F.LLI De CECCO DI FILIPPO FARA S. MARTINO S.p.A., Plaintiff-Appellee, v. UNITED STATES, Defendant-Appellant, and Borden, Inc., Gooch Foods, Inc., and Hershey Foods Corp. (now New World Pasta, LLC), Defendants, 216 F.3d 1027

Opinion for the court filed by Circuit Judge MICHEL. Opinion concurring-in-part, dissenting-in part filed by Circuit Judge SCHALL. MICHEL, Circuit Judge. The United States appeals from the decision of the United States Court of International Trade in this anti-dumping duty case, Borden, Inc., Gooch Foods, Hershey Foods Corp. v. United States, No. 96—08—01970 (Ct. Int’l Trade Dec. 16, 1998) (“Borden II”), with respect to just one of the numerous respondents, F.lli De Ceceo di Filippo Fara S. Martino S.p.A. (“De Cec-eo”).

The Court of International Trade affirmed the portion of the Remand Rede-termination of the Department of Commerce (“Commerce”) that imposed, consistent with the Court of International Trade’s March 26, 1998 remand order in Borden, Inc., Gooch Foods, Hershey Foods Corp. v. United States, 4 F.Supp.2d 1221 (Ct. Int’l Trade 1998) (“Borden I ”), a revised dumping margin of 24.31 percent on De Ceceo as an adverse inferential rate against a non-cooperating party.

In its Borden II decision, the Court of International Trade refused to consider Commerce’s presentation of a new methodology, based on data gathered in Commerce’s original investigation of cooperating companies, to corroborate the 46.67 percent rate that the court had previously rejected as discredited and uncorroborated in view of Commerce’s own investigation. See Borden II, slip op. at 2. The government timely appealed to this court. The appeal was submitted for our decision following oral argument on April 4, 2000.

Because we hold that the trial court did not err in (1)rejecting the 46.67 percent anti-dumping margin as discredited and uncorroborated; (2) suggesting a possible alternative rate based on the verified dumping margins of the cooperating respondents; or, (3) refusing to allow Commerce a second opportunity to attempt to corroborate the 46.67 percent rate with a new methodology, we affirm. BACKGROUND This case involves the determination of dumping margins for imports from Italian pasta manufacturers.

Unlike the other manufacturers, De Ceceo did not fully cooperate with Commerce during the inquiry period, leading Commerce to make an “adverse inference” when determining the dumping margin of the pasta products at issue. The relevant statute provides that if Commerce finds that an interested party has failed to cooperate by not acting to the best of its ability to comply with a request for information from [Commerce], [Commerce] in reaching the applicable determination under this subtitle, may use an inference that is adverse to the interests of that party in selecting from among the facts otherwise available.

Such adverse inference may include reliance on information derived from (1) the petition, (2) a final determination in the investigation under this subtitle, (3) any previous review under

section 1675 of this title ... or, (4) any other information placed on the record. 19 U.S.C. § 1677e(b) (1994) (emphasis added). The statute further provides: (

c) Corroboration of secondary information When [Commerce] relies on secondary information rather than on information obtained in the course of an investigation or review, [Commerce] shall, to the extent practicable, corroborate that information from independent sources that are reasonably at their disposal. 19 U.S.C. § 1677e(c) (1994) (emphasis added). In its initial anti-dumping determination, Commerce concluded that De Ceceo had not cooperated.

See Notice of Preliminary Determination of Sales at Less Than Fair Value and Postponement of Final Determination: Certain Pasta From Italy, 61 Fed.Reg. 1344, 1345 (Jan. 19, 1996) (“Preliminary Determination”). Because De Ceceo failed to cooperate, Commerce used an “adverse facts otherwise available” dumping margin for De Ceceo. See id. Commerce chose a simple average of the range of margins stated in the notice of initiation: 46.67 percent. See id.

Because the notice relied on the allegations in the petition from the domestic manufacturers, we refer to it throughout this opinion as the “petition rate.” Commerce later permitted De Ceceo an opportunity to answer supplemental questions to verify that information that De Ceceo did report. De Ceceo, however, was unable to rectify the deficiencies in its submissions to Commerce, and Commerce imposed a 46.67 percent margin in its Final Determination. See Notice of Final Determination of Sales at Less Than Fair Value: Certain Pasta From Italy, 61 Fed.Reg. 30,326, 30,-329 (June 17, 1996) (“Final Determination”).

De Ceceo filed suit in the Court of International Trade challenging both Commerce’s finding that it had not cooperated and Commerce’s use of the petition rate as the adverse facts otherwise available rate. De Ceceo argued that Commerce was statutorily required to corroborate the petition rate by using other information readily available. See Borden I, 4 F.Supp.2d at 1247. Following a motion for judgment on the agency record pursuant to Ct.

Int’l Trade R. 56.2, the Court of International Trade held that Commerce had not sufficiently established that De Ceceo was uncooperative and remanded to Commerce for reconsideration and a redetermination as to the dumping margin. In addition, the trial court held that Commerce could not rely on the petition rate for the adverse inference because Commerce’s own investigation revealed that the 46.67 percent rate was “thoroughly discredited” and uncorroborated. Id. at 1247 & 1248.

The court, however, suggested that Commerce might “use the highest verified margin of [24.31 percent]” applicable to those respondents that did cooperate with Commerce’s investigation. Id. at 1247. In its Remand Redetermination, Commerce found that De Ceceo was indeed uncooperative and applied the 24.31 percent anti-dumping margin to De Ceceo. See Redetermination on Remand: Final Determination of Sales at Less than Fair Value: CeHain Pasta from Italy at 5-6 (Aug. 28, 1998) (“Remand Redetermination”).

In addition, in an effort to corroborate the petition rate Commerce applied a new methodology to analyze the data gathered in its investigation of the cooperating respondents. The new methodology compared the petition rate to the margins calculated on specific transactions of record for the cooperating respondents.

While the transactions of record were not reported as such in either the Final Determination or the Remand Redetermination, Commerce reported in the Remand Rede-termination that the petition rate of 46.67 percent was within the range of the margins calculated on transactions of record for cooperative respondents. Thus Commerce argued that the Court of International Trade should have considered the petition rate corroborated. See id. at 15.

When the case returned to the Court of International Trade, the government requested that the court reconsider its original order and permit Commerce to corroborate the 46.67 percent petition rate using the methodology first applied by Commerce on remand. See Borden II, slip op. at 2-3. The Court of International Trade, however, affirmed the imposition of a 24.31 percent rate on De Ceceo and rejected Commerce’s request for reconsideration, declaring: It is too late for Commerce to attempt to justify its original 46.67% margin with new support or methodology.

The issue was before the court prior to remand, and all arguments relating thereto should have been made at that time.... Commerce’s statement that, pursuant to a new methodology, it established the 46.67% margin was “within the range of margins calculated on transactions for cooperative respondents” ... is not a sufficient reason to reopen this issue. Id. The Court of International Trade affirmed the finding that De Ceceo was not cooperative and affirmed that portion of Commerce’s Remand Redetermination that applied a 24.31 percent rate to De Ceceo. See id. at 4.

Commerce appealed to this court, which has jurisdiction pursuant to 28 U.S.C. § 1295(a)(5) (1994). DISCUSSION The government argues that the Court of International Trade erred in not remanding the case to Commerce under instructions that would allow it to corroborate the petition rate with further evidence or to select any other proper rate, rather than rejecting the petition rate and suggesting a rate that Commerce might impose on De Ceceo.

In addition, the government argues that the trial court compounded its original mistake by refusing to reconsider the petition rate in light of the corroborating evidence cited in the Remand Redeter-mination. The government concedes, however, that the trial court was correct in holding that the 46.67 percent petition rate was not properly corroborated in the Final Determination. When reviewing anti-dumping determinations made by Commerce, this court applies anew the standard of review applied by the Court of International Trade in its review of the administrative record. See Atlantic Sugar, Ltd. v.

United States, 744 F.2d 1556, 1559 n. 10 (Fed.Cir.1984) (“We review the [Court of International Trade’s] review of an ITC determination by applying anew the statute s ... express judicial review standard.”). In doing so, we uphold Commerce’s determination unless it is “unsupported by substantial evidence on the record, or otherwise not in accordance with law.” 19 U.S.C. § 1516a(b)(l)(B)(i) (1994). This court reviews de novo the trial court’s answers to all questions of law, including statutory

interpretation questions; eviden-tiary decisions, including waiver for failure to timely present evidence or raise an issue, are reviewed for abuse of discretion. Findings of fact by the trial court are reviewed for clear error. See, e.g., Texport Oil Co. v. United States, 185 F.3d 1291,1294 (Fed.Cir.1999).

The principal issue before this court, then, is whether the Court of International Trade erred as a matter of law in rejecting the petition rate of 46.67 percent and in suggesting to Commerce that it may apply to De Ceceo the 24.31 percent margin, which was the highest rate verified for any of the cooperating respondents. Likewise, we must decide whether the trial court abused its discretion in refusing to consider anew the petition rate in light of the allegedly corroborating methodology and evidence offered by Commerce to support the higher, 46.67 percent, anti-dumping margin originally imposed on De Ceceo. I.

The government argues that in Borden I the Court of International Trade erred in prohibiting the use of the petition rate and should instead have remanded the action to Commerce either for (1) proper corroboration of the original petition rate, or (2) selection by Commerce of any other rate that was consistent with the statute. The government argues that the trial court’s order was in conflict with the statute because it “impermissibly limited Commerce’s discretion in selecting the adverse facts available rate” for De Ceceo. We disagree.

As the government argues, the anti-dumping statute leaves much to agency discretion in making anti-dumping determinations. See Smith-Corona v. United States, 713 F.2d 1568, 1571 (Fed.Cir.1983) (“The Secretary has broad discretion in executing the [anti-dumping] law.”). In the case of uncooperative respondents, the discretion granted by the statute appears to be particularly great, allowing Commerce to select among an enumeration of secondary sources as a basis for its adverse factual inferences. See 19 U.S.C. § 1677e(b).

Similarly, this court has repeatedly held that Commerce’s special expertise makes it the “master” of the anti-dumping law, entitling its decisions to great deference from the courts. See, e.g., Micron Tech., Inc. v. United States, 117 F.3d 1386, 1394 (Fed.Cir.1997); Torrington Co. v. United States, 68 F.3d 1347, 1351 (Fed.Cir.1995). Thus, factual determinations supporting anti-dumping margins are best left to the agency’s expertise. See, e.g., Smith-Corona, 713 F.2d at 1571.

Here, it is clear, moreover, that the statute has no requirement that Commerce is limited to the highest rate imposed on a cooperating company when selecting a rate for a non-cooperating respondent. Furthermore, the statute explicitly allows for use of “the petition” to determine relevant facts when a respondent does not cooperate. See 19 U.S.C. § 1677e(b). Thus, we are convinced that it is within Commerce’s discretion to choose which sources and facts it will rely on to support an adverse inference when a respondent has been shown to be uncooperative.

Particularly in the case of an uncooperative respondent, Commerce is in the best position, based on its expert knowledge of the market and the individual respondent, to select adverse facts that will create the proper deterrent to noncooperation with its investigations and assure a reasonable margin. Commerce’s discretion in these matters, however, is not unbounded. As the Court of International Trade pointed out in its decision following Commerce’s Remand Redetermination, the purpose of

section 1677e(

b) is to provide respondents with an incentive to cooperate, not to impose punitive, aberrational, or uncorroborated margins. See Bordon II, slip op. at 4. It is clear from Congress’s imposition of the corroboration requirement in 19 U.S.C. § 1677e(

c) that it intended for an adverse facts available rate to be a reasonably accurate estimate of the respondent’s actual rate, albeit with some built-in increase intended as a deterrent to non-compliance. Congress could not have intended for Commerce’s discretion to include the ability to select unreasonably high rates with no relationship to the respondent’s actual dumping margin. Obviously a higher adverse margin creates a stronger deterrent, but Congress tempered deterrent value with the corroboration requirement.

It could only have done so to prevent the petition rate (or other adverse inference rate), when unreasonable, from prevailing and to block any temptation by Commerce to overreach reality in seeking to maximize deterrence. Here Commerce seeks to impose upon De Ceceo an extremely high dumping margin that it now concedes was uncorroborated in its original determination.

In addition, the trial court found as a fact that the petition rate of 46.67 percent had been “thoroughly discredited,” noting that other high-end producers, similar to De Ceceo whose U.S. prices were among the highest, were found to have lower anti-dumping margins than low-end producers, and even low-end producers had rates significantly lower than the petition rate. Borden I, 4 F.Supp.2d at 1247. Thus, all available evidence indicates that 46.67 percent was many times higher than De Ceceo’s actual dumping margin.

Accordingly, the court found that there was “nothing to support” an inference that the petition rate might reflect De Cecco’s actual dumping margin. Id. The Court of International Trade’s findings were not clearly erroneous and its conclusions were correct. Thus, we hold that the Court of International Trade did not err in rejecting the petition rate and remanding to Commerce for a new determination. In fact, the government appears to concede this point. The government argues, however, that the court erred in holding that the petition rate “may not be utilized” on remand to establish De Cecco’s dumping margin.

Under the circumstances of this case, we do not agree. The highest rate applied to any cooperating respondent was 24.31 percent. In addition, De Ceceo was considered a high-end producer and other high-end producers received margins far below 24.31 percent. Furthermore, the court properly found the petition rate of 46.67 percent was uncorroborated by Commerce. Thus, sufficient evidence in the administrative record was not shown to support the application of the 46.67 percent petition rate to De Ceceo, and we discern ample evidence to support the court’s decision that the petition rate was discredited.

To allow Commerce to impose an unjustifiably high, punitive rate that has been discredited by the agency’s own investigation would permit the agency to exceed the bounds of its admittedly broad discretion.

Thus, given that the petition rate was properly found to be discredited by Commerce’s own investigation, we do not view the Court of International Trade’s decision that the petition rate “may not be utilized” to establish a dumping margin for De Ceceo to be contrary to the statute, based on less than substantial evidence in the administrative record, or based on a clearly erroneous finding of fact by the trial court. II.

With regard to the rate of 24.31 percent, the government argues that “by dictating the exact margin rate to be applied to an uncooperative respondent in this case, the trial court completely eliminated Commerce’s discretion in selecting the adverse facts available rate.” In support of its argument, Commerce quotes Florida Power & Light v. Lorion, 470 U.S. 729, 744, 105 S.Ct. 1598, 84 L.Ed.2d 643 (1985), which holds that if the record before the agency does not support the agency action ... the proper course, except in rare circumstances, is to remand to the agency for additional investigation or explanation....

The reviewing court is generally not empowered to conduct a de novo inquiry into the matter being reviewed and to reach its own conclusions based on such an inquiry. Id. We do not disagree with the government’s argument that Commerce, not the courts, must make determinations as to proper anti-dumping margins.

Here, however, we do not see that the Court of International Trade overstepped its statutory review authority or violated the rule articulated in Florida Power & Light The court’s suggestion to Commerce that it “may use the highest verified margin of [24.31 percent],” we hold, is not an impermissible limitation on Commerce’s discretion. Indeed, in our view, the court’s remand order imposes no limitation on Commerce at all.

Having found that the petition rate was discredited by Commerce’s own investigation and not corroborated by any “independent source,” the Court of International Trade declared merely that if Commerce properly draws an adverse inference, it may use the highest verified margin of [24.31], as, under these facts, even that margin is very likely to be adverse to De Ceceo. Borden I, 4 F.Supp.2d at 1247 (emphasis added). Later in its opinion the Court of International Trade went on to direct that: Commerce shall make findings as to whether De Ceceo acted to the best of its ability.

Following that determination, it shall draw appropriate inferences and select a margin based on facts available, either adverse or not adverse, as required by the inference drawn. Id. at 1248. Rather than limiting Commerce’s discretion, the Court of International Trade merely informed Commerce that it might satisfy its burden of providing a rate that was corroborated and not discredited by using a margin that had been verified. Notwithstanding Commerce’s assertions to the contrary, nothing in the court’s order required Commerce to select the 24.31 percent rate.

In fact, the court specifically directed Commerce to “select a margin based on facts available,” id., and to “select a new dumping margin for De Ceceo,” id. Even if the Court of International Trade had done as Commerce suggests and dictated that the maximum rate to be applied to De Ceceo was 24.31 percent, the court did not abuse its discretion. The facts of this case make it highly likely that the real dumping margin for De Ceceo would be well under 24.31 percent.

De Ceceo is considered a high-end producer and the other high-end producers, Deverde and De Matteis, received dumping margins of only 2.80 percent and 0.67 percent (de minimis) respectively. See Notice of Anti-dumping Duty Order and Final Determination of Sales at Less than Fair Value: Certain Pasta From Italy, 61 Fed.Reg. 38,547, 38,548 (July 24, 1996). The low-end producers, on the other hand, had higher dumping margins; those, however, averaged 16.71 percent. See id. The highest of the low-end rates was 24.31 percent; the “all-others” rate was 12.09 percent. Id.

Thus, applying the highest low-end producer rate to high-end producer De Ceceo essentially is, in itself, an adverse inference. Furthermore, when all facts point to a rate well below 24.31 percent, to allow imposition of a higher rate renders the corroboration requirement of

section 1677e(

c) meaningless. By requiring corroboration of adverse inference rates, Congress clearly intended that such rates should be reasonable and have some basis in reality. The facts of this case are such that the Court of International Trade judge could correctly conclude that any rate higher than 24.31 percent would be unreasonable, without any grounding in reality. In addition, 24.31 percent was more than sufficient to deter non-cooperation because it was almost certainly an adverse rate.

Thus, in this case, with facts that so clearly indicated that a rate over 24.31 percent was unreasonable and unrealistic, imposition of 24.31 percent as a maximum rate would not have been an abuse of the court’s discretion. III. Finally, the government argues that the Court of International Trade erred in not allowing Commerce a second chance, after remand, to corroborate the 46.67 percent petition rate. We, however, find no abuse of discretion in the court’s refusal to reconsider the 46.67 percent rate in light of a new methodology based on data that was gathered prior to Commerce’s Final Determination.

As previously discussed, Commerce’s own investigation revealed the 46.67 percent petition rate to be unreliable. Indeed, on appeal to this court, the government concedes the rate was not corroborated in the Final Determination. Having once found, and correctly so, that the petition rate was “thoroughly discredited,” the Court of International Trade was not required, by law or policy, to entertain another attempt by Commerce to corroborate the rate, especially when, as here, the information relied on in the new methodology was known to Commerce at the time of the Final Determination.

Commerce thus had sufficient opportunity to corroborate the rate, by presenting its alternative methodology to the data it gathered in the verification stage, prior to remand and failed to do so. When, as here, the petition rate is uncorroborated, and Commerce’s own investigation undermines the use of the petition rate, we will not require the Court of International Trade to give Commerce a second opportunity for corroboration based on pre-existing data.

To do so would be a waste of the court’s time and would create a perverse incentive for Commerce not to make adequate efforts to corroborate the rate it selected in the first instance, as the statute requires. This we decline to do. CONCLUSION In its remand order, the Court of International Trade properly barred Commerce’s use of the petition rate to establish De Cecco’s anti-dumping margin and made no error in suggesting to Commerce, as one option, a verified rate that it could apply instead of the discredited petition rate.

The Court of International Trade did not limit Commerce’s broad discretion in making adverse inferences regarding uncooperative respondents in an anti-dumping investigation merely by suggesting one rate that would withstand scrutiny. Because, finally, the petition rate was discredited by Commerce’s own investigation, we find no error in the- court’s refusal to consider Commerce’s belated attempt to corroborate that rate with pre-existing data. We therefore AFFIRM. .

In its March 26, 1998 decision, the trial court directed that Commerce might use a rate of 21.34 percent for De Ceceo; it later revised that rate to 24.31 percent in its final judgment of December 26, 1998, because 24.31 was the highest rate ultimately applied to a cooperating respondent. Borden II, slip op. at 4. . While we do not interpret the Court of International Trade decision as imposing a maximum, it is clear that Commerce interpreted it as such. In addition, De Cecco’s briefs and oral arguments indicated that it interpreted the court's order as dictating a maximum rate. .

While the Court of International Trade opinion could perhaps have been clearer in articulating the reasoning behind both its bar on the use of the 46.67 percent rate and on its rationale for the use of the 24.31 percent rate, we will not reverse for imperfect opinion drafting. Judicial decisionmaking is not an opinion writing contest. .Whether the specific methodology applied by Commerce to corroborate the petition rate was, or could ever be, sufficient to verify an adverse factual inference applied to a non-cooperating respondent is not before this panel.

Accordingly, we do not reach the issue and intimate no view as to whether such a methodology could ever satisfy the statutory requirement of corroboration.