Allied-Signal Aerospace Company appeals from the judgment of the United States Court of International Trade denying Allied-Signal's motion for judgment on the agency record and affirming the final determination of the United States Department of Commerce, International Trade Administration (ITA) after an administrative review of antidumping duty orders covering antifriction bearings from France. Allied-Signal Aerospace Co. v. United States, 802 F. Supp. 463 (Ct.Int'l Trade 1992). In reviewing the dumping margins for bearings imported by Allied-Signal, the ITA used the highest dumping rates assigned to any company in the original antidumping investigation as the "best information otherwise available" under 19 U.S.C. § 1677e(c) (1988). We reverse and remand.

In response to a petition filed in 1988 on behalf of the domestic antifriction bearing industry, the ITA initiated an investigation to determine whether to impose antidumping duties on antifriction bearings and parts thereof from France. Upon the conclusion of its investigation, the ITA determined that such bearings were being, or were likely to be, sold in the United States at less than fair value (LTFV). 54 Fed.Reg. 19092 (Dep't Comm.1989). Pursuant to that LTFV determination, the ITA issued the following "weighted average" dumping margins covering antifriction bearings from France:

Ball Bearings                Margin Percentage
INA                               66.18
SKF                               66.42
SNR                               56.50
All others                        65.13

Cylindrical Roller Bearings  Margin Percentage
INA                               11.03
SNR                               18.37
All others                        17.31

Id. at 19096. On May 15, 1989, the ITA published final orders imposing antidumping duties on antifriction bearings imported from France consistent with its LTFV determination. Id. at 20902.

On the first anniversary of the publication of the antidumping duty orders, the ITA notified interested parties of an opportunity to request administrative review of the orders pursuant to 19 U.S.C. § 1675 (1988). 55 Fed.Reg. 19093 (May 8, 1990). Allied-Signal, a United States purchaser and importer of antifriction bearings, submitted a request for administrative review of ball bearings and cylindrical roller bearings produced by SNFA S.A., a French manufacturer of customized aerospace bearings. On June 11, 1990, the ITA initiated the review for the period from November 9, 1988 through April 30, 1990. 55 Fed.Reg. 23,575.

The ITA subsequently issued a questionnaire to SNFA requesting factual information pertinent to the review, pursuant to 19 C.F.R. § 353.22(c)(2) (1992). The ITA questionnaire requested general corporate information (Section A), data on U.S. sales (Section B), data on home market sales (Section C), and cost of production information and constructed value data (Section D) for the review period. SNFA, however, provided information responding only to Section A of the questionnaire due to what it alleged were procedural and financial obstacles prohibiting it from providing the remaining requested data. Because SNFA was deemed to have provided an "inadequate response" to the questionnaire, the ITA resorted to the "best information available" (BIA), which it concluded was represented by "the highest margin calculated in the final determination of sales at less than fair value (LTFV) or this review." 56 Fed.Reg. 11179.

The ITA published its preliminary results from the administrative review on March 15, 1991. For the period from November 9, 1988 through April 30, 1990, the ITA issued the following preliminary dumping margins:

Ball Bearings                Margin Percentage
SKF"France                        6.97
Fiat Aviazione                    0.00
SNECMA                            0.08
Aerospatiale                      8.09
Turbomeca                         10.43
Pratt & Whitney Canada            4.89
SNR Roulements                    0.36
INA Roulements                    66.42
SNFA                              66.42
Dowty Rotol                       66.42

Cylindrical Roller Bearings  Margin Percentage
SKF"France                        No sales
Fiat Aviazione                    No sales
SNECMA                            0.29
Aerospatiale                      22.73
Turbomeca                         1.88
Pratt & Whitney Canada            4.89
SNR Roulements                    0.31
SNFA                              22.73
INA Roulements                    22.73
Dowty Rotol                       No Sales

56 Fed.Reg. 31750 (emphases added).

In reaching its final results, the ITA again resorted to the best information available and explained its BIA policy:

For the purposes of these final results of review, we have applied two tiers of BIA:

1. When a company refused to cooperate with the [ITA] or otherwise significantly impeded these proceedings, we have used as BIA the higher of (1) The highest of the rates found for any firm for the same class or kind of merchandise in the same country of origin in the less than fair value investigation (LTFV) or (2) the highest rate found in this review for the same class or kind of merchandise in the same country of origin.

2. When a company substantially cooperated with our requests for information including, in some cases, verification, but failed to provide the information requested in a timely manner or in the form required, we have used as BIA the higher of: (1) The firm's LTFV rate for the subject merchandise (or the "all others" rate from the LTFV investigation, if the firm was not individually investigated), or (2) the highest calculated rate in this review for the class or kind of merchandise from the same country of origin.

56 Fed.Reg. 31705. The ITA stated that because "SNFA responded only to section A of our questionnaire and provided no further responses ... [we therefore] applied the first tier of BIA." Id. Thus, the dumping margin determinations of 66.42% for ball bearings and 18.37% for cylindrical roller bearings manufactured by SNFA were each the highest margins assigned to any company from the original LTFV determination for those types of bearings.

On August 29, 1991, Allied-Signal filed a complaint in the United States Court of International Trade challenging the ITA's final dumping margin determination. Allied-Signal did not dispute that the ITA was required under 19 U.S.C. § 1677e(c) to use the best information available in making its determination. However, Allied-Signal argued that in selecting the highest rates from the original LTFV investigation as the best information available, the ITA used its "two-tier" BIA methodology in a punitive manner, thereby frustrating the remedial purpose of the antidumping laws. Allied-Signal also argued that the ITA abused its discretion in failing to consider the available verified information submitted for the administrative review record by other similarly situated bearing manufacturers, which it claimed more accurately reflected market conditions during the review period.

Allied-Signal moved for judgment on the record, and that motion was denied. The court concluded that the agency was justified in applying the "first tier" of the methodology because SNFA failed to adequately respond to the questionnaire. 802 F. Supp. at 466. The court also concluded that the agency's selection of the highest prior margins was reasonably within its broad discretion in selecting the best information available. Id. Lastly, the court rejected Allied-Signal's contention that the ITA punitively applied its BIA methodology to SNFA's bearing entries because Allied-Signal did not offer evidence to show that the margins selected by the ITA were demonstrably less probative of current market conditions than the review margins. Id. at 467. Accordingly, the court affirmed the final determination in all respects and dismissed the case.

We review the decision of the Court of International Trade by applying de novo the statutory standard of review to the underlying agency determination. U.H.F.C. Co. v. United States, 916 F.2d 689, 696 (Fed.Cir.1990). We must determine whether the trial court correctly concluded that the ITA's determination was not "unsupported by substantial evidence on the record, or otherwise not in accordance with law." 19 U.S.C. § 1516a(b)(1)(B) (1988).

Allied-Signal alleges that for the period of review, the ITA improperly assigned antidumping duty margins to SNFA's bearings based on what Allied-Signal characterized as a "punitive" two-tier BIA methodology. The punitive nature of the ITA's duty assessment, according to Allied-Signal, is evidenced by the ITA's refusal to consider alternative simplified reporting requirements for SNFA and its rejection of lower margins based upon verified information allegedly more probative of current market conditions. Allied-Signal claims that the ITA's final determination was not in accordance with law because the ITA's punitive use of its BIA policy frustrated the remedial purpose of the antidumping laws.

The ITA's selection of the best information available in a given case is necessarily predicated on the agency's interpretation of the governing statutory and regulatory provisions. Thus, in examining the propriety of that selection, we must consider the scope of the statutory authority conferred upon the ITA regarding the use of the best information available, the ITA's implementation of that authority through its promulgated regulations, and the ITA's execution of its regulations through its articulated "BIA policy."

Section 776 of the Tariff Act of 1930 contains two important rules governing the use of "best information available" in antidumping determinations. First, the statute mandates that the ITA "shall verify all information relied upon" in reaching a final determination in an antidumping investigation or an administrative review. 19 U.S.C. § 1677e(b) (1988). If the ITA is unable to verify the accuracy of the information it receives, then "it shall use the best information available to it as the basis for its action." Id. Second, "whenever a party or any other person refuses or is unable to produce information requested in a timely manner and in the form required, or otherwise significantly impedes an investigation," both the ITA and the U.S. International Trade Commission (ITC) are required to use "the best information otherwise available." 19 U.S.C. § 1677e(c) (1988). These "BIA rules" were among the "administrative reforms" enacted by Congress in the Trade Agreements Act of 1979, Pub.L. No. 96-39, in an effort to reduce the length of antidumping and countervailing duty investigations, which was often attributable to the delay in the assessment and collection of data. See H.R.Rep. No. 317, 96th Cong., 1st Sess. 24 (1979).

The implementing regulation promulgated by the ITA tracks the statutory provision covering the use of the best information available. See 19 C.F.R. § 353.37(a) (1992). In addition, the regulation provides that where an interested party "refuses to provide factual information requested by the [ITA] or otherwise impedes the proceeding, the [ITA] may take that into account in determining what is the best information available." 19 C.F.R. § 353.37(b) (1992). Thus, in addition to evaluating the adequacy of the response, the regulation allows the ITA to consider the degree of a particular respondent's cooperation in the administrative review as a factor in determining what constitutes the best information available. Cf. Atlantic Sugar, Ltd. v. United States, 744 F.2d 1556, 1560, 2 Fed.Cir. (T) 130, 134 (1984) ("[O]ne may view the best information rule ... as an investigative tool, which that agency may wield as an informal club over recalcitrant parties whose failure to cooperate may work against their best interest.").

In giving effect to its regulation, the ITA had earlier consistently subscribed to a "BIA policy" in which it assigned to a nonresponsive firm the higher of the highest rate calculated in the current administrative review or the firm's own prior rate. That policy was recently articulated by the ITA as follows:

When a company refuses to provide the information requested in a timely manner, or otherwise significantly impedes the [ITA's] review, the [ITA] assigns to that company the highest margin for the subject merchandise of (1) The highest margin calculated for that company in any previous review; (2) the highest margin calculated for any respondent that supplied adequate responses in this review; or, (3) the margin for that company calculated in the less than fair value (LTFV) investigation.

Cyanuric Acid From Japan: Final Results of Administrative Review, 56 Fed.Reg. 2741, 2742 (Dep't Comm. Jan. 24, 1991) (emphases added). We have previously accepted this approach as a reasonable and permissible interpretation of the BIA statute because it encourages the timely submission of pertinent data and "implements the basic purpose of the statute--determining current margins as accurately as possible." Rhone Poulenc, Inc. v. United States, 899 F.2d 1185, 1191 (Fed.Cir.1990).

The ITA departed from its prior BIA policy by employing, for the first time in its administrative review of antidumping duty orders covering antifriction bearings, a "two-tier BIA methodology." The methodology has been described as a "new policy for dealing with firms that refuse to cooperate with the [ITA] or otherwise significantly impede the proceedings." Final Results of Antidumping Duty Administrative Review: Clear Sheet Glass From Taiwan, 56 Fed.Reg. 44075 (Dep't Comm. Sept. 6, 1991).

Under the new BIA scheme, when a respondent refuses to cooperate with the ITA's requests for information or significantly impedes the administrative review, the ITA selects the higher of the highest rate assigned for any firm in the LTFV investigation or the highest rate calculated in the administrative review. This "first tier" treatment results in the selection of the most adverse margin possible as the best information available. In contrast, respondents who substantially cooperate but nonetheless fail to provide the requested information in a timely manner or in the required form are subject to a "second tier," which is substantively similar to that of the ITA's prior BIA policy. When the second tier is employed, the ITA assigns to a respondent the higher of its own prior LTFV rate or the highest rate calculated in the current administrative review.

The critical difference in BIA treatment under the first and second tiers lies in the range of LTFV margins subject to consideration as the best information available. Under the first tier, the ITA may consider the margin of any company involved in the prior LTFV investigation. Under the second tier, the ITA may only consider a particular firm's own prior LTFV rate. The difference has significant consequences when the BIA scheme is actually employed. Although each tier requires the ITA to choose the higher of a certain LTFV rate and the highest review rate, the ITA has acknowledged that in practice, "generally the highest rate for any company from the less than fair value (LTFV) investigation" is used for the best information available under the first tier and that under the second tier, "generally the highest rate found for any company in these reviews" is used. 56 Fed.Reg. 11179.

Allied-Signal asserts that the two-tier methodology adopted by the ITA in determining the best information available is not in accordance with law because it focuses on a respondent's level of cooperation, as perceived by the ITA, and is used by the agency to punish noncompliance. In particular, Allied-Signal argues that the first tier precludes the ITA from making an accurate determination of current margins because it forces the agency to "blindly" select the most adverse margins as the best information available and to ignore the most recent information submitted during the administrative review. We disagree.

Although Congress expressly mandated that the ITA use the best information available when faced with a party who is unwilling or unable to participate in the administrative review proceedings, it did not explicitly define what type of information constituted the "best" information.1 Hence, because Congress has "explicitly left a gap for the agency to fill" in determining what constitutes the best information available, the ITA's construction of the statute must be accorded considerable deference. Chevron U.S.A., Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 843-44, 104 S. Ct. 2778, 2781-82, 81 L. Ed. 2d 694 (1984).

We reject Allied-Signal's charge that rates derived from the new methodology are necessarily punitive rather than based on the best information. As we stated in Rhone Poulenc, Inc. v. United States, 899 F.2d 1185 (Fed.Cir.1990), "[i]n order for the agency's application of the best information rule to be properly characterized as 'punitive,' the agency would have had to reject low margin information in favor of high margin information that was demonstrably less probative of current conditions." Id. at 1190. The new BIA scheme, like its predecessor, merely establishes a presumption that the highest prior margins are the best information available. That presumption can be rebutted by the respondent with evidence showing the actual margin to be less. See id.

Additionally, the ITA's new policy is consistent with the statutory purpose underlying the BIA rules provided in 19 U.S.C. § 1677e, which is to facilitate the determination of dumping margins as accurately as possible within the confines of extremely short statutory deadlines. Under such circumstances, "cooperation by the parties to the [administrative review] is essential ... to gather the data needed for an accurate determination." Atlantic Sugar, 744 F.2d at 1560. We have recognized that "the ITA cannot be left merely to the largesse of the parties at their discretion to supply the ITA with information. This is particularly the case when the ITA is attempting to obtain information to conduct statutorily mandated administrative reviews because unlike ITC, the ITA has no subpoena power." Olympic Adhesives, Inc. v. United States, 899 F.2d 1565, 1571 (Fed.Cir.1990).

By taking into account a respondent's degree of cooperation, the two-tier methodology induces the submission of timely, accurate, and complete information by respondents, who are often in the best position to obtain and compile such information. Additionally, by allowing the ITA to make the "common sense inference" that the highest prior margins are the most indicative of current market conditions, Rhone Poulenc, 899 F.2d at 1190, the ITA avoids rewarding the uncooperative and recalcitrant party for its failure to supply requested information. "An agency's own construction of its own regulations has been regarded as especially due ... respect." Ford Motor Credit Co. v. Milhollin, 444 U.S. 555, 565-67, 100 S. Ct. 790, 796-97, 63 L. Ed. 2d 22 (1980).

Accordingly, we hold that the new two-tier methodology adopted by the ITA and used in its administrative review of antidumping duty orders is a reasonable and permissible exercise of the ITA's statutory authority to use the best information available when a respondent refuses or is unable to provide requested information.

Having addressed the validity of the ITA's new BIA policy, we now turn to its application to SNFA's bearings. Allied-Signal argues that the ITA erred in its decision to apply the first tier of the BIA methodology. It contends that it did not refuse to cooperate with the ITA and that the circumstances underlying its nonresponsiveness do not justify the application of the first tier. Based on our review of the record, we agree that the ITA erred in its application of the two-tier BIA methodology.

Whether the first or second tier properly applies to a nonresponsive respondent essentially turns on the level of cooperation exhibited by that respondent during the review. In order to apply the first tier to a particular respondent, the ITA must conclude that the respondent "refused to cooperate with the [ITA] or otherwise significantly impeded" the review. However, if the respondent "substantially cooperated ... but failed to provide the information in a timely manner or in the form required," the second tier is applicable.

In the instant case the record is clear that SNFA did not refuse to cooperate. Prior to the initiation of the administrative review, officials from SNFA and the ITA discussed proposals for a simplified process of review for those companies like SNFA that were in the "all others" category of the LTFV investigation. SNFA concluded that it could not adequately respond to the questionnaire because the particular costing system it had in place and its limited accounting resources made compilation of constructed value data difficult. Additionally, its small-volume production of specialized, custom-made bearings allegedly led to an absence of home market sales data.

SNFA alerted the ITA to its difficulty in responding to the questionnaire and indicated its willingness to participate in a simplified review process. Although the ITA acknowledged as much in its correspondence with SNFA and expressed interest in seeking to accommodate SNFA, it ultimately concluded that SNFA had refused to cooperate. The ITA explained that it employed the first tier because

SNFA did not provide a substantive response to our questionnaire. It responded to section A (general information) of the questionnaire, but did not respond to sections B, C, or D. This does not constitute a reasonable attempt by SNFA to comply with the [ITA's] requirements for conducting the review. The lack of information significantly impeded the [ITA's] investigation, since there was no basis for calculating actual dumping margins nor even data to consider.

56 Fed.Reg. 31706. However, in view of the undisputed fact that SNFA supplied as much of the requested information as it could and offered to provide the remaining information in a simplified form, we must conclude that it was unreasonable for the ITA to have characterized SNFA's behavior as a refusal to cooperate.

Moreover, application of the first tier in SNFA's case does not promote the underlying objectives of the BIA rules. One of the fundamental purposes of the rules is to induce the timely submission of information to assist the ITA in determining accurate dumping margins. Thus, for example, the ITA may well have been justified in resorting to the first tier in selecting the best information available for bearings produced by INA Roulements S.A. after it announced that it would not participate in the administrative review and accordingly did not respond to the questionnaire at all. 56 Fed.Reg. 11179. In contrast, SNFA stated its willingness to cooperate and responded with the requested information to the extent it could.

Neither is the goal of encouraging future compliance furthered by the application of the first tier to SNFA, because it apparently has no ability to respond more completely than it already has done. SNFA failed to provide a complete response to the requested information because it was unable to, not because it refused to. SNFA will likely continue to be unable to adequately provide the requested data in future reviews unless a simplified process is established. Upon recognizing its institutional and financial limitations, SNFA offered an alternative proposal involving the determination of new "all other" rates, based on information submitted during the review, for those small companies that found participation in a full review difficult and financially burdensome.

We hold that the ITA erred in employing the first tier of its two-tier methodology and that SNFA's bearings at issue are subject to second tier treatment. A dumping margin determination derived from an erroneous application of the ITA's own two-tier BIA methodology cannot be based on the "best information otherwise available" pursuant to section 1677e(c). Furthermore, by having ignored alternative and simplified reporting methods to generate relevant information, the ITA in the instant case may not have used the best information in assessing SNFA's dumping margins, as it is statutorily mandated to do. Accordingly, we reverse the judgment of the Court of International Trade affirming the ITA's final determination and remand the case to the court to direct the ITA to recalculate the dumping margins at issue under the second tier of the two-tier BIA methodology.

The two-tier BIA methodology employed by the ITA in selecting the best information available for nonresponsive parties is a permissible and reasonable exercise of its statutory authority. However, the ITA's final determination covering SNFA's entries of antifriction bearings for the period under review was based on a misapplication of that methodology and thus was not in accordance with law. Therefore, we reverse the judgment of the Court of International Trade and remand for further proceedings consistent with this opinion.

No costs.

REVERSED and REMANDED.