Smith Corona Corp. v. United States

698 F. Supp. 240, 12 Ct. Int'l Trade 854, 12 C.I.T. 854, 1988 Ct. Intl. Trade LEXIS 258

• Court: United States Court of International Trade
• Decided: September 20, 1988
• Docket: Consolidated Court 87-02-00157
• Status: Published

Opinion

OPINION AND ORDER

AQUILINO, Judge:

The International Trade Administration, U.S. Department of Commerce (“ITA”) has issued the Final Results of Revised Scope Determination for Antidumping Duty Order on Portable Electric Typewriters from Japan Pursuant to Court Re-mand 1 The remand was pursuant to an order in conjunction with this court’s slip op. 87-145, 11 CIT -, 678 F.Supp. 285 (1987), and has resulted in a determination that portable electric typewriters containing calculators fall within the ambit of the antidumping-duty order, 45 Fed.Reg. 30,-618 (May 9, 1980), while portable electric typewriters capable of text memory do not.

The plaintiff has interposed a motion for judgment on the agency record, as supplemented during remand, requesting, among other things, that the court

(1) hold that the ITA’s final remand re-determination with respect to portable electric typewriters incorporating text memory is unsupported by substantial evidence, (2) hold that the ITA’s final remand redetermination with respect to portable electric typewriters incorporating a calculator function is supported by substantial evidence and in accordance with law, (3) reverse the redetermination with respect to text-memory portable typewriters, (4) remand the action to the agency for publication of a revised determination .... 2

The intervenor-defendants oppose plaintiffs motion, and all but Nakajima All Co., Ltd. (“Nakajima”) (and the defendant) seek reversal of the ITA’s redetermination regarding typewriters with calculators. In-tervenor-defendants Brother Industries, Ltd. and Brother International Corp. (“Brother”) assert in addition that the ITA’s analysis of the scope issue by means of the criteria set forth in Diversified Products Corporation v. United States, 6 CIT 155, 572 F.Supp. 883 (1983), is not in accordance with law, although the result is correct with regard to the typewriters with text memories. Similarly, intervenor-de-fendants Matsushita Electric Industrial Co., Ltd., Kyushu Matsushita Electric Industrial Co., and Panasonic Company and Panasonic Industrial Company, Divisions of Matsushita Electric Corporation of America (“Panasonic”) urge the court to find that the typewriters at issue had been expressly excluded from the original order 3 . Inter-venor-defendants Canon Inc. and Canon U.S.A., Inc. (“Canon”) stress the same point and also contend that “[tjhere is ... no record evidence supporting the assertion that portable automatic typewriters did not exist at the time of the original investigation” 4 , and that TSUS item 676.0510 referenced in the 1980 order should control the scope.

The parties have fully briefed and argued their respective positions.

I

As discussed in slip op. 87-145, the original antidumping-duty order covered typewriters which were portable and electric. By the time of the first administrative review thereof, technology had brought forth typewriters which are portable and “electronic”. This advancement led the ITA, in conjunction with publication in 1983 of the preliminary results of that first administrative review, to invite interested parties to comment on whether these portable electronic typewriters are within the scope of the 1980 order. Thereafter, the agency concluded that they are — in reasoning reported sub nom. Portable Electric Typewriters From Japan; Final Results of Administrative Review of Anti-dumping Duty Order, 48 Fed.Reg. 7,768, 7,769-70 (Feb. 24, 1983), and repeated at length in slip op. 87-145. The agency’s reasoning was based on (1) general physical characteristics, (2) the expectations of the ultimate purchaser, (3) the ultimate use of the merchandise and (4) the channels of trade in which the merchandise moves, the same approach approved by the court in Diversified Products Corporation v. United States, supra, for determining whether merchandise developed after issuance of an antidumping-duty order is nevertheless within the order’s scope. 5

In any event, the ITA concluded that its order covers portable electric and electronic typewriters, which are referred to from time to time hereinafter as “PETs”. Technology has not stood still since then, however, with some machines equipped with calculators and more recently text memories, the latter of which give rise to characterizations of “automatic”. In slip op. 87-145, the court found the record “clear that the typewriters at issue ... were not in existence at the times of either the filing of [the] original petition or the issuance of the May 1980 order” 6 and that the order did not expressly exclude them from its coverage. See 11 CIT at -, 678 F.Supp. at 289-90. See also Royal Business Machines, Inc. v. United States, 1 CIT 80, 507 F.Supp. 1007 (1980), aff'd, 669 F.2d 692 (CCPA 1982); Diversified Products Corporation v. United States, 6 CIT at 160, 572 F.Supp. at 888. This finding was made in the face of intervenor-defendants’ contentions to the contrary.

The ITA’s scope determination originally challenged in this case was that the PETs with calculators or text memories were excluded from the 1980 order. However, in response to plaintiff’s motion to reverse that determination, the defendant took the following position:

After reviewing the administrative record in this ease along with arguments advanced by plaintiff in support of its motion for judgment upon the agency record, the ITA concedes that its determination that PETs incorporating calculators or text memory were specifically excluded from the original investigation and order is not supported by substantial evidence in the administrative record. Furthermore, Commerce concedes that its prior conclusion to the effect that under the criteria established in Diversified Products, ... PETs incorporating calculators or text memory should be excluded from the scope of the outstanding antidumping duty order on portable electric typewriters from Japan is not supported by substantial evidence. Therefore, the Department requests that that part of this consolidated case which deals with the scope issue (the entire Court No. 84-1-00046 and Count 8 of the complaint in Court No. 87-02-00157) be remanded so that it can reconsider its scope determination and publish a revised scope determination which would be supported by substantial evidence in the administrative record. 7

In contrast to this position, the inter-venor-defendants urged the court to search the record for substantial evidence. The court did so, but found the facts tending to support their position insubstantial and therefore granted the remand.

Now that the record has been supplemented, the intervenor-defendants do not show that it contains evidence that the PETs presently at issue had been excluded from the 1980 antidumping-duty order. Rather, they repeat arguments already considered unpersuasive in slip op. 87-145, 11 CIT -, 678 F.Supp. 285.

II

Under 19 U.S.C. § 1516a(b)(l)(B), the standard of review to which the court must adhere in considering the remand results is whether they are supported by substantial evidence on the record and are otherwise in accordance with law. See, e.g., Ceramica Regiomontana, S.A. v. United States, 10 CIT -, 636 F.Supp. 961, 965 (1986), aff'd, 810 F.2d 1137 (Fed.Cir.1987). As pointed out, for example, in Carlisle Tire & Rubber Co. v. United States, 9 CIT 520, 523, 622 F.Supp. 1071, 1074-75 (1985), the “substantial evidence standard ‘frees the reviewing courts of the time-consuming and difficult task of weighing the evidence, it gives proper respect to the expertise of the administrative tribunal and it helps promote the uniform application of the statute’ ”, quoting Consolo v. Federal Maritime Commission, 383 U.S. 607, 620, 86 S.Ct. 1018, 1026-27, 16 L.Ed.2d 131 (1966). Nevertheless, findings must be based on “ ‘such relevant evidence as a reasonable mind might accept as adequate to support a conclusion.’ ” Universal Camera Corp. v. NLRB, 340 U.S. 474, 477, 71 S.Ct. 456, 459, 95 L.Ed. 456 (1951), quoting from Consolidated Edison Co. v. NLRB, 305 U.S. 197, 229, 59 S.Ct. 206, 217, 83 L.Ed. 126 (1938).

Such evidence exists in the record, as supplemented upon remand, with regard to the ITA’s determination that PETs with calculators are within the scope of the 1980 antidumping-duty order. Just as the original record was found inadequate to sustain their exclusion from the order, the ITA has reached the same conclusion in its redeter-mination. In finding PETs incorporating calculators covered by the order, the agen cy relied on the Diversified Products criteria.

As for the first of those criteria, the general physical characteristics of the PETs with and without calculating mechanisms are virtually identical. For example, as pointed out by the ITA:

... This is apparent when comparing Canon’s model S-15 prior to and following the addition of the calculating function. We compared Canon Inc.’s model S-15 shown in the Wholesale Price Schedule ... dated February 16, 1988 ... to the model S-15 depicted in Canon’s Wholesale Price Schedule for July 1, 1986, Appendix 5. These products are exactly the same except for the calculating mechanism. Remand Results, p. 7.

The respondents argued below that there are significant physical differences. Brother, for one, described the internal distinctions of its EP-5 from standard PETs as follows:

... A standard electronic PET, incorporating a central processing unit (“CPU”), a read only memory (“ROM”), and a random access memory (“RAM”), is essentially a process control computer. It lacks a user memory found in more sophisticated machines, including the [PETs with calculators] and [PETs with text memories]....

... The EP-5’s mathematical computing capability results from the software embedded onto the ROM chip. The EP-5’s ROM circuitry design creates a discrete logic pattern, which is distinctive from that of a standard PET's ROM. Software is also embedded on the electronic PET ROM. However, the electronic PET ROM and the EP-5 ROM are not interchangeable or substitutable. To obtain the calculating capabilities associated with a [PET with a calculating mechanism], only a ROM with calculating software designed into [it] can be used.

The distinctiveness of the ROMs cannot be understated. The physical differences therein give a standard electronic PET and EP-5 their unique capabilities. 8

This may be true, but the court is not persuaded that what amounts to the implantation of a single ROM chip in an otherwise identical machine results in a significant, general physical distinction. Diagrams demonstrate that this implant is not substantial. Moreover, to repeat, PETs with and without a calculating mechanism generally exhibit the same external characteristics. See, e.g., S.R.Doc. 12, Attachments 2 and 5.

As for consumer expectations, the ITA reviewed the record, as supplemented, and noted the absence of a “price premium” for a calculator in finding that the expectation of the consumer is not influenced by that additional feature. See Remand Results, p. 8; S.R.Doc. 27, Exhibit 6. If the embedded calculators and the complexities of their chip designs and internal circuitry distinguish the PETs at issue from those originally under the order, the evidence regarding prices at which those machines have been available does not indicate that any such distinction is substantial. For example, in its Remand Results, the ITA noted at (pages 8-9) that

Canon significantly decreased its price for the S-15 after adding the calculating feature to the model. In Canon’s Comments, Appendix 2, the model S-15 lists for a suggested retail price of $329.95 ... for February 1, 1986. At this time the S-15 did not have a calculating function ....

Subsequent to this, ... the Wholesale Price Schedule for July 1, 1986 lists a suggested retail price of $299.95 for the S-15 which now incorporates a calculating function but is in all other aspects identical to the earlier model S-15. This is a substantial reduction in selling price over this short time period and is not consistent with price movements of other models listed in Canon’s two price schedules.

The record further reflects that PETs with and without calculators have nearly identical uses. The calculator “only allows the operator to perform simple calculations that can be done in one horizontal line and does not allow for columnar calculations.” Remand Results, p. 7. Based on this limited capability, it is clear that the primary use of such a PET cannot be as a calculator, rather still as a typewriter. Cf S.R. Doc. 13, Exhibit 11. Indeed, as to the final criterion, none of the parties dispute, and the record reflects, that the channels of trade for PETs with and without calculators are identical.

In sum, the ITA’s determination after remand that the addition of a calculating mechanism to a portable electronic typewriter does not exclude that product from the same class or kind of merchandise subject to the investigations is supported by substantial evidence in the record and is otherwise in accordance with law.

Ill

The same cannot be said, however, for the ITA’s determination as to the automatic PETs. That is, a careful examination of the record, as supplemented, covering those typewriters fails to reveal substantial evidence in support of their exclusion from the realm of portable electric typewriters. The information now in the record which the ITA apparently views as “substantial” consists primarily of magazine articles, brochures of various models of typewriters, advertisements from discount stores, packing materials, and users’ guides, which do not provide much underpinning for the determination reached after the agency’s analysis of the Diversified Products criteria. Furthermore, the other information in the record — affidavits attempting to conjure up consumer motives, quotations of costs and prices in no particular contexts, the transcript of another case on classification of an automatic typewriter — does not add facts substantial enough to sustain that determination on the record.

Of course, this does not mean that there are no differences between the PETs with text memories and the portable typewriters in existence when the administrative proceedings first began. The Remand Results reason as follows with regard to automatic typewriters then and now:

Like electronic typewriters, which were initially considered “office” typewriters, automatic typewriters were marketed to the business and government sectors where heavier use is required and repetitive tasks performed, whereas the elec-tro-mechanical typewriters subject to the original investigation were lighter in weight and marketed for home and student use. Technological advances over the years have brought automatic typewriters down to a portable level and they are now also marketed as consumer products.

The Department has determined that automatic typewriters were not subject to the investigations because they were not portable. Therefore, we have considered the criteria set forth in Diversified Products in determining whether automatic typewriters which now are portable are of the “class or kind” of merchandise subject to the order. We have concluded that, based on the criteria set forth in Diversified Products, automatic typewriters are not of the same class or kind as typewriters subject to the original investigations. Remand Results, p. 4.

That conclusion is not premised, however, on evidence that a reasonable mind might accept as adequate to show that the addition of text memories makes the typewriters something other than PETs.

A

As to general physical characteristics, automatic portable typewriters were determined by the ITA to be only “slightly different” than electronic PETs without text memories in outward appearance. Id. However, according to the ITA,

[m]ore importantly, the internal circuitry of an automatic typewriter is substantially different from that of the electromechanical units originally investigated and electronic typewriters determined to be within the scope of the order in 1983. Remand Results, p. 4.

While plaintiff’s papers point to the similarity in the component make-up — specifically, random access memory, read only memory, and a central processing unit — of both automatic and non-automatic PETs 9 , the defendant claims that the “key distinction” is the sophistication of those components, which “ ‘allow for the same memory capabilities as found in word processors and computers’ ” 10 . The defendant and the intervenor-defendants essentially argue that, unlike the single function — typing— performed by electro-mechanical and electronic portable typewriters, those with text memories have “word-processing” capabilities (e.g., encoding, storing, reading, and printing information) which require “vastly more complex” software that is “manifested in the microelectronic circuitry embedded on the chips” 11 . Further:

... Once the software has been embedded onto the ROM, the electronic PET ROM and the [automatic PET] ROM are neither interchangeable or substitutable without changing the machine’s performance capabilities. S.R.Doc. 13, p. 15.

The respondents below had made similar arguments with regard to PETs with calculating mechanisms. However, the record is not convincing that a chip implanted to provide text memory amounts to a substantial difference from one embedded to provide calculating capability. And, as pointed out in Point II above, the ITA did not consider such an enhancement of consequence for the calculator PETs.

While the supplemented record supports that result, it does not support a contrary conclusion as to PETs with text memories.

As for “functional differences” 12 between automatic and non-automatic typewriters, the defendant cites a comparison by Brother of the features of its “more sophisticated” 13 model with text memory, the CE-380, with an out-of-production elec-tro-mechanical model, the XL-20 + 3, which was investigated at the time of the original order. See Defendant's Memorandum, pp. 19-20. Described as “technologically ... eons ahead” of the older machine 14 , the CE-380 contains an “optional 16k memory card and thesaurus”, 160-character liquid crystal display, spelling correction, and correction mechanism by electronic lift-off. See S.R.Doc. 13, Exhibit 7. By comparison, the most advanced features of an XL-20 + 3 were a built-in correcting ribbon and a “power electric carriage return”. See id., Exhibit 3.

While such components constitute physical differences, they do not add up to a different class or kind of merchandise. Any comparison of features of a manufacturer’s older or obsolete machines with those of its most advanced should show differences, but such a phenomenon is not necessarily conclusive of substantially distinct general physical characteristics. Indeed, even when comparing the XL-20 + 3 with another out-of-production, electronic model, the Brother EP-20, the features are also “functionally” different. The EP-20 has been described as a

personal electronic printer [that] is a lightweight, portable, battery-powered unit with a calculating mechanism.... A calculating function handles the four basic functions, and a second shift function will print extra characters from international languages.

... A 16 digit display allows the user to make corrections before actual print-ing_ R.Doc. 7.

The correction mechanism contained in this machine is one of many distinct elements which comprise its physical charac teristics and distinguish it from the electromechanical model. Yet, the ITA has now determined that the EP-20 falls within the scope of the order covering the XL-20 + 3. Cf. Remand Results, pp. 6-7. Thus, notwithstanding the technological advances evidenced both internally and externally by the EP-20 15 , understandably, it remains within the same class or kind of merchandise as the XL-20 + 3.

Other portable electronic typewriters, those without calculating mechanisms, were similarly characterized as “fundamentally distinct[ ]” from electro-mechanical PETs during the 1983 scope proceeding, to wit:

... [T]he respondents state that electronic typewriters are the product of a different technology, that they are radically different in physical characteristics, that the additional features such as memory capability cannot be duplicated b [sic ] a portable electric machine_ 48 Fed. Reg. at 7,769.

Nevertheless, the physical differences did not lead the ITA to determine that the electronic typewriters were not of the same class or kind of merchandise subject to the 1980 order. 16

Likewise, the record evidence noted by the ITA and elaborated upon by the inter-venor-defendants on the internal circuitry of an automatic typewriter does not support a finding that it, unlike its non-automatic electronic and electro-mechanical counterparts, is not also subject to the order. Word-processing capability notwithstanding, the physical differences between typewriters with text memories and those without do not detract from their fundamental purpose. According to the defendant, the “printing of letters on paper in the manner of a printing press” is the typewriter’s “primary function” 17 , not its editing capabilities or its ability to print characters automatically without manual operation of the keyboard, as the intervenor-defendants claim. See, e.g., S.R.Doc. 13, pp. 40-41; S.R.Doc. 21, pp. 17-19.

Of course, the prominence of one component compared with another can be a determinative factor as to whether the physical characteristics of a product are distinct from those of the original merchandise. Compare, e.g., Color Television Receivers, Except for Video Monitors, from Taiwan, 51 Fed.Reg. 46,895, 46,902 (Dec. 29, 1986), with Television Receivers, Monochrome and Color, from Japan, 52 Fed. Reg. 8,940, 8,946 (March 20, 1987). Those scope determinations indicate, however, that PETs with text memories, whose primary function is admittedly the same as that of PETs without memories, do not differ physically to such an extent as to fall within another class or kind of merchandise.

The defendant nevertheless argues (at page 21 of its memorandum) that the similarity of the primary function

is not sufficient to establish that automatic typewriters are within the scope of the ... order. The reason is that the antidumping order specifically excluded certain typewriters — the automatic typewriters then in existence — that had the very same primary function.

But this argument skirts the fact recited in the remand results and quoted above, page 245, supra, that it was the non-portability of those earlier automatic machines, which were “office” typewriters, that meant their exclusion from the order.

Portability, an electrically-activated keyboard, and an ability to imprint characters like those of a letterpress are common characteristics of PETs. See, e.g., S.R.Doc. 28 at A-2 to A-3. When those characteristics are combined with “pure typewriter mode operation” 18 in a unit designed for easy manipulation of a variety of paper types 19 , they amount to PETs of the class or kind subject to the original order.

Some intervenor-defendants claim that this approach is too general and would encompass “word processors, lap top computers, most calculators”. E.g., Brother Memorandum, p. 57. Those products, however, have other characteristics which distinguish them from PETs. Most obviously, what has become known as a “word processor” displays at least one-half page of text on an electronic screen. See, e.g., S.R. Doc. 13, p. 36. An automatic PET, on the other hand, displays at best some two lines of text. Moreover, such a PET features neither the printing speed nor the storage capacity of a word processor. 20 Simply increasing the complexity of the machine’s microelectronic circuitry does not transform a typewriter into a word processor, and certainly not into a computer. And the more sophisticated semiconductor chip likewise does not create a distinct class or kind in between a portable electronic typewriter and a word processor, as has been suggested. See, e.g., Canon Brief, p. 16. The general physical characteristics remain substantially the same, and a PET with text memory clearly retains its typewriter identity.

B

The ITA takes the position that the ultimate use of an automatic PET, in considering a second Diversified Products criterion, is substantially different from that of a non-automatic PET:

... [A] primary distinguishing characteristic of an automatic typewriter is its capability to print automatically without the direct operation of a keyboard. This function also makes the automatic typewriter more similar to a word processor. 21

The defendant also argues that the “review and revision function” is a distinguishing use of automatics, as is the ability to “remove the page being typed, review the text, and make corrections a day, a week or even a month later”. Defendant’s Memorandum, p. 23. The plaintiff, on the other hand, argues that the review and revision capabilities are so limited as to preclude the use of automatic PETs as substitutes for word processors. See Plaintiff’s Memorandum, pp. 39-43.

The record supports plaintiff’s position. The evidence indicates that the “word processing” capabilities of the automatics are not at high enough “levels of sophistication” 22 to support a finding that machines with text memories are in the same class or kind of merchandise as word processors or personal computers. Only those automatic PETs with larger information storage capacities come anywhere near to performing effectively those features described above as “distinguishing”. 23 The papers submitted make references to PETs with memory storage capacities of 16k, yet the record indicates that the typical model with memory has a lesser capacity, the equivalent of only two or three pages of text. 24 Thus, a memory permits review and revision of text, but only to the extent that

it fits within the few pages the memory allows. The small display screens make seeing what you’re doing a bit hard, as do the often-cumbersome means provided for moving back and forth through the text. You’d probably end up printing a draft version, then entering the corrections. 25

Without an additional, external memory device, the automatic PETs are “insufficient” for preparing “lengthy, multipage documents or requiring] extensive memory to store formats and repetitive forms”. S.R. Doc. 10, Exhibit 4(3). Thus:

... We recommend electronic typewriters to customers who do not need a lot of storage. It makes sense to save the computer for long documents and keep an electronic typewriter for quick notes and letters. 26

Other “major differences” distinguishing word processors from typewriters, as revealed in the record, include “the size of the system, the special functions, the speed of operation of the microprocessors, the quality of the text display, ... and the cost”. Id., Exhibit 2, p. 12, para. 30. What the court gleans from this is that, as a result of these distinctions, a consumer interested primarily in text storage and editing capabilities purchases a dedicated word processor or a personal computer with a word processing program. See, e.g., id., Exhibit 7(9). Specific uses exclusive to those machines, in contrast to typewriters, are described in the supplemented record. For example, in one magazine article a salesperson is advised to recommend the purchase of a dedicated word processor for users producing long documents with complex formats and revisions. For someone tying “into a network, updating] a spread sheet or data base file, and do[ing] word processing tasks, ... sell a microcomputer with word processing and applications software”. Id., Exhibit 11(5).

The record further indicates that, in order to achieve significant word-processing ability with an electronic typewriter, a system of various, additional components— such as a video screen and an external memory device — is required. 27 While there are references to the PETS at issue as “word-processing typewriters” in the papers submitted 28 and in the record, the trade apparently uses that terminology to describe such a system of components added on to a “common typewriter with memory”. Id., Exhibit 7(9). Stated another way, it is “modularity and upgradability” which place typewriters within the range of capabilities of word processors or personal computers. See id., Exhibit 7(7). But without such enhancements, the underlying machine, which is all that is at issue herein, remains just an “ordinary typewriter”. Id.

The uses to which automatic PETs are put are the same as those for “ordinary” PETs. 29 According to papers submitted by Nakajima, the general uses include preparing letters, preparing texts and reports, filling out forms, preparing statistical tables, and preparing lists 30 . That text memory may facilitate the performance of some of those tasks does not establish that automatic typewriters are therefore inter changeable with word processors or personal computers 31 , as the defendant has concluded. See Remand Results, p. 5. Rather, the record reflects that PETs are purchased for their “typewriter functions”, for which other products either cannot be used — or cannot be used as efficiently. See S.R.Doc. 10, Exhibits 7(11) and 11(6). Such functions include “writing short notes, addressing several envelopes or typing a few Rolodex cards, to name a few” 32 for there is

still the difficulty of using the computer for a lot of the day to day typing functions ... [s]imple functions, like typing forms or memos, envelopes. S.R.Doc. 21, Exhibit 22 at 20.

Word processors or personal computers, unlike typewriters, “are not efficient when you have to frequently switch from task to task”. S.R.Doc. 10, Exhibit 11(6) at 46. Thus, they hardly qualify as replacements for electronic typewriters even though arguably more sophisticated. See id,., Exhibit 7(8). Stated another way, in order to accomplish the widest range of tasks, there is a place for both a word processor and a typewriter. See id., Exhibit 11(9) at 58.

Although certain tasks are carried out more efficiently on typewriters with text memories than without, it is not that function which makes the machine, as described by a marketing expert, an “indispensible tool”. See S.R.Doc. 10, Exhibit 7(4) at 70. As a stand-alone product, its word-processing capabilities are simply too limited. Whether automatic or not, typewriters perform “flexible functions that are very difficult for a computer to do”. S.R.Doc. 21, Exhibit 22. In short, the record as a whole does not support the ITA’s finding that the “ultimate use of automatic [PETs] is substantially different from that of typewriters originally investigated”, to quote from the Remand Results, page 5.

C

Similarly, the record does not support a finding that consumers’ ultimate expectations of automatic PETs — a third consideration under Diversified Products — are significantly different from those of PETs without text memories. As evidence to support the ITA’s view that “the consumer’s motive for purchasing an automatic typewriter is different from his motive for purchasing a non-automatic typewriter” 33 , the defendant points to PET advertisements submitted by the parties 34 , which it claims give “prominence” to the “word processing capabilities of typewriters”. See Defendant’s Memorandum, p. 26. What those promotional materials contain in general are pictures of the typewriters and lists of available features, some of which include references to “text memory” or “word processing”. However, even if all of the advertisements focused exclusively on those references 35 , other parts of the record reflect that the underlying reason a consumer purchases an electronic typewriter with text memory is an expectation that it will be able to perform as a typewriter, albeit with some storage and editing capabilities. If a consumer were primarily interested in those capabilities, a typical automatic PET with a few pages’ worth of memory and either no or very little display screen would hardly be satisfactory. While the consumer could enhance such a typewriter with a disk drive, keypad attachment, CRT monitor and word-processing software, the underlying machine would remain a PET. Indeed, the advertisements, sales displays, brochures, packaging and other manufacturers' materials in the record confirm that the merchandise at issue is sold as a typewriter, and not as another product. 36

By way of comparison, as described in the record, the Smith Corona PWP system “is an admirable attempt at turning a typewriter into a word-processor” 37 . It is claimed to be a “complete word-processing system" which includes an “electronic typewriter with spelling dictionary and correction memory”. See, e.g., S.R.Doc. 21, Exhibit 11(19). Such a claim anticipates that the consumer expects the capabilities of a typewriter, as well as of a word processor. However, if typewriter attributes were not of primary interest, there would be little reason to purchase that system, built as it is around an automatic typewriter, notwithstanding that such a system has some capabilities similar to those of a word processor. But dedicated word processors or personal computers are superior products for those consumers primarily interested in text editing and storage capabilities, with limited need for functions that only typewriters can fulfill.

Electronic typewriters with word-processing capabilities that arguably compare with those of dedicated word processors or personal computers are part of modular systems. The record indicates that the Smith Corona PWP is available for $599.00 38 , but, at that price, “you could practically buy an IBM-PC clone, the nucleus of a far more powerful and versatile system” 39 . Hence, the stronger inference is that the system’s more sophisticated features do not induce its purchase. Rather, its capacity to perform typewriter functions is determinative. Otherwise, a consumer presumably would purchase that product which would provide the greater text storage and editing capabilities for the money.

The ITA noted in its Remand Results that automatic PETs “demand a higher retail price than those without text memory” and that the “ability to store text will add to the typewriter’s cost”. Remand Results, p. 5. As stated by intervenor-defend-ant Nakajima, since

buyers are willint [sic] to pay more for the ability to input, store, recall and edit documents ... it is clear that consumers’ expectations as to automatic typewriters are different than their expectations of non-automatics. Nakajima Memorandum, p. 31.

While a selective consideration of cost indicates automatic PETs are priced higher than non-automatics 40 , the record also demonstrates that typewriters with other features, such as “spell check”, also command higher prices than those without them. For example, two Brother models offer identical features, including 5k text memory, except that the AX-20 offers a spelling dictionary, while the AX-15 does not. 41 The former retails for $100.00 more than the latter. 42 Under the ITA’s rationale, the “price premium” for the model with spell check would indicate that the consumer’s motive for purchasing such a typewriter differs from that for purchasing one without the feature 43 rather than suggest that the purchaser is simply willing to pay the price for that added feature.

Other information in the record — affidavits of individuals in the typewriter industry 44 , including retailers and manufacturers 45 , market studies 46 , and customer surveys 47 — likewise does not provide support for a finding that consumers who purchase an automatic PET expect a machine distinct from a non-automatic one. Rather, it indicates that consumers consider text memory an attractive feature among others on a typewriter that is portable:

... [S]ome of the key features on electronic typewriters are features that have been around since the inception of the electronic typewriter. Variable pitch, a variety of typestyles, one line correction. If you look at the features that people are buying and getting excited about, it’s those basic features. 48

In its memorandum, intervenor-defend-ant Brother infers (at page 45) that the consumer is very aware of the available automatic features and “expects a unique and special product” when purchasing a PET with text memory. Its marketing director, on the other hand, indicated that consumers are not necessarily aware of what is available and are impressed by all of the features on today’s PETs. See S.R. Doc. 21, Exhibit 22. Without evidence to support their assertions, several other in-tervenor-defendants state that consumers expect to use PETs with text memories for “multiple reproduction” 49 of documents and “mass mailings of individualized letters”. Nakajima Memorandum, p. 31. As shown in the record, however, use of automatic PETs as printers is not efficient, “taking nearly two to four minutes to print a business letter that an ordinary dot-matrix printer whip[s] out in 32 seconds” 50 . Because of this, as well as the fact that paper must be advanced manually, any expectation of a PET as an efficient office machine or as a machine which “contains the sophistication of a word processor” 51 would be hard to realize.

D

In considering the fourth Diversified Products criterion, the ITA determined that automatic and non-automatic PETs are sold through the same channels of trade, specifically, through discount stores, mass merchandisers, office suppliers and department stores. 52 This determination is not diminished by the fact that other products such as microwave ovens, telephones and word processors are similarly marketed. See, e.g., S.R.Doc. 10, Exhibit 12.

Although the plaintiff argues that more weight should be given to this criterion, the cases cited for the proposition that the ITA’s scope rulings purportedly turned on the channels of trade are not controlling; the agency has discretion in how it chooses to balance the various criteria. See, e.g., Consumer Products Division, SCM Corp. v. Silver Reed America, Inc., 753 F.2d 1033 (Fed.Cir.1985). In any event, the record shows, and the agency has determined, that most of the typewriters at issue, like those admittedly covered by the antidumping-duty order, are sold in different channels of trade than heavy office machines, which are marketed through specialized dealers. See, e.g., Remand Results, p. 6. This — coupled with the fact that most PETS, automatic or non-automatic, are sold for home use 53 — indicates that the significant distinction is between office machines and PETs, not PETs with and without text memories. Gf, e.g., R.Doc. 298, Appendix 25.

E

As indicated above, information was submitted to the ITA with regard to cost differences between automatic and non-automatic PETS, including (1) testimony that the cost of adding to a PET the hardware for 8k text memory is about $2.10 per unit 54 ; (2) market price quotations indicating prices for memory chips ranging from 8k to “32K x 8 NMOS static RAM” 55 ; (3) an affidavit attesting to the minimal cost of adding the memory function 56 ; (4) factory production costs for the Brother non-automatic CE-50 and the automatic CE-58 57 ; (5) a table reflecting “ROM Software Development Man Hours and Costs” for the Brother EP-5 with a calculator, the automatic AX-28, and the non-automatic AX-22 58 ; and (6) an affidavit stating, without reference to source, that adding text memory is “approximately 15% of the total unit cost”, as is the addition of a liquid crystal display. See S.R.Doc. 10, Exhibit 2, para. 39.

Given the number of PETs at issue and their variations, however, the information submitted does not provide a firm basis for evaluating the figures presented. For example, the production-cost table submitted by Brother for its CE-50 and CE-58 models does not indicate the year(s) to which the figures apply, although it can be assumed that the cost of producing the hardware and the software for automatic PETs has decreased over time. In any event, the ITA did not rely on cost in its analysis, but the record does indicate that advances in technology have enabled manufacturers to offer features to consumers that once were available only on more expensive office equipment. And to offer such features on PETs within a price range affordable to larger numbers of consumers can hardly remove those machines from the class or kind of merchandise that has been under review.

F

In summary, the information relied upon by the ITA in its Diversified Products analysis, as well as the arguments presented by the parties, do not amount to substantial evidence in support of the determination that the typewriters at issue are not within the class or kind of merchandise covered by the 1980 antidumping-duty order. While the record does indicate that text memory is another technologically-advanced feature of a continually evolving product, that feature does not make an automatic PET something other than “a machine with a manually operated keyboard which produces characters like those of a letterpress as a substitute for handwriting” and which is “customarily sold at retail with a carrying case”, to refer to the International Trade Commission’s definition of a PET for its investigation.

While recognizing that the “possibility of drawing two inconsistent conclusions from the evidence does not prevent an administrative agency’s finding from being supported by substantial evidence” 59 , the court is not persuaded that that is the case here. Though voluminous, the overall record simply does not provide the support for the critical factors underlying 60 the challenged determination.

IV

In view of the foregoing, the determination of the ITA after remand that portable electric typewriters incorporating calculators are within the scope of the Antidump-ing Duty Order on Portable Electric Typewriters from Japan, 45 Fed.Reg. 30,-618 (May 8, 1980), is supported by substantial evidence on the record and is in accordance with law, but the determination of the agency after remand that portable electric typewriters with text memories are not within the scope of the aforesaid antidump-ing-duty order is not supported by substantial evidence on the record, and that determination is therefore hereby reversed and the ITA is directed to issue a redetermination that portable electric typewriters with text memories are within the scope of the aforesaid antidumping-duty order.

So ordered.

1 . Dated March 18, 1988 and referred to hereinafter as "Remand Results".

2 . Motion of Smith Corona Corporation for Judgment on the Agency Record as Supplemented During Remand, pp. 2-3.

3 . See Matsushita Memorandum of Law in Opposition to Plaintiff's Motion for Judgment on the Agency Record as Supplemented During Remand [hereinafter cited as "Panasonic Memorandum”], pp. 10-13. See abo Memorandum of Nakajima All Co. Ltd. in Opposition to Smith Corona Corporation’s Motion for Judgment on the Agency Record as Supplemented During Remand [hereinafter cited as "Nakajima Memorandum”], p. 3, n. 2.

4 . Brief of Defendant-Intervenors Canon Inc. and Canon U.S.A., Inc. in Opposition to Smith Corona’s Motion for Judgment on the Agency Record as Supplemented During Remand [hereinafter cited as “Canon Brief”], p. 27.

5 . Compare 48 Fed.Reg. at 7,769 with 6 CIT at 162, 572 F.Supp. at 889, which court opinion also refers to cost as a fifth factor. See Kyowa Gas Chemical Industry Co. v. United States, 7 CIT 311, 312 (1984) [available on WESTLAW, 1984 WL 3729].

6 . 11 CIT at -, 678 F.Supp. at 289.

7 . Defendant's Response to Plaintiff's Motion for Judgment on the Agency Record, p. 5.

8 . Supplemental Record Document ("S.R.Doc.”) 13, pp. 10-12 (citations and footnotes omitted).

9 . See Plaintiff's Memorandum of Points and Authorities In Support of Motion of Smith Corona Corporation for Judgment on the Agency Record as Supplemented During Remand [hereinafter cited as "Plaintiffs Memorandum"], pp. 19-21.

10 . Defendant's Memorandum in Partial Opposition to Motion of Plaintiff, Smith Corona Corporation, for Judgment on the Agency Record as Supplemented During Remand [hereinafter cited as "Defendant’s Memorandum”], p. 18, quoting from the Remand Results, p. 4.

11 . Brother Memorandum, p. 35. See, e.g., Panasonic Memorandum, pp. 7-8. Several parties submitted diagrams of the internal circuitry of text-memory and non-text-memory typewriters. See Record Document (“R.Doc.”) 245, Exhibit 9; S.R.Doc. 14, Exhibit 2; S.R.Doc. 27, Exhibit 1.

12 . Canon Brief, p. 13.

13 . S.R.Doc. 13, p. 30, n. 26.

14 . Id. at p. 14.

15 . Compare S.R.Doc. 14, Exhibit 3 (brochure for XL-20 + 3 with photograph) with Memorandum in Support of Smith Corona’s Motion for Judgment on the Agency Record, Appendix 4 (advertisement for EP-20).

16 . In its memorandum at page 19, the defendant implies that the plaintiff misplaces its reliance on the 1983 determination (by the ITA) that electronic machines are within the scope of the order covering electro-mechanical ones despite their physical differences. Claiming that that determination was not based on an analysis of the Diversified Products criteria, the defendant impliedly distinguishes it from the scope determination now under review on the ground that consideration of such criteria herein is “necessary".

As in the instant scope ruling, however, the 1983 determination clearly states that the ITA considered the physical characteristics of the electronic PETs. See 48 Fed.Reg. at 7,769. The necessity of that analysis was never addressed, and there is no indication that that issue affected the outcome.

17 .Defendant’s Memorandum, p. 21.

18 . S.R.Doc. 10, Exhibit 7(6) at 74.

19 . See id., Exhibit 11(5) at 60.

20 . See, e.g., S.R.Doc. 10, Exhibit 5(1); S.R.Doc. 14, Appendix 4; S.R.Doc. 27, Exhibit 7; Supplemental Record Confidential Document ("S.ConfDoc.”) 5, Exhibit 7.

21 . Remand Results, p. 5. See also Nakajima Memorandum, p. 24, citing S.R.Doc. 10, pp. 19-20, Exhibit 2, paras. 14, 18, 25, 26 and Exhibits 4, 7 and 13.

22 . S.R.Doc. 10, Exhibit 2 at 12.

23 . See, e.g., S.R.Doc. 10, Exhibit 5(1); S.R.Doc. 14, Appendix 4; S.R.Doc. 27, Exhibit 7; and S.ConfDoc. 5, Exhibit 7, at 687.

24 . See generally id. at 687, 689 and 690.

25 . Id. at 687. See also S.R.Doc. 10, Exhibit 7(4) ("the one- or two-line display alone is a major limitation when compared with a video screen”) and Exhibit 7(7) ("A one line or two line display screen is not going to serve you with anything near the value of the unit”).

26 . S.R.Doc. 10, Exhibit 11(6) (Computer Merchandising, Electronic Typewriters: Where Do they Fit?, p. 45 (July 1-14, 1985)).

27 . See, e.g., S.R.Doc. 10, Exhibit 5(1); S.R.Doc. 14, Appendix 4; S.R.Doc. 27, Exhibit 7; S.ConfDoc. 5, Exhibit 7.

28 . See, e.g., Nakajima Memorandum, p. 17.

29 . This does not hold true for all automatic typewriters, however. As the record reflects, those in existence at the time of the original antidumping investigation were heavy, expensive office machines in which the automatic features were devoted to the performance of repetitive tasks and the composition of large documents. See S.R.Doc. 27, Exhibits 10-12; Plaintiff’s Memorandum, Appendix 6. Such uses are in contrast to those of portable automatics.

Similar analysis was relied on in the ITA's determination that PETs with calculators are within the scope of the 1980 order. See Remand Results, p. 7.

30 .See S.R.Doc. 10, Exhibit 2 at 16, para. 40. Compare S.R.Doc. 21, Exhibit 22 at 19 ("the average person probably doesn’t have much of a need for databases or spreadsheets").

31 . See, e.g., S.R.Doc. 10, Exhibit 5(1); S.R.Doc. 14, Appendix 4; S.R.Doc. 27, Exhibit 7; S.ConfDoc. 5, Exhibit 7, at 685.

32 . S.R.Doc. 10, Exhibit 11(6).

33 . Remand Results, p. 5.

34 . See S.R.Doc. 10, Exhibit 11; S.R.Doc. 14, Exhibit 4; S.R.Doc. 21, Exhibit 17.

35 . In general, the advertisements simply do not focus on text-memory, although they often point to it along with other attributes, on occasion in bold print.

36 . Cf. Gold Star Co. v. United States, 12 CIT -, -, 692 F.Supp. 1382, 1385 (1988) ("The present merchandise is sold on the U.S. market not as a [printed circuit board] nor as a [color picture tube], but as a color television receiver; the object of the original antidumping duty order”) [emphasis in original].

37 . S.R.Doc. 10, Exhibit 5(1); S.R.Doc. 14, Appendix 4; S.R.Doc. 27, Exhibit 7; S.ConfDoc. 5, Exhibit 7, at 685.

38 . See, e.g., S.R.Doc. 2, Exhibit 17.

39 . S.R.Doc. 10, Exhibit 5(1); S.R.Doc. 14, Appendix 4; S.R.Doc. 27, Exhibit 7; and S.ConfDoc. 5, Exhibit 7, at 689.

40 . The defendant characterized as “hardly useful” plaintiffs comparison of the prices of two of its own non-automatic PETs with two automatic PETs manufactured by Brother. See Defendant’s Memorandum, p. 27.

According to the plaintiff, the Brother machines retailed for less than the plaintiffs despite the presence of text memory. See Plaintiffs Memorandum, p. 35. Contrary to defendant’s characterization, a comparison of prices of all models that are available to the consumer regardless of manufacturer would reflect more accurately consumer expectations. S.R.Doc. 10, Exhibit 5(1), for example, provides exactly such a comparison among typewriters produced by a broad range of manufacturers, and the Remand Results indicate (at page 5) that the ITA considered the comparison useful.

41 . See S.R.Doc. 10, Exhibit 5(1); S.R.Doc. 14, Appendix 4; S.R.Doc. 27, Exhibit 7; and S.ConfDoc. 5, Exhibit 7, at 690-91.

42 . See id.

43 . The court notes in passing that that feature is equally as prominent as text memory in materials in the record.

44 . See S.R.Doc. 10, Exhibit 2; S.R.Doc. 12; S.R. Doc. 21, Exhibit 21. With regard to consumer expectations, the affidavits contain statements to the effect that purchasers of automatic PETs expect them to perform sophisticated tasks like word processers or personal computers. Other than their personal experiences in the industry, however, the affiants do not disclose grounds for their viewpoints.

45 . See S.ConfDoc. 5, Exhibit 4A, which contains standardized affidavits of owners, executives and other personnel of outlets of PETs. Exhibit 4B contains additional statements.

46 . See S.ConfDoc. 5, Exhibit 9B, which is a market study performed by the plaintiff in 1984 on the basis of a mail survey in which consumers were asked what typewriter feature they desired most.

47 . See S.ConfDoc. 5, Exhibits 9A and 9C.

48 . S.R.Doc. 21, Exhibit 22 at 15. Such a viewpoint refutes, of course, the position of inter-venor-defendant Nakajima that text memory is a "primary function ... widely recognized by consumers”. Nakajima Brief, p. 28. That is, “there are still people out there who can’t distinguish between electric and electronic” typewriters. S.R.Doc. 21, Exhibit 22 at 15.

49 . Panasonic Memorandum, p. 9.

50 . S.R.Doc. 10, Exhibit 5(1); S.R.Doc. 14, Appendix 4; S.R.Doc. 27, Exhibit 7; and S.ConfDoc. 5, Exhibit 7, at 687. See also S.R. Doc. 10, Exhibits 11(6) and 7(4), at 72.

51 . Nakajima Memorandum, p. 27.

52 . See Remand Results, p. 6; S.ConfDoc. 5, Exhibits 2 and 4; S.R.Doc. 21, Exhibit 17; S.R.Doc. 28 at 11 and A-7.

53 . See, e.g., S.R.Doc. 21, Exhibit 22 at 16.

54 . See R.Doc. 343, pp. 83-86.

55 . See S.ConfDoc. 5, Exhibit 3A.

56 . See id., Exhibit 3B.

57 . See S.R.Doc. 13, pp. 18-19; Brother Memorandum, p. 50.

58 .See id. at 17. The plaintiff contends that the figures given should be allocated over the total number of typewriters produced to reflect unit cost. By its estimate, that cost is about $1.00. See S.R.Doc. 27, pp. 13-14.

59 . Consolo v. Federal Maritime Commission, 383 U.S. 607, 620, 86 S.Ct. 1018, 1026, 16 L.Ed.2d 131 (1966).

60 . Shell Oil Co. v. Federal Energy Regulatory Commission, 707 F.2d 230, 235 (5th Cir.1983).