). By conditioning receipt of a benefit on the expression of a particular view to the governmental agency charged with making a decision, the petition support requirement necessarily impedes companies from “com-municat[ing] their will” to the relevant government officials.
Third, the petition support requirement affects speech in a designated public forum. The government creates a designated public forum when it makes a space “generally available to a certain class of speakers.” Ark. Educ. Television Comm’n v. Forbes, 523 U.S. 666, 679, 118 S.Ct. 1633, 140 L.Ed.2d 875 (1998). The ITC not only creates a designated public forum for domestic producers by inviting them to share their views on a petition, but it in fact requires them to do so. See 19 U.S.C. § 1333(a), (f); Producers’ Questionnaire at 1. Moreover, the petition support requirement requires both that the domestic producer support the petition, and that it allow its support to be publicly known. See 19 U.S.C. § 1675c(d) (requiring publication of qualified recipients of distributions). I would conclude in these circumstances that an ITC proceeding is a limited public forum for speech by domestic producers.
We are, of course, not the first court of appeals to consider the constitutionality of a government regulation that provides a benefit to a party as a reward for prior political expression. The closest analogous case in the regional circuits is the Sixth Circuit’s decision in Lac Vieux, 172 F.3d 397, appeal after remand 276 F.3d 876 (6th Cir.2002), cert. denied, 536 U.S. 923, 122 S.Ct. 2589, 153 L.Ed.2d 779 (2002). In that case, two casino developers had spent substantial sums of money to advertise and *1378promote the passage of a ballot initiative to legalize casino gambling in Detroit, Michigan. Id. at 400. After the ballot measure passed, the Detroit City Council adopted an ordinance giving preference for casino licenses to developers who had actively promoted the ballot initiative. Id. at 401. The Lae Vieux Desert Band of Lake Superior Chippewa Indians — a potential casino developer that had not lobbied for passage of the ballot initiative but wanted a casino license — challenged the ordinance on First Amendment grounds. Id. at 402. The Sixth Circuit held that the ordinance “impose[d] a burden based on the content of political speech” and that the ordinance was content based and therefore subject to strict scrutiny. Id. at 409-10.
The majority dismisses Lac Vieux in a footnote, reasoning that it “did not reward the achievement of the enforcement of government policy through litigation, but instead involved ‘political support’ for legislative efforts.” Maj. Op. at 1356 n. 32. I agree that an ITC investigation is not an election by ballot initiative, but I do not think that this distinction is of any significance. “[T]he free discussion of governmental affairs” protected by the First Amendment encompasses more than merely campaigning. Mills, 384 U.S. at 218, 86 S.Ct. 1434. Moreover, because the ITC requires domestic producers to provide their views on a petition and is required to take those views into account, the petition support requirement, like the ordinance in Lac Vieux, does concern a company’s “political support” for a proposition (as in Lac Vieux) or a petition (as in this case).
I would conclude that because the petition support requirement is viewpoint discriminatory toward political speech in a public forum, it is subject to strict scrutiny. To survive, it must be “necessary to serve a compelling state interest and ... narrowly drawn to achieve that end.” Perry, 460 U.S. at 45, 103 S.Ct. 948. Even assuming that the interests asserted by the majority (reward for assistance) and the government (remedy for the most seriously injured domestic producers) were compelling, I cannot conclude that the petition support requirement is narrowly drawn to achieve either. As discussed in detail above, less restrictive means exist to achieve either interest. See supra at 23-26. I would therefore conclude that the petition support requirement is unconstitutional.
IV
Because I would affirm the judgment of the Court of International Trade that the petition support requirement is unconstitutional,8 I briefly address the remaining issues concerning severance and SKF’s amended certification.
A. Severance
“[WJhenever an act of Congress contains unobjectionable provisions separable from those found to be unconstitutional, it is the duty of this court to so declare, and to maintain the act in so far as it valid.” El Paso & N.E. Ry. Co. v. Gutierrez, 215 U.S. 87, 96, 30 S.Ct. 21, 54 L.Ed. 106 (1909). *1379Timken argues that even if the petition support requirement is unconstitutional, the Court of International Trade erred by severing the statute so that opponents of a petition were eligible for benefits. Instead, Timken contends that the statute should be severed so that only petitioners — not any other “interested part[ies] in support of the petition” — would be eligible for distributions.
There are two problems with Timken’s proposed approach. First, it would run contrary to Congress’s intent, clear from the face of the statute, to distribute collected duties to “affected domestic producers.” “[T]he touchstone for any decision about remedy is legislative intent, for a court cannot ‘use its remedial powers to circumvent the intent of the legislature.’ ” Ayotte v. Planned Parenthood, 546 U.S. 320, 330, 126 S.Ct. 961, 163 L.Ed.2d 812 (2006) (quoting Califano v. Westcott, 443 U.S. 76, 94, 99 S.Ct. 2655, 61 L.Ed.2d 382 (1979) (Powell, J., concurring in part and dissenting in part)). Here, Congress’s intent is clear from the overall structure of the Byrd Amendment. The Byrd Amendment authorizes distributions to “affected domestic producers.” 19 U.S.C. §§ 1675c(b)-(d). The petition support requirement is only one of several parts of the definition of “affected domestic producers” — -an “affected domestic producer” must also be a “manufacturer, producer, farmer, rancher or worker representative (including associations of such person)” and must “remain[] in operation.” Id. § 1675e(b)(l). Additionally, a producer that has “ceased the production of the product covered by the order or finding” is excluded from the statutory definition of “affected domestic producer.” Id.
Plainly, Congress intended to distribute funds collected as a result of antidumping duty orders to more “affected domestic producers” than simply the petitioner who initiated the action. If Congress had intended to limit distributions to petitioners, the statute simply would have authorized distributions to “petitioners.” There would be no need for an elaborate definition of “affected domestic producer,” with its various requirements and exclusions. Congress’s intent therefore must necessarily have been not to reward petitioners for assistance, but to provide a monetary remedy to injured members of the domestic injury, to offset the injuries caused by dumping. In fact, the very title of the Byrd Amendment — the Continued Dumping and Subsidy Offset Act of 2000— evinces to this purpose. See also 145 Cong. Rec. S497, 497 (1999) (statement of Sen. DeWine) (“Current law also does not contain a mechanism to help injured U.S. industries recover from the harmful effects of foreign dumping and subsidization.”). It would be inconsistent with this intent to remedy the constitutional defects in the Byrd Amendment by limiting recovery to petitioners.
The second problem with Timken’s approach is that it would not actually cure the First Amendment defect of the petition support requirement. Notably, Timken made its severance argument in the context of a finding by the Court of International Trade that the petition support requirement violated the Equal Protection Clause because it was not rationally related to a legitimate government purpose. While it might be true that limiting distributions to petitioners only — rather than petitioners and parties that supported the petition — might cure any problem that the petition support requirement had overcoming the rational basis test, the statute would still fail strict scrutiny under the First Amendment, even if severed as *1380Timken proposed, because it would still condition the receipt of funds on expression of a political viewpoint and petitioning activity — namely, filing a petition that argues that an antidumping duty order should enter. Moreover, the statute would still fail strict scrutiny, because less restrictive means are available to serve the interests identified by the majority (reward for assistance) and the government (remedy for the most seriously injured domestic producers). Thus, I would conclude that the Court of International Trade properly severed the Byrd Amendment by removing the petition support requirement.
B. SKF’s Cross Appeal
SKF argues on cross appeal that the Court of International Trade was wrong to hold that Customs was not required to accept SKF’s amended certification for fiscal year 2005 distributions under the Byrd Amendment. Customs rejected SKF’s amended certification as untimely. “[T]his court reviews the trial court’s decision de novo, reapplying the same standard utilized by that court” — here, the standard of review under the Administrative Procedure Act. Consol. Bearings Co. v. United States, 348 F.3d 997, 1004 (Fed.Cir.2003). Under the Administrative Procedure Act:
The reviewing court shall — ... hold unlawful and set aside agency action, findings, and conclusions found to be
(A) arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law;
(B) contrary to constitutional right, power, privilege, or immunity;
(C) in excess of statutory jurisdiction, authority, or limitations, or short of statutory right;
(D) without observance of procedure required by law; [or]
(E) unsupported by substantial evidence ....
5 U.S.C. § 706 (2006).
The timing of certifications pursuant to the Byrd Amendment is governed by Treasury regulations. “At least 90 days before the end of a fiscal year, Customs will publish in the Federal Register a notice of intention to distribute assessed duties received as the continued dumping and subsidy offset for that fiscal year.” 19 C.F.R. § 159.62(a). That notice contains instructions for filing a certification to claim a distribution. Id. § 159.62(b)(2). “In order to obtain a distribution of the offset, each affected domestic producer must submit a certification ... that must be received within 60 days after the date of publication of the notice in the Federal Register, indicating that the affected domestic producer desires to receive a distribution. The certification must enumerate the qualifying expenditures incurred by the domestic producer since the issuance of an order or finding for which a distribution has not previously been made....” Id. § 159.63(a) (emphasis added).
SKF admits that the certification that it submitted within the sixty-day time frame contained expenditure data only for a one manufacturing facility. It did not seek to amend its certification until after the Court of International Trade held that the petition support requirement was unconstitutional. SKF admits that its amended certification was untimely, but argues in essence that submitting a complete certification would have been futile, because “it was a foregone conclusion that Customs would reject SKF[’s] certification.” Br. of Plaintiff-Cross Appellant SKF USA Inc. *1381at 67. It further argues that its failure to submit a complete certification was harmless.
I disagree. Plainly, SKF’s certification was not futile, because the Court of International Trade, reviewing Customs’s rejection of the certification, held that the Byrd Amendment was unconstitutional and, as a result, that Customs should not have rejected SKF’s certification. If SKF believed when it filed its certification that its challenge to the constitutionality of the Byrd Amendment was worth consideration by Customs, the Court of International Trade, and this court, then SKF should have expended its own time and effort to provide a complete and timely certification for all of its expenses. I would affirm Customs’s refusal to accept SKF’s amended certification under the Administrative Procedure Act’s standard of review.
* * *
For the foregoing reasons, I respectfully dissent.
. Similar questions appeared on the version of the questionnaire that SKF completed in 1989. See Final Questionnaire of SKF USA, Inc. at 106-07.
. The equivalent to this question that appeared in the version of the questionnaire that SKF completed in 1989 was question 1.2, which asked “Please indicate, by checking the appropriate box, the position that your firm takes with respect to the petition. (CHECK ONLY ONE)” and offered the choices "Supports the petition,” "Opposes the petition,” and "Does not wish to take a position on the petition.” See Final Questionnaire of SKF USA, Inc. at 6.
. The majority states that I "rely[] primarily on the government's representations at oral argument” to conclude that the purpose of the Byrd Amendment is not the "reward for assistance” rationale that the majority advances. Maj. Op. at 1352. As discussed in detail, the statutory text of the Byrd Amendment, its legislative history, the conduct of antidumping investigations in practice, and the example of this very case all make clear that the purpose of the Byrd Amendment was not to reward companies for assisting the government in antidumping investigations. The fact that the government agrees that “reward for assistance” was not the purpose of the Byrd Amendment is only one of many reasons that I would reject it.
. The majority also relies on a recent case from the First Circuit that, as an alternative ground for its decision, reasoned that the transfer of data that identified which physicians had prescribed specific pharmaceuticals was commercial speech. See Maj. Op. at 1355 (citing IMS Health Inc. v. Ayotte, 550 F.3d 42, 54-55 (1st Cir.2008)). In that case, the First Circuit rejected a "narrower definition of commercial speech limited to activities 'propos[ing] a commercial transaction,’ ”- and instead reasoned that the data transfer at issue "at most embodied] expression related solely to the economic interest of the speaker and its audience.” IMS, 550 F.3d. at 54. The IMS case plainly involved the sale of data — i.e., a commercial transaction that involved payment for the supposed "speech,” which the court reasoned was actually not speech at all, but rather conduct. Id. To the extent that the majority concludes that IMS stands for the broader proposition that any speech that involves the "economic interests of the speaker” is commercial speech, I respectfully submit that either the majority’s reading of IMS is incorrect, or IMS was incorrectly decided.
. The majority questions my seeming failure to explain why an even narrower construction of the Byrd Amendment — "limiting the rewards to petitioners alone” — would not meet the final requirement of the Central Hudson test. See Maj. Op. at 1357 n. 35. Because the majority does not adopt this narrower con-struclion, the significance of the majority’s criticism is not clear. In any event, as discussed in detail below, I disagree that limiting distributions to petitioners alone would cure the First Amendment problem. See infra at 1357-58.
. In a footnote, the majority asserts an alternative basis for affirmance: "[e]ven if we apply the test for speech combined with conduct in United States v. O’Brien, 391 U.S. 367, 377, 88 S.Ct. 1673, 20 L.Ed.2d 672 (1968), for reasons that are clear from the text the Byrd Amendment would still be constitutional.” Maj. Op. at 1355 n. 28.1 respectfully disagree. For the same reasons that it cannot satisfy Central Hudson's "not more extensive than is necessary” requirement, the Byrd Amendment cannot meet O’Briens requirement that any "incidental restriction on alleged First Amendment freedoms [be] no greater than essential to the furtherance of [the government’s asserted] interest.” O’Brien, 391 U.S. at 377, 88 S.Ct. 1673.
. I agree with the majority’s conclusion that, if the Byrd Amendment were subject to rational basis review under the Equal Protection Clause, it would survive — though I do so for different reasons. Though the petition support requirement is not a good proxy for the seriousness of a domestic producer’s injury, I would not conclude, as the Court of International Trade did, that it is an irrational proxy. I would therefore affirm the judgment of the Court of International Trade solely on the alternative basis that the petition support requirement violates the First Amendment.