Asociacion Colombiana De Exportadores De Flores v. United States , 13 Ct. Int'l Trade 526 ( 1989 )

717 F. Supp. 834 (1989)

The ASSOCIATION COLOMBIANA de EXPORTADORES de FLORES, et al., Plaintiffs,
v.
The UNITED STATES, et al., Defendants,
and
Floral Trade Council of Davis, California, Defendant-Intervenor.

Court No. 87-04-00622.

United States Court of International Trade.

June 29, 1989.

Arnold & Porter, Patrick F.J. Macrory, Spencer S. Griffith and Gwyn F. Murray, Washington, D.C., for Association Colombiana de Exportadores de Flores, et. al.

Heron, Burchette, Ruckert & Rothwell Thomas A. Rothwell, Jr. and James M. Lyons, Washington, D.C., for Floramerica.

Stuart E. Schiffer, Acting Asst. Atty. Gen., David M. Cohen, Director, Commercial Litigation Branch (Jeanne E. Davidson) Civ. Div., U.S. Dept. of Justice; Anne W. White, Office of the Chief Counsel for Import Admin., U.S. Dept. of Commerce, Washington, D.C., for defendant.

Stewart & Stewart, Eugene L. Stewart, Terence P. Stewart, James R. Cannon, Jr., and Jimmie V. Reyna, Washington, D.C., for Floral Trade Council of Davis, Cal.

OPINION

BACKGROUND

RESTANI, Judge:

Representatives of both the Colombian and the domestic fresh cut flower industry challenge the results of the United States International Trade Administration (ITA) remand determination, wherein ITA amended its previous assessment of dumping rates for some respondents, while leaving other respondents' rates the same. The court's original opinion in this matter ordering remand to ITA is Asociacion Colombiana de Exportadores de Flores v. United States, 13 CIT ___, 704 F. Supp. 1114 (1989) hereinafter referred to as Asocolflores I. That opinion contains the relevant factual information applicable to this case, and almost all of the relevant legal discussion, and should be read in conjunction with this decision. Asocolflores is a Colombian producer and exporter organization, and the lead plaintiff herein. For ease of argument, plaintiffs generally will be referred to as "Asocolflores," "plaintiffs," or "respondents."^[1] Where appropriate the court will refer to individual producers or exporters. Floral Trade Council, also a plaintiff, but not referred to as such in these consolidated actions, represents domestic flower interests and is referred to as "FTC." Defendant is the United States of America.

A. Asocolflores' Challenge to the Rate for Non-Responding Companies

In its original determination, for companies which were included in its sample for investigation but which did not respond properly to ITA's requests for information, ITA utilized the highest rate of dumping then reflected in the record. This rate was the highest rate for any company whose data was successfully verified. The company with the highest such rate was Uniflor, an exporter. Upon remand, because of a change in methodology approved by the court, see discussion Section B infra, Uniflor's rate dropped forty percentage points to approximately forty-six percent.^[2]See Remand Determination at 10-11. This remains the highest rate derived from verified data.

On remand, for these non-responding companies ITA decided to use petitioner's rate as the highest rate of record. Id. at 11. Pursuant to statute, 19 U.S.C.A. § 1677e (West Supp.1989), ITA may use "best information available" (BIA) to assign a dumping rate to such companies if they do not provide ITA with the information it requests. The statute also states that information "submitted in support of the petition" may be chosen by ITA as best information available for purposes of a final determination. 19 U.S.C.A. § 1677e(b).

Asocolflores objects to ITA's use of petitioner's rate as BIA on the basis that a verified rate should be used because such a rate by its very nature (i.e. that it is derived from information confirmed by ITA) is "better" information than petitioner's rate, which is based only on data provided by petitioners. See Plaintiffs' brief at 3-4. Defendant for its part claims that this is a discretionary matter, and that its newly articulated policy is to use the highest rate available for non-cooperative parties, which in this case is the rate alleged in the petition. ITA further avers that under this policy, parties which are cooperative but yet cannot substantially comply with ITA's requests, likely would receive rates lower than the petition rate. As this policy is newly articulated there are no court precedents directly on point.^[3]

As far as the non-responding parties are concerned they are in a particularly difficult position. Presumably, petitioner's rate was known and any interested respondent could have complied with ITA's request if it had data to establish a lower rate. None of these non-complying respondents or anyone representing them has demonstrated to the court that it should relieve these particular companies of petitioner's rate, if such companies are considered on an individual basis. In fact, it appears that these companies have not participated directly in the administrative or judicial proceedings and are not in a position to object on behalf of themselves. The inquiry, however, does not end here. The rates chosen for these non-complying respondents have application beyond their effects on the non-complying respondents themselves. The BIA rates for non-complying respondents were averaged with other rates from companies comprising ITA's random sample to arrive at an "all other rate." This rate is applied to those companies which were not part of ITA's random sample and which were not asked to participate in ITA's investigation.

In Asocolflores I, the court approved inclusion of the BIA rate assigned to non-responding companies (which then was Uniflor's verified rate) in ITA's calculation of an average "all other rate" for those companies that were not part of ITA's sample. At that time the court found ITA did not err in finding Uniflor sufficiently representative so as to be includable in the sample, and that the integrity of the sample required inclusion of rates for non-responding companies that were within the sample. See Asocolflores I, 704 F.Supp. at 1121, n. 11, and 1126. The court finds, however, that under the facts of this case use of petitioner's unverified rate, as part of the averaged rate, simply because it is the "highest" rate, is unreasonable.

One question to be answered here is whether, in the absence of proof of collusion among the parties, ITA should use an unverified rate in the random sample average if a verified rate is available. The court concludes that, in general, too great a burden would be placed upon the agency if proof of collusion were required before petitioner's rate is used in such a situation. But in a case such as this where, for at least three of the four non-responding companies, there seems to be demonstrable proof that the parties were not colluding, use of petitioner's rate seems to be particularly inappropriate. The record indicates that these companies had their own reasons for non-cooperation which were unrelated to the margin level, and the fact-pattern makes collusion nearly an impossibility in this case. The four companies either had business difficulties or no continuing sales of the relevant merchandise, so they may not have found it worth their while to respond.

In other cases, if it is acceptable for ITA to use petitioner's rate for individual companies and a lack of collusion is not so apparent from the fact pattern, the preferred methodology may be to use petitioner's rate in the "all other" average, because it is the rate actually assigned to those companies. In the present case, however, because the verified rates are so much lower than petitioner's rate and because some or all of the non-responding companies appear to have had no motivation to respond, even if their rates would have been far below petitioner's, blanket use of petitioner's rate to calculate the "all other rate" is not appropriate.

On remand, ITA should look at the reasons of record for non-response. If the record indicates that a company is out of business, that it made a one-time sale, or that it had a similar reason unrelated to deposit rates for not responding, a verified rate should be used for purposes of establishing an "all other rate."^[4] In other cases, not now before the court, exigencies might or might not permit the use of petitioner's rate where the all other rate is based on a random sample. When dealing with the partial application of quasi-punitive rates to innocent parties, that is, those not part of the sample, further inquiry and explanation is required by ITA before the final average "all other" rate is determined.^[5]

B. Asocolflores Challenge to Uniflor's Material Cost Estimate

In its original determination ITA used Uniflor's unverified estimate of increases in material costs for the first half of 1986. Uniflor's data was generally verifiable; thus, it was not assigned a BIA rate. To fill in some missing data, the court directed ITA on remand to use averages of verified data, if they revealed higher 1986 costs than estimated by Uniflor. Asocolflores I, 704 F.Supp. at 1118. The court found this to be a fair approach, as well as being in keeping with ITA's own statement as to its methodology for those companies which had been fully verified, except for minimal items. Plaintiffs now allege that it is unfair of ITA to continue to use the estimate, just because the average of verified cost increases is lower than Uniflor's estimate. The court concludes that this objection is too late. If parties with an interest in Uniflor's rate wished to abandon Uniflor's own estimate of its costs, they should have so advised the court prior to remand. Furthermore, if they had no confidence in the estimate, prior to final determination they should have asked the agency to use an average rate.

C. Floramerica's and FTC's Challenges as to Correction of ITA Clerical Errors

In Asocolflores I, the court instructed ITA to correct certain clerical errors noted by the parties regarding Floramerica's dumping margins. See Asocolflores I, 704 F.Supp. at 1126. Those clerical errors were allegedly: 1) that ITA incorrectly calculated the foreign market value for standard carnations for the months of January 1986 and November 1985, and 2) that the U.S. price for June 1985 had been computed erroneously. The parties are in agreement regarding the error as to computation of U.S. price. Accordingly, ITA's correction of this figure is accepted. With regard to the foreign market value error, Floramerica claims that ITA has now corrected the error for January 1986, but that an error still exists in the November 1985 calculation figure. Plaintiff Floramerica's brief at 3. FTC, for its part, claims that ITA has again miscalculated both the November 1985 and January 1986 values. See FTC's Response to Plaintiffs' Comments at 3-4. FTC claims that this mistake arose because ITA took figures from frames 1853 and 1854 (Reel 3 of the Confidential Record) rather than from frames 1889 and 1890. Id. at 4.

Defendant claims that the figures noted in frames 1889 and 1890 constitute "preliminary, unrevised figures for the entire fourth quarter, not merely for the month of November 1985." Defendant's Brief at 18-19. Upon review of these documents, it seems clear to the court that the documents alleged by FTC to be correct are in fact only preliminary assessments, as defendant posits. The court therefore affirms ITA's correction of this clerical error. As FTC's allegations regarding the January 1986 figures are based on the same incorrect assumption, that is, that ITA should have used what are now shown to be preliminary figures rather than final calculations, the court also affirms ITA's conclusion as to January 1986.

D. FTC's Objections to Calculations of Cost of Production

1. Timana's Material Costs for 1986

ITA's method for calculating Timana's material costs for 1986, is affirmed. ITA describes its remand methodology as follows:

... ITA has determined that the most accurate verified material cost figure to use in lieu of the unverified December 1985—May 1986 material costs is the verified average material cost per bunch sold in the June-November 1985 period (one-half the year-end verified material costs for 1985 divided by bunches sold during June-November 1985). Since the time period June-November 1985 is close to December 1985—May 1986, the costs can be expected to be close and can be adjusted for cost increases or decreases between these two periods. Moreover, by using the cost per bunch (i.e., per unit), ITA can multiply the June—November 1985 unit cost figure by the units sold in December 1985—May 1986 to reflect the seasonal differences of higher sales in the first half of the year and to arrive at a material cost for December 1985—May 1986 using best information available. C.R. Doc. 95 Frame 1650.

Slip Op. 89-3 also orders that any estimates for 1986 material costs based on 1985 verified costs should reflect the average change in costs from 1985 to 1986. ITA could only find verified material costs for four producers and the earliest period verified was June-November 1985. The change in material costs from June-November 1985 to December 1985-May 1986 for these four producers resulted in an average decrease of 7.8 percent. Thus, Timana's material costs of [] Colombian pesos per bunch, verified for June-November 1985 and applied to bunches sold in December 1985—May 1986, would be decreased to [] Colombian pesos to reflect the average verified decrease in material costs from the period June-November 1985 to the period December 1985—May 1986.

Remand determination at 6-7.^[6]

FTC challenged ITA's determination on this point on several grounds. First, FTC challenges ITA's method of arriving at an average verified increase. FTC's and ITA's disagreement on the amount of the average increase reflects different data pools.^[7] ITA apparently averaged in data from one company not acceptable to FTC, that is, Flores Esmeralda. FTC, on the other hand, argues that verified data from Flores Generales should have been included in the average.

The court has concluded that none of the companies whose data was used in calculating the average increase were demonstrated to be so unrepresentative that their data should have been excluded from the calculation of an average increase.^[8] FTC has failed to demonstrate that Flores Generales should have been included. The portions of the record cited by FTC and defendant do not demonstrate that Flores Generales' material costs were verified to the extent that the difference between 1985 and 1986 costs should have been included in the average of verified data which the court directed ITA to use. It is undisputed that Flores Generales' material cost data includes some amount of overhead costs. This did not prevent ITA from verifying the more inclusive "cost of manufacturing" figure, but ITA did not verify the entire "cost of materials" subset. FTC argues that because only a relatively small overhead figure is shown to be included, i.e., 5% or less, Flores Generales' material cost figure should be considered verified. The court did not direct ITA to use figures which were nearly verified, only those that were verified.^[9] In addition, ITA may have had reason to exclude the separate cost of materials figure as part of the average increase calculation because the increase calculated for Flores Generales was out of line with those derived from more completely verified information. See Administrative Record, Doc. No. 96 (reel 3) at frame 1834. Flores Generales started production in 1985, so that its cost allocation methodology may have yielded bizarre results when the data was used to calculate an increase figure. In any case, ITA essentially followed the court's instruction in utilizing fully verified data. ITA also was not required to utilize the highest cost figures available, as FTC argues. See Asocolflores I, 704 F.Supp. at 1118.

Second, FTC contends that ITA did not follow the court's direction to account for fluctuating costs over the year-long business cycle. The court concludes that ITA has done as much as it must do on this point, utilizing the information available to it, to be fair to each side.^[10] If anything, apart from its refusal to use "punitive" data, ITA has leaned toward FTC on this issue. See supra note 6.

2. La Pampa's Material Costs

As to La Pampa's costs, as indicated supra, note 8, if ITA had verified data it should have been used. Defendant has demonstrated, however, that whether averaged data for 1986 increases in costs are used, or whether La Pampa's own data is used, no difference in La Pampa's rate results.^[11]

3. Cull Sales Values as a Deduction to Cost

Implicit in the court's earlier opinion is its conclusion that there was, at a minimum, conflicting evidence on whether La Pampa "burned" all culls, and that ITA was not required to conclude that all culls were burned by La Pampa. The court directed ITA to use for fully responding companies without verifiable data on this element, the average of verified data on cull sales, as it stated it had in its original investigation. Asocolflores I, 704 F.Supp. at 1118. ITA complied on remand.

4. Interest Expenses of Flores Esmeralda

In its original opinion the court remanded this matter to ITA to explain its policies on offsetting working capital expenses and credits, that is, short term interest expenses and credits. Asocolflores I, 704 F.Supp. at 1119. In addition, the court asked ITA to explain how it arrived at the conclusion that the interest credits were earned on current operating capital, the proffered rationale for allowing the interest credits as an offset to interest expenses arising from current operations. Id.

As to the first point, ITA now has altered its position and there is no challenge to the offset methodology used on remand. The sole question remaining is whether ITA's determination that the current interest expenses were totally offset by interest earned on income from current operations is based on substantial evidence.

ITA explained that it examined Flores Esmeralda's books, thereby verifying that short term interest expenses were totally offset by short term interest earned. Remand Determination at 3. ITA then refers to generally accepted accounting principles (GAAP), which provide both that short term interest is treated as part of working capital, and that working capital is normally restricted to current operating cycles. ITA's assumption, of course, is that Flores Esmeralda records, which look like they reflect GAAP, as well as ordinary business practices, actually do so.^[12] There is the possibility, raised now by FTC, that because of Colombian economic conditions short term investments were made with other than current operating capital. While ITA may have relied too heavily on the facial appearance of the accounting records in the face of FTC's challenge, the record does not appear to contain data on the viability of short term investment of long-term assets which would allow the court to conclude that ITA erred in not looking beyond the company records. Without contradictory data, the company's records are substantial evidence on which ITA may rely.

The court finds that most of the potential harm to FTC regarding the previous calculation of Flores Esmeralda's interest expense has been eliminated by the change in offset methodology and that FTC has not met its burden of proving lack of substantial evidence in support of, or failure of ITA adequately to investigate the derivation of interest credits.

E. FTC's Challenge to Exclusion of Certain of Uniflor's Sales as Unrepresentative

As should be clear from the court's earlier discussion herein, the court viewed the record as supporting exclusion of the Uniflor sales that were made with full expectation of payment, but on which collection could not be made, as unrepresentative of Uniflor's U.S. sales behavior. On remand, ITA altered its position and arrived at the same conclusion as did the court. Remand Determination at 10. ITA has further explained on remand that in a past final determination, in which ITA also was attempting to arrive at an estimated deposit rate in similar circumstances, such sales were excluded. Id. at 10-11 (citing Fabric and Expanded Neoprene Laminate from Taiwan, 52 Fed.Reg. 37,193, 37,194 (October 5, 1987)). ITA has also explained why it treats such sales differently in annual reviews. See id., (citing Neoprene Laminate from Japan, 52 Fed.Reg. 36,295 (September 28, 1987)).^[13] The court affirms ITA's exclusion of these sales.

F. Miscellaneous Issues Not Subject of Remand

Upon reviewing Asocolflores I, the court's earlier determination herein, it appears to the court that the reader may not perceive that Section I.B.4. thereof refers to an adjustment to foreign market value based on constructed value by means of an exporter's sales price offset. See 19 C.F.R. § 353.15(c) (1988).

In addition, Asocolflores and Floramerica complain that upon remand the court wrongfully did not instruct ITA to include certain home market selling expenses in its calculation of an exporter's sales price offset for Floramerica. This was not the subject of remand and need not be considered by the court at this time. As this is obviously a very important issue to Floramerica, however, the court has reviewed the matter once more. The court remains convinced that ITA should not be required to recalculate its margins to reflect post-final determination explanations of information, because such explanations were not timely submitted to ITA. It should be noted that there is no dispute as to Floramerica's duty to prove its entitlement to this adjustment. ITA decided and the court affirmed that Floramerica did not meet this burden timely.

Among other things, Floramerica seems to argue that because this issue will affect whether Floramerica receives an operative rate or a de minimis rate, the court should order ITA to "correct"^[14] this selling expense calculation. To rule for Floramerica on such a basis alone the court must be willing to say that parties to proceedings are not required to give ITA timely, clear and accurate information, because if a party's mistake results in an important change in rates, ITA will be compelled to recalculate the rates after final determination.

Congress has required that estimated deposit rates in original proceedings be calculated within very short time periods. One cannot simply throw data at ITA without clear explanations, or provide it with misleading explanations. Accordingly, ITA did not err in declining to recalculate, but rather, apparently made a policy decision in favor of compelling accuracy and timely presentation of data, even if it means here that both ITA and the party involved will be forced to spend time in an administrative review of the rate.

Although the parties did not focus on the extent of the court's equitable powers in these actions, it appears to the court that despite ITA's rational behavior, the court may have discretion, particularly where it is remanding other matters, to order ITA to recalculate in the interest of justice and to save judicial, as well as administrative, resources. The court has declined here to exercise any such discretion it may have. This is a particularly complicated case involving several interrelated investigations, numerous producers and novel methodologies. Due to the size and complexity of the case the court believes justice is better served by keeping remand issues as narrow as is permissible and by restricting all parties to arguments based on the presentation of their cases before final determination, except on remand issues.^[15] If the court orders remand in this situation where ITA did not err in its assessment of the data, as presented in the record, and where it did not abuse its discretion in declining to recalculate, other parties, particularly those opposed to Floramerica's position, might rightly ask to raise other issues not timely addressed or to present new evidence or arguments on behalf of a position. One only need read Asocolflores I and the previous portions of this opinion for indications that this is not mere speculation. Accordingly, in the interest of fulfilling Congress' desire for expedition, the court adheres to its previous opinion on this issue.

ITA has twenty days to file its remand results on the one issue remaining. See supra Section A. As this is not an accounting issue, the court expects compliance with this deadline unless extraordinary reasons are presented as to why it cannot be met.

NOTES

[1] One group of producers, Floramerica and its related Colombian companies, filed a separate brief.

[2] This is not an insignificant rate. Some large producers had de minimis or very low rates.

[3] This policy is partially reflected in the administrative determinations in this case, but it was elaborated upon during a post-remand telephone conference among the parties and the court.

[4] Asocolflores has requested Uniflor's rate, the highest verified rate available. There is nothing in the record of this case to indicate ITA should choose a lower rate at this point in this action.

[5] It is disingenuous of defendant to argue that any producer could have participated without being part of the sample. Small producers cannot participate easily and if many parties had volunteered ITA could not have investigated them. The purpose of the sample was to avoid an unwieldy investigation. In addition, parties not volunteering are not guilty of something which requires punitive action. Parties not part of the sample have a right to expect that ITA's sampling techniques will yield a fair rate related to actual industry behavior, not a rate unduly related to petitioner's claims.

[6] ITA may have used a needlessly complicated method. The more straightforward approaches which occur to the court as satisfying the direction of Asocolflores I, however, appear to yield results less favorable to FTC, assuming data from the same companies are used to compute an average increase figure. FTC is the only party objecting to any of ITA's methodologies on this point. The original determination challenged by FTC did not yield a result significantly different from the result achieved by ITA's new method.

[7] This issue was clarified in several conferences among the court and the parties. Letter submissions and briefs, which have been made part of the record of this action, further elucidate this matter.

[8] Flores Esmeralda, which had verified data that was included in the average, was hit by a tornado in 1986. This destroyed part of crops of two flower types. FTC, however, has not demonstrated by citation to evidence of record that Flores Esmeralda's decrease in material costs is largely due to an extraordinary weather event. Other parties also had materials costs which declined on an absolute basis. More importantly, FTC has not demonstrated that on a per unit basis, which was the methodology used by ITA to which FTC objected, Flores Esmeralda's material cost change figure was at all unusual.

ITA may have "erred" in excluding La Pampa's data from its average increase calculation. Apparently, from the parties' pre-remand arguments, the court concluded in Asocolflores I that La Pampa's material cost data for 1986 was unverified. Defendant now states that they were verified. FTC states that the costs were unverified and has not argued for inclusion of La Pampa's data in the average. In this case, the inclusion or exclusion of the possibly verified data from La Pampa in the average likely had no appreciable effect. For this reason presumably, no request for clarification of the court's statement was made.

[9] Had ITA, during its investigation, needed a verified separate cost of materials figure for Flores Generales, it might have been able to obtain the information to "fill in the gaps." In Asocolflores I the court did not require ITA to reopen its investigation to complete verification of separate cost of materials data. Rather, the court directed ITA to use the verified data on hand. The court's direction was based on ITA's original explanation of methodology and FTC's arguments.

[10] The December 1984—May 1985 period was not part of the investigation and, therefore, was not a source of verified costs for comparison purposes.

[11] FTC is not pressing this challenge, and as indicated, it believes La Pampa's data is unverified.

[12] The court has cited, with approval in various cases, e.g. Ipsco, Inc. v. United States, 12 CIT ___, 701 F. Supp. 236, 238, note 3 (1988), ITA's policy of using a "firm's expenses as recorded in its financial statements as long as those statements are prepared in accordance with the home country's generally accepted accounts principles (GAAP) and do not significantly distort the firm's financial position or actual costs." (Citing Carbon Steel Products from Brazil, 49 Fed.Reg. 28,298, 28,302 (Jan. 25, 1984)).

[13] The practice with regard to annual reviews is not before the court.

[14] The court has referred to ITA's "error," only because post-final determination information reveals that an "error" probably occurred.

[15] To put this matter in a broader perspective one may wish to review the court's decisions in the related cases in which the court has upheld equally strict standards of accuracy and timely presentation of arguments and data, to the detriment of FTC, a party which opposes Floramerica in this action. See Floral Trade Council of Davis, Calif. v. United States, 12 CIT ___, 698 F. Supp. 925 (1988); Floral Trade Council of Davis, Calif. v. United States, 12 CIT ___, 704 F. Supp. 233 (1988); Floral Trade Council of Davis, Calif. v. United States, 12 CIT ___, 704 F. Supp. 241 (1988); Floral Trade Council of Davis, Calif. v. United States, 13 CIT ___, 707 F. Supp. 1343 (1989). While these decisions did not benefit Floramerica directly, it appears to the court that what Floramerica seeks here is application of a different standard than that which the court has applied to all other parties in these related cases.