American Sporting Goods v. United States

*1303 OPINION GOLDBERG, Senior Judge. This matter is before the Court following trial de novo. At issue is the proper classification of six entries of certain footwear that the plaintiff calls “chula” sandals (“sandals”). The parties agree on the basic nature of the merchandise in dispute and that it is classifiable as “[o]ther footwear with outer soles and uppers of rubber or plastics: [o]ther footwear” under heading 6402 of the 1994 Harmonized Tariff Schedule of the United States (“HTSUS”). They part company with respect to a single issue: whether the external surface area of the upper (“ESAU”) of the sandals is more than 90 percent rubber or plastic. If, as the plaintiff claims, such is the case, the parties agree that the sandals are classifiable under subheading 6402.99.15, HTSUS, and dutiable at a rate of 6 percent ad valorem. If, however, the ESAU is less than or equal to 90 percent rubber or plastic, then Customs’s classification of the sandals under subheading 6402.99.30, HTSUS, and assessment of duty at a rate of 37.5 percent ad valorem must stand. Upon review of the merchandise, the exhibits, the testimony of record, and the applicable law, the Court finds in favor of the plaintiff. I. BACKGROUND Prior to the dispute that gave rise to this litigation, the plaintiff imported and distributed chula sandals manufactured according to its design and specifications by an independent factory in Shanghai, China. 1 In May and June, 1994, the plaintiff imported a total of six shipments comprising 65,736 pairs of such sandals in various sizes and styles for children, youth, boys, men, and women, and entered them under HTSUS 6402.99.15 as “[o]ther footwear with outer soles and uppers of rubber or plastics: [o]ther footwear: [o]ther: [h]av-ing uppers of which over 90 percent of the external surface area ... is rubber or plastics .. .:[o]ther: [o]ther....” In June 1994, Customs tested a single sample from the first shipment, a size 10 in one of the two men’s styles, and determined that its ESAU was 89.7 percent rubber or plastic. Consequently, Customs liquidated all six shipments of sandals under subheading 6402.99.30, HTSUS, as “[o]ther footwear with outer soles and uppers of rubber or plastics: [o]ther footwear: [o]ther: [o]ther: [f]ootwear with open toes or open heels; footwear of the slip-on type, that is held to the foot without the use of laces or buckles or other fasteners.... ” The plaintiff protested the classification and provided Customs with five additional sandals in other styles and sizes, as well as analyses of the other half of each pair conducted by an independent commercial laboratory that determined the ESAU of each sample to be above 90 percent rubber or plastic. Customs declined to conduct tests on these additional samples, denied the protest on November 9, 1994, and denied a request for reconsideration on February 13, 1995. The plaintiff then appealed the denial of protest to the Court of International Trade, and the matter proceeded to bench trial. The Court exercises jurisdiction pursuant to 28 U.S.C. § 1581(a). *1304 II. DISCUSSION A. Findings of Fact Based on the evidence adduced at trial, together with the supporting exhibits, the Court hereby adopts the following findings of fact. 1. Production of the subject merchandise Mr. John Thomas, vice-president of the plaintiff, 2 testified that eligibility for the lower tariff rate applicable under subheading 6402.99.15, HTSUS, was, together with aesthetic appeal, the primary consideration in the design of the sandals. Before mass production of the subject merchandise began, the plaintiff prepared a prototype pattern in a size 8, and confirmed that it exceeded the 90 percent threshold by several percentage points, as intended. Prototypes for the other models and sizes were unnecessary because the sandal design was scalable, so that the percentage of rubber and plastic in the ESAU would not change markedly from one type of sandal to the next. The importance of maintaining the “duty feature” was communicated to the Shanghai factory fulfilling the plaintiffs orders, a request familiar to the factory from its production for other U.S. importers. However, the production of sandals is not a matter of scientific precision. Production variances occur because the upper is connected to the sole by hand. The sole is mounted on a last, and the straps of the sandal are pulled through the slots in the sole and fastened; how slackly or tautly this is done necessarily affects the total ESAU of the sandal. Because hand-lasting introduces greater risk of production variances, the plaintiffs employees visually inspect ten percent of all pairs in a shipment for unacceptable production variances, and examine the entire lot if five percent of the initial pairs sampled evince a defect. 2. Measurement of the ESAU of footwear Customs Laboratory Method 64-01 (“Method 64-01”). establishes the proper methodology for measuring the ESAU of footwear. 3 According to Method 64-01, either of two instruments is permissible to conduct such measurements: a polar compensating planimeter (“polar planimeter”), or an image analyzer. Because both of these instruments are capable of measuring only two-dimensional surfaces, the three-dimensional footwear uppers must be reduced to two dimensions in order to be measured. To that end, Method 64-01 directs the tester to: (1) identify all external surfaces to be included in the determination of the size of the upper; (2) cut off all such external surfaces, and, if necessary, cut them again so that they lie flat; (3) trace around the border of surfaces to be measured in white if necessary to heighten the contrast with surrounding surfaces; (4) lay the detached external surfaces flat and photocopy the image; 4 (5) trace around *1305 the area of each material type (e.g., plastic or textile) on the photocopied image of the ESAU; (6) use a polar planimeter or image analyzer to measure at least twice the relative surface area of each material; and (7) calculate the -average value for the area of each material and the average total area of all materials, and then calculate the relative percentage area for each material. According to Customs’s guidelines, the external surface to be included in the determination of the area of the upper “is, in general, the outside surface of what you see covering the foot ... when the [footwear] is worn.” Treas. Decision 93-88 (Oct. 25th, 1993). Thus, with respect to the sandals at issue here, a tester would not include the lower, inner portion of the strap binding the wearer’s foot in place, to the extent that it were overlaid by (and attached by velcro to) the upper, outer portion of that same strap. Unspecified by Method 64-01 and Customs guidelines, however, is the extent to which the upper strap should overlie the lower one. This issue has ramifications in the instant case, because a portion of the lower strap is composed of textile (i.e., not rubber or plastic) that is exposed if the velcro on the upper strap is not matched precisely to the velcro on the lower strap. 3. Deficiencies in Customs’s test for ESAU In June 1994, the Customs Testing Laboratory analyzed the sample pulled at random from the plaintiffs initial shipment of subject merchandise. The first Customs analyst, Mr. Melvin Barber, tested the sandal twice using an image analyzer, and determined the ESAU to be composed of 89.3 and 90.1 percent rubber or plastic on the first and second occasions, respectively. Because his results were close to and above the 90 percent threshold, a different Customs tester, Mr. Brian Kennedy, performed a third test, and obtained a result of 89.7 percent rubber or plastic. The raw data for the three tests was aggregated, yielding a final, official determination that the ESAU was 89.74 percent rubber or plastic. On the basis of this determination, Customs liquidated the sandals under subheading 6402.99.30, HTSUS, and denied the plaintiffs subsequent protest. The sample and a single rather poorly-contrasted photocopy of the sample were retained, but the individual calculations with respect to each part of the sample were lost. Neither Mr. Barber nor Mr. Kennedy was deposed or testified at trial. In 1997, after the instant suit was initiated, Customs retested the original sample five times, and obtained measurements of the ESAU ranging from 88.2 to 89.9 percent rubber or plastic. In preparation for the instant litigation, the plaintiff commissioned Consumer Testing Laboratories (“CTL”), 5 to examine the sample tested by Customs as well as Customs’s analyses of that sample. Mr. He-mant Patel, 6 who at that time was CTL’s laboratory technical manager, used a polar planimeter 7 to test both the image ob *1306 tained by Customs in 1994 from the original sample, as well as photocopies of the sample that he made himself. 8 He found that the ESAU of Customs’s own 1994 image was 90.47 percent rubber or plastic. His calculations for each of the three photocopies that he made of the original sample yielded an ESAU of 92.10, 91.29, and 91.08 percent rubber or plastic, respectively- In addition, Mr. Patel found that approximately 0.25 square inches of the ESAU, composed of rubber or plastic, was improperly not severed from the outer sole of the original sample. He calculated that including this portion in the determination of the ESAU would increase the rubber or plastic proportion of that area by approximately 0.05 percentage points. Dr. Doe-meny conceded Custom’s error, but stated that the magnitude of the error would not materially alter Customs’s classification decision. Mr. Patel suggested that the discrepancies between Customs’s results and his own could be inferred from deficiencies in Customs’s 1997 tests. First, Mr. Patel emphasized that an accurate analysis requires that the samples be completely flat when photocopied or scanned. The presence of shadowing, and the oval shape of circles on the straps, indicated that the images that Customs obtained in 1997 were not made when the sample was lying completely flat. Second, he testified that the velcro visible on the lower strap in the 1997 images indicated that Customs’s tester had failed to “normalize” the straps— i.e., to match velcro-to-velcro, which both he and Mr. Thomas testified was the normal position for the straps. 9 Dr. Paul Doemeny and Ms. Marian Sa-marin of the Customs Service Laboratory 10 both took the stand in part to refute Mr. Patel’s testimony. Ms. Samarin explained that the 1997 tests were conducted purely in consideration of the pending litigation, and that the Customs testers were deliberately tweaking the sample as an experiment. 11 However, neither Dr. Doe-meny nor Ms. Samarin had direct knowledge of how the 1994 image was obtained, so neither was able to testify as to whether the sample in that image was lying completely flat. In addition, both gave contradictory testimony with respect to whether the lower and upper straps should be normalized so that they match velcro-to-velcro. Dr. Doe-meny testified that in the Initial test, the straps were matched velcro-to-velcro, but subsequently stated that they should be laid according to their “natural crease.” Ms. Samarin testified that it was not nor *1307 mal Customs practice to match velcro-on-velcro. However, in her deposition testimony, she claimed that it was normal practice to match velcro-to-velcro, but that to do so would not be appropriate in this case because the strap would bulge, and that instead the strap should be cut to eliminate such a bulge. There was no evidence that any Customs analyst consistently followed this or any other practice, however. 4. Plaintiffs tests for ESAU After learning of Customs’s initial test result finding an ESAU of 89.7 percent rubber or plastic, one Ms. Jane Y. Mo, an employee of the plaintiff who did not testify, 12 apparently contacted the Shanghai factory to confirm that the sandals were produced to the plaintiffs specifications including with respect to the percentage of rubber and plastic in the ESAU. Mr. Thomas also instructed Ms. Mo to select 13 additional samples from the entries for analysis by CTL. 14 At CTL, Mr. Jim Bibeault 15 tested the five samples, which included sizes 9 and 11 of men’s sandals in the same style as the initial sample tested by Customs; ladies’ sizes 5 and 7; and a youth’s size 6. The aggregate percentage of rubber or plastic in the ESAU of each sample was 91.7; 91.8; 91.3; 92.9; and 91.7, respectively. Mr. Bibeault’s work and results were supervised by Mr. Patel, and Mr. Bibeault testified that he paid particular care to the necessity of ensuring that the sample was flat before photocopying, and of normalizing the straps. B. Conclusions of Law Customs’s tariff classification decisions enjoy a statutory presumption of correctness, and the burden of proving otherwise rests upon the party challenging such decisions. 28 U.S.C. § 2639(a). However, the presumption of correctness “does not add evidentiary weight; it simply places the burden of proof on the challenger.” Anhydrides & Chems., Inc. v. United States, ,130 F.3d 1481, 1486 (Fed.Cir.1997). 16 Thus, although the methods of weighing, measuring, and testing merchandise used by [C]us-toms officers and the results obtained ' are presumed to be correct, ... this presumption may be rebutted by show *1308 ing that such methods or results are erroneous. Furthermore, the presumption does not have evidentiary value and may not be weighted against relevant and material proof offered by the plaintiffs. If a prima facie case is made out, the presumption is destroyed, and the Government has the burden of going forward with the evidence. Aluminum Co. of America v. United States, 60 C.C.P.A. 148, 151, 477 F.2d 1396, 1398-99 (1973) (quoting with approval Consolidated Cork Co. v. United States, 54 Cust.Ct. 83, 85 (1965)) (citations omitted). In Aluminum Co., the court considered the claim of a plaintiff who argued that Customs had erred by classifying the subject merchandise as fluorspar containing not over 97 percent by weight calcium fluoride. The court found that the plaintiff established its prima facie case by submitting evidence of multiple analyses of the merchandise, conducted according to Customs’s own established testing method, each of which showed the calcium fluoride content to exceed 97 percent. 60 C.C.P.A. at 151, 477 F.2d at 1399. After weighing the evidence, the court found that the preponderance of the evidence supported the plaintiffs proposed classification, and ruled accordingly. Id. at 151-52, 477 F.2d at 1399-1400. Thus, the plaintiff in a case such as this may make out a prima facie case either by showing that Customs’s results or methods are erroneous, Consolidated Cork, 54 Cust-Ct. at 85, or by “submitting evidence of analysis [that the plaintiff] applied to the merchandise which gave a result different from that claimed by the Government.” Aluminum Co., 60 C.C.P.A. at 151, 477 F.2d at 1399. The plaintiff here has done both. The question becomes, therefore, whether this evidence, when weighed against that produced by Customs, is such that the plaintiff has carried its burden of proving by a preponderance of the evidence that the rubber or plastic content of the ESAU of the subject merchandise is greater than 90 percent. The Court finds that the plaintiff has met its burden, for three related reasons. First, the plaintiff raised some doubts about the precision of Customs’s 1994 test, by noting the minor but still important failure to sever the upper cleanly from the sole, and by highlighting Customs’s inability to define a standard practice with respect to matching the upper and lower straps, suggesting a confused or inconsistent approach in that regard. While it may not be fair to impute the deficiencies in Customs’s 1997 tests to the 1994 tests (about which Customs’s records are scant), as Customs’s witnesses and counsel expressly disclaimed reliance on the 1997 tests, that fact points to a second consideration. Stated plainly, the plaintiff adduced evidence of its own test on six different samples, each of which was determined to have an ESAU above 90 percent rubber and plastic, whereas Customs relied only on its original, underdocumented test on a single sample. 17 More extensive testing is particularly likely to be more reliable with respect to merchandise such as the plaintiffs that is subject to normal production variances. Nor was Customs able to point *1309 to any significant flaws in the plaintiffs own test results. Still, the plaintiffs evidence can hardly be called overwhelming, 18 and it might not have been enough to triumph were it not for the fact that Customs went to trial already backed up against its own goal line. That is to say, the third and decisive factor in this case is that Customs’s evidence that the ESAU was equal to or less than 90 percent rubber or plastic is inherently weakened by the fact that its own tests produced results so close to the borderline. To be sure, there are no ties in Customs classification decisions, and if the evidence clearly showed that the ESAU of the subject merchandise were 89.99 or even 90.00 percent rubber or plastic, Custom’s classification would be upheld. At the same time, the Court believes it is incumbent upon the Customs Service to recognize that in especially close cases— where even one of Customs’s own tests exceeds the 90 percent threshold — the agency has a special duty to ensure that its determinations are accurate and well-substantiated. As the court explained in Consolidated Cork, “the final determination in situations where the merchandise approaches the borderline set by the tariff act depends upon the accuracy of the methods used and their application by the chemists who performed the tests.” 54 Cust.Ct. at 87. The court in that case scorned the decision of Customs officials to test only 11 ounces from an entry of 22,050 pounds of granulated cork. “For such a result to be drawn from so small a sample, extreme niceness of weighing and measurement is required. The slightest error would be fantastically multiplied in the final result.” 54 Cust.Ct. at 88. The court likewise observed that the plaintiffs experts conducted fifteen tests compared to Customs’s single test. See 54 Cust.Ct. at 88 (“Since 15 tests were made, it is possible to check the accuracy of the tests against each other.... As the record stands, these tests are more likely to be accurate and representative of the density of the entire shipment than the single one made by the Government chemist.”). The Court does recognize the practical limitations on the resources of both the Customs Service and the import community, and does not suggest that it is either practicable or desirable that the testing process continue ad museum, particularly if it initial results do not suggest a close case. In this instance, however, Customs’s results were close to the threshold, and the plaintiff adduced additional evidence showing both deficiencies in Customs’s tests and the probable outcome of those tests had they been error-free. While the plaintiffs case is not overwhelming, it is enough to meet its burden to prove by a preponderance of the evidence that the ESAU of the subject merchandise was greater than 90 percent rubber or plastic. III. CONCLUSION The preponderance of the record evidence establishes that Customs erred in classifying the subject merchandise under subheading 6402.99.30, HTSUS, and that the merchandise is properly classifiable *1310 under 6402.99.15, HTSUS. The Court will enter judgment accordingly. Footnotes [1] According to the plaintiff, Customs's adverse tariff ruling led it to modify subsequent production of chula sandals to increase the percentage of rubber or plastic in the ESAU. This alteration proved to be aesthetically unappealing and destroyed the commercial viability of the product, leading the plaintiff to cease production and importation of the sandals. [2] Mr. Thomas is responsible for design, development, and quality control, and has more than 30 years experience in product design. His testimony was substantially unimpeached. [3] The parties agree that Method 64-01 sets forth the accepted methodology for testing the ESAU of footwear, but dispute whether their respective testers properly adhered to it. Because the reliability of the methodology is not in doubt, the standards for evaluating reliability enumerated by the Supreme Court in Daubert v. Merrell Dow, 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993), and by the Court of Appeals in Libas, Ltd. v. United States, 193 F.3d 1361 (Fed.Cir.1999), are not relevant to the disposition of this case. [4] Despite the instruction in Method 64-01, strictly speaking it is no longer necessary to produce a photocopy for use by an image analyzer, as an image analyzer is usually paired with a photoreceptive scanner that records a two-dimensional image of the sample. *1305 Nevertheless, the necessity of ensuring that the sample lies flat remains constant in either case. [5] CTL is a commercial testing laboratory that regularly tests the ESAU of footwear for persons in the import community. The Court of International Trade accepted CTL's footwear ESAU analyses in a previous case. See Hi-Tec Sports USA v. United States, 21 CIT 212, 213, 958 F.Supp. 637, 638 (1997). [6] Mr. Patel holds a master's degree in mechanical engineering from the Massachusetts Institute of Technology, and has tested footwear at CTL since the 1970’s, conducting several thousands ESAU tests. He is now a vice-president at CTL. [7] Mr. Patel used a Planex-5 planimeter, which had been tested and shown to be accurate within -0.062 percent, well within the ±0.2 percent tolerance specified by Method 64-01. [8] Mr. Patel did not test the first set of photocopies that he made of the sample, because he found that he had not followed the correct procedure for obtaining a proper photocopied image. [9] Mr. Patel testified — although his recollection of this event was hazy — that a Customs official lecturing at a footwear conference had stated that the normal procedure was to match velcro to velcro. One of the two Customs officials at the conference testified that he could not recall giving such an instruction. Mr. Thomas explained that, from a design and marketing standpoint, matching velcro-to-vel-cro produced the most attractive appearance for the sandal. [10] Dr. Doemeny holds a Ph.D. in Chemistry and has worked as an assistant research chemist at the Customs Testing Laboratory since 1973. Ms. Samarin has worked at the Customs Testing Laboratory as a full-time footwear analyst since 1990. [11] The Court gives credence to this testimony. At the same time, it bears notice that Ms. Samarin did maintain that the original sample was positioned completely flat for the 1997 tests. Ms. Samarin testified that the apparent oval shape of the circles could be shadows caused by the scanner for the image analyzer, and Dr. Doemeny testified that the apparent shadows in the 1997 images could actually be velcro. [12] The substance of Ms. Mo’s actions were recounted by Mr. Thomas. Although Customs did not object on this ground, much of the testimony concerning Ms. Mo’s actions is hearsay. The Court recounts it here because it is useful background information and is not central to the disposition of the case. [13] Mr. Thomas testified that he instructed Ms. Mo to pull, or have the warehouse manager pull, the samples at random, and that she would have done so, but there is no real evidence either to confirm or refute this detail. [14] The plaintiff also sent a sample or samples for analysis to U.S. Testing, a different commercial testing laboratory, but there is no evidence that such an analysis was ever conducted. [15] Since June 1993, Mr. Bibeault has worked as a testing engineer at CTL, where he performs approximately 300 polar planimeter tests a year. [16] Moreover, the presumption of correctness applies only to the factual basis of such decisions, and not to their legal component, with respect to which the Court of International Trade exercises de novo review. See Universal Elecs., Inc. v. United States, 112 F.3d 488, 492 (Fed.Cir.1997); Anhydrides, 130 F.3d at 1485-1486. However, because the Court agrees with the parties that the subject merchandise should be classified under either of their competing proposed classifications, the disposition of this case turns on a purely factual issue, viz., whether the ESAU of the sandals is greater than 90 percent plastic. Cf. Anhydrides, 130 F.3d at 1482-83 ("The application of the correctly interpreted tariff classification to a particular article is a question of fact.”). [17] In this case, Customs has persisted in a mistaken belief that some special talismanic power or authority attaches to the results of the test on which it based its original classification decision. The Court must make its determination on the basis of the record before it, comprising the evidence introduced at trial, rather than that developed by Customs. Automatic Plastic Molding, Inc. v. United States, 2002 WL 31243112, **2-3 (CIT 2002) (quoting United States v. Mead Corp., 533 U.S. 218, 233 n. 16, 121 S.Ct. 2164, 150 L.Ed.2d 292 (2001)). In light of Customs's poor custody of the data for the initial test results and its failure to make available the persons who conducted the 1994 tests, any other rule would make it very difficult indeed for the plaintiff to prevail. [18] In particular, the Court notes that the plaintiff failed to substantiate testimony that the additional samples were pulled at random, and failed to have more than one commercial testing facility analyze the subject merchandise. Cf. Aluminum Co., 60 C.C.P.A. at 151, 477 F.2d at 1400 (noting with approval that multiple analysts at multiple facilities had tested plaintiff's merchandise).