# Tapered Roller Bearings, Four Inches or Less in Diameter, and Components Thereof, From Japan

> Briefs, arguments, decisions, and more.

URL: https://www.frixlaw.com/law-library/documents/fr%3A94-27783

## Record

- **Collection:** Federal Register
- **Document type:** Uncategorized Document
- **Published:** November 10, 1994

## Text

DEPARTMENT OF COMMERCE
[A-588-054]

Tapered Roller Bearings, Four Inches or Less in Diameter, and
Components Thereof, From Japan

AGENCY: International Trade Administration/Import Administration,
Department of Commerce.

ACTION: Notice of Final Results and Partial Termination of Antidumping
Duty Administrative Reviews.

-----------------------------------------------------------------------

SUMMARY: On December 30, 1993, the Department of Commerce published in
the Federal Register the preliminary results of administrative reviews
of the antidumping finding on tapered roller bearings, four inches or
less in outside diameter, and components thereof, from Japan. The
reviews cover 14 manufacturers/exporters of the subject merchandise to
the United States generally during the periods April 1, 1979 through
July 31, 1980; August 1, 1980 through July 31, 1981; August 1, 1981
through July 31, 1982; August 1, 1982, through July 31, 1983; August 1,
1983 through July 31, 1984; August 1, 1984 through July 31, 1985; and
August 1, 1985 through July 31, 1986. Based on our analysis of comments
received, the dumping margins for some companies have changed from the
margins contained in the preliminary results.

EFFECTIVE DATE: November 10, 1994.

FOR FURTHER INFORMATION CONTACT: Maureen McPhillips (Koyo), Chip Hayes
(NSK), Valerie Turoscy (Toyota), Lisa Raisner (Yamaha, Nissan, Mazda,
MC International, Nachi-Fujikoshi, Niigata Converter, Toyosha, Suzuki,
Maekawa, Sumitomo Yale, Sumitomo Corp., Mitsubishi), Chip Hayes,
Charles Vannatta, or John Kugelman, Office of Antidumping Compliance,
International Trade Administration, U.S. Department of Commerce,
Washington, D.C. 20230, telephone: (202) 482-5253.

SUPPLEMENTARY INFORMATION:

Background

On December 30, 1993, the Department of Commerce (the Department)
published in the Federal Register (58 FR 69336) the preliminary results
of administrative reviews of the antidumping finding on tapered roller
bearings (TRBs), four inches or less in outside diameter, and
components thereof, from Japan (41 FR 34374, August 18, 1976). We have
now completed these reviews in accordance with section 751 of the
Tariff Act of 1930, as amended (the Tariff Act).

Scope of the Reviews

Imports covered by these reviews are TRBs, four inches or less in
outside diameter when assembled, including inner race or cone
assemblies and outer races or cups, sold either as a unit or
separately. During the review periods such merchandise was classifiable
under item numbers 680.3932, 680.3934, and 680.3938 of the Tariff
Schedules of the United States Annotated (TSUSA). This merchandise is
currently classified under the Harmonized Tariff Schedule (HTS) item
numbers 8708.70.6060, 8708.99, and 8482.99.4510. These TSUSA and HTS
item numbers are provided for convenience and Customs purposes only.
The written description remains dispositive.
These reviews cover TRB sales by Koyo Seiko Company, Ltd. (Koyo),
for the periods 1979 through 1986, NSK Ltd. (formerly Nippon Seiko,
K.K.) (NSK), for the periods 1980 through 1986, Mitsubishi Corp.
(Mitsubishi), for the periods 1980 through 1985, Sumitomo Yale Co.,
Ltd. (Sumitomo Yale), for the periods 1980 through 1985, and Sumitomo
Corporation (Sumitomo Corp.), Nachi-Fujikoshi, Niigata Converter,
Toyosha, Toyota, Yamaha, Suzuki, and Maekawa Bearing Manufacturer
(Maekawa), for the periods 1985 through 1986. Nachi-Fujikoshi and
Niigata Converter claimed no shipments during the 1985-86 period of
review (POR). Consequently, for both firms we have used each firm's
rate from the last prior period in which they had shipments. Finally,
as stated in the preliminary results, for these final results we are
terminating the reviews of Nissan, Mazda (formerly Toyo Kogyo Co.
Ltd.), and MC International.

Analysis of Comments Received

We gave interested parties an opportunity to comment on the
preliminary results. We received written comments from petitioner, the
Timken Company (Timken), and respondents NSK, Koyo, Mazda, and Sumitomo
Corp. Rebuttal briefs were submitted by Timken, NSK, and Koyo. At the
request of Timken, NSK, Koyo, and Mazda, we held a hearing on February
18, 1994.
Comments are addressed in the following order:
1. General Issues
2. Terminations
3. Annual Average Foreign Market Value (FMV), Model Match, and Cost
Test Methodology
4. Calculation of FMV
5. Calculation of U.S. Price (USP)
6. Cost of Production (COP) and Constructed Value (CV)
7. Use of Best Information Available (BIA)
8. Clerical Errors

Comments Regarding General Issues

Comment 1: Timken proposes that assessment of duties should be
based on the actual value of the entries suspended by the U.S. Customs
Service, as Timken states was done in Antifriction Bearings from
France, et al., 58 FR 39729 (July 26, 1993) (AFBs III), and was upheld
for the first antifriction bearing (AFB) administrative review by the
Court of International Trade (CIT) in Koyo Seiko Co., Ltd. v. United
States, 796 F. Supp. 1526, 1529, modified, 806 F.Supp. 1008 (1992)
(Koyo Seiko). Timken acknowledges that, although the Department's
approach in the AFB cases was necessary due to the Department's
sampling methodology, compelling reasons exist in these reviews to
adopt the same approach.
Further, Timken cites Timken Company v. United States, Slip Op. 94-
1 at 11-12 (January 3, 1994) (Timken III), wherein the CIT did not
order a remand for further investigation, but required the Department
to order liquidation of the merchandise in accordance with the AFBs III
methodology (i.e., collect the full amount of antidumping duties
calculated to be due in this case by assessing these duties over all
suspended entries of the merchandise). Timken submits that this is the
only reasonable method of assessing antidumping duties in this long-
delayed proceeding.
Timken argues that assessing antidumping duties in these reviews
may be problematic because ``it is likely'' that entries subject to the
dumping finding have been incorrectly liquidated. Similarly, Timken
notes that the administrative record contains evidence that an importer
incorrectly entered certain merchandise as outside the scope of the
antidumping orders on TRBs, referring to U.S. Customs Service, HQ
222367 (December 28, 1990) (Customs HQ), originally submitted as
Exhibit B of Koyo's October 30, 1991 submission.
Timken acknowledges that identification of the merchandise is
complicated by the fact that the subject merchandise was often entered
as part of a multi-product shipment and points out that, unless the
Department adopts an alternative assessment methodology, the correct
amount of duty may not be assessed at liquidation. Timken suggests that
before the Department issues final assessment instructions, it should
require Customs to report the value of suspended entries during the
review periods for each manufacturer subject to review. With this
information, Timken argues, the Department could then formulate
assessment instructions that would ensure collection of the actual
amount of duty determined to be payable by each company.
Koyo believes the Department should reject Timken's claim for the
following reasons:
Timken has provided no evidence to support its claim that
entries subject to the dumping finding have been liquidated in error.
Timken's cite to the CIT decision in Koyo Seiko as support
for assessment on the basis of entered value is misleading because the
use of sampling rendered assessment on entered value necessary in the
AFB cases.
Section 751(a)(2) of the Tariff Act requires that the
excess of FMV over USP ``shall be the basis for the assessment of
antidumping duties on entries of the merchandise included within the
determination and for deposits of estimated duties.''
Timken's cite to Timken III is not valid because in that
case Timken challenged the Department's failure to address adequately
Timken's arguments regarding discrepancies between the value of U.S.
sales reported by respondents and the value of suspended entries
reported by the Customs Service. The CIT concluded that the Department
had fulfilled its statutory obligation and that Timken should have
brought an action in mandamus to compel Customs to perform its duty.
Moreover, Koyo argues that the CIT's reference to assessment based on
entered value was obiter dicta, as the Department had not instructed
the Customs Service to assess antidumping duties based on entered value
in the review at issue in that appeal.
Furthermore, Koyo states that the Department cannot
legally order the assessment of duties on entries which have already
been liquidated. Koyo states that liquidation of entries is final
unless the liquidation is properly protested in a timely manner, citing
19 U.S.C. Sec. 1514(a) and United States v. Utex Intern, Inc., 857 F.2d
1408, 1410-11 (Fed. Cir. 1988).
NSK argues that assessment of antidumping duties must be imposed on
an entry-by-entry basis only on unliquidated entries. NSK asserts that
the Department has no statutory authority to assess total antidumping
duties (covering liquidated and unliquidated entries) over unliquidated
entries.
Department's Position: As outlined in the Department's Advanced
Notice of Proposed Rule Making (56 FR 63696 (December 5, 1991)), with
respect to exporter's sale price (ESP) assessment, the Department has
experienced significant difficulties in issuing liquidation
instructions to Customs in cases involving large volumes of ESP
entries. Several factors contribute to this difficulty: (1) the lag
between entry date and sale date in ESP transactions, (2) the
fungibility of the product, (3) the volume of transactions reviewed,
and (4) the parties' own inability to link entries to sales or to
provide the Department with the documentation to do so. Given these
difficulties, as noted in the Advanced Notice of Proposed Rule Making,
the Department has explored various alternative methods of issuing ESP
assessment instructions. Such alternatives, and variations, include:
(1) When entry data are available, the Department can derive a per-
unit margin by dividing the calculated dumping duties due (based on
sales) by the number of units entered and then instruct Customs to
apply the per-unit margin against all units entered during the review
period. Similarly, the Department can derive a percentage margin by
dividing the calculated dumping duties due (based on sales) by the
entered value of the merchandise entered during the POR and then
instruct Customs to apply this percentage margin against the entered
Customs value of merchandise entered during the POR. This proposed
methodology simplifies and streamlines assessment and liquidation
dramatically, yet still enables Customs to collect the exact amount of
dumping duties due on ESP sales that occurred during the review periods
(although the lag between entry and sale dates in an ESP situation will
still make assessment problematic).
(2) When entry data are not available, we can derive a percentage
margin by dividing the calculated dumping duties due (based on sales)
by the net USP (which is our best calculation of the entered value of
the merchandise entered during the POR), and then instruct Customs to
apply that percentage margin against the Customs value of suspended
entries of merchandise entered during the POR.
(3) The Department can derive a per-unit dumping margin by dividing
the calculated dumping duties on ESP sales that occurred during the POR
by the number of units sold and then apply the per-unit margin against
all units entered during the POR.
We have evaluated these options in the context of these reviews
covering entries from 1979 through 1986. We have determined that
absolute assessment (option 1), as advocated by Timken, is not feasible
for these reviews. Collection of the entered value of suspended entries
of covered merchandise requires an amount of precision on a country-
wide scale that, though possible for current ongoing reviews given
today's level of electronic database maintenance, is clearly not
feasible for the time periods in question in these reviews. Also, other
than a limited number of Yamaha entries and the single instance cited
by Timken (Customs HQ), we have no evidence that any entries of covered
TRBs were incorrectly liquidated. Timken's concern about possible
incorrectly liquidated entries is also unfounded conjecture and not
based on substantial evidence. We disagree with Timken that these
concerns constitute compelling reasons to use the AFBs III methodology.
As for Timken's concern that TRBs entered as a part of a multi-product
shipment may escape coverage of the finding, there is no evidence on
the record to support this conclusion. Further, this option is not
feasible because we did not ask the respondents for the entered value
data necessary to implement this approach, and do not believe it
appropriate to do so now, given the age of the reviews.
Since option 1 is not feasible, we have considered options 2 and 3.
In order to be consistent with the Department's practice outlined in
the second administrative review of the AFB orders (Antifriction
Bearings from the Federal Republic of Germany (57 FR 28395, June 24,
1992) (AFBs II), we will calculate POR-specific appraisement rates
using option 2, and we will instruct Customs to assess dumping duties
on the entered value of suspended entries of covered TRBs during each
POR.
Comment 2: Timken notes that cash deposits were not required for
Koyo and NSK until the completion of the first administrative review in
June 1990. Timken maintains that the Department, in each of the
administrative results of review issued to date in this case, has
refused to assess interest on antidumping duties secured by bonds
rather than cash deposits. Timken acknowledges that the CIT has upheld
the Department's determination not to assess interest and notes that
Timken has appealed those decisions to the Court of Appeals for the
Federal Circuit (Federal Circuit) (Timken Company v. United States, 15
CIT 526, 777 F.Supp. 20 (1991), aff'd after remand, 819 F.Supp. 1093
(CIT March 4, 1993), appeal docketed, Fed. Cir. No. 93-1312 (April 16,
1993)).
Timken asserts that section 778(a) of the Tariff Act contemplates
collection of interest on all ``underpayments of amounts deposited on
merchandise entered'' and that the Department's rationale for
distinguishing entries secured by bonds, as opposed to cash deposits,
distorts the statutory purpose and benefits those respondents that
failed to make cash deposits of estimated duties. Timken requests that
the Department address this issue once again in these final results.
NSK argues that the Tariff Act, the Department's practice, and
judicial precedent hold that no interest is assessable on entries
covered by bonds rather than cash deposits.
Koyo counters that Timken raises no new arguments in this review in
support of its position and that unless and until the Department's
longstanding interpretation of the Tariff Act is rejected by the
Federal Circuit, the Department should continue to instruct Customs not
to assess interest on antidumping duties secured by bond when
liquidating the entries subject to these reviews.
Department's Position: We disagree with Timken. Section 778(a) of
the Tariff Act provides that interest shall be payable only on
overpayments and underpayments of ``amounts deposited.'' The CIT held
in Timken Co. v. United States, 777 F.Supp. 20 (1991), that the words
``amounts deposited'' refer only to cash deposits of estimated
antidumping duties upon entry, and not to other kinds of security, such
as a bond. In considering Timken's appeal, the Federal Circuit held
that ``the requirement to make cash deposits of estimated duties, under
the duty order, triggers the interest provision. Without the duty
order, the importer has no obligation to pay interest. The Court of
International Trade did not err in upholding ITA's determination that
NSK and Koyo Seiko are not liable for interest * * *'' (Timken Co. v.
United States, Fed. Cir. No. 93-1312, -1455 (Fed. Cir., September 27,
1994)).
This proceeding concerns a 1976 finding, in which the statute
required bonds, not cash deposits. The Trade Agreements Act of 1979
(the 1979 Act) did not allow the collection (or payment) of interest
where the entry of merchandise was permitted under bond rather than by
the cash deposit of estimated dumping duties. Further, the 1979 Act
contained no provision for the conversion from bonds to cash deposits
for existing antidumping findings, except through the administrative
review process.
Therefore, because we required cash deposits for the first time on
entries of merchandise manufactured by Koyo and NSK on June 1, 1990,
interest will only be collected or refunded on under- or overpayments
of cash deposits on entries after that date.
Comment 3: Timken states that in its submission of October 8, 1991,
it requested that the Department convert existing bonds, posted by Koyo
and NSK as security for antidumping duties, to cash deposits. Since the
Department has never formally responded to its request, Timken is
renewing that request in its case brief for these administrative
reviews.
Timken maintains that the role of Customs in matters pertaining to
antidumping duties is purely ``ministerial,'' requiring Customs to
implement the instructions of the Department to the extent those
instructions are not inconsistent with Customs regulations, citing
Diversified Products Corp. v. United States, 6 CIT 155, 572 F.Supp. 883
(1983). According to Timken, the Department, therefore, has the
authority to require Customs to collect cash deposits at this time,
even if such deposits are nominal in amount. Timken argues that the
conversion of existing bonds to cash deposits would permit the
Department to collect interest on under-deposits of antidumping duties
as required by section 778(a) of the Tariff Act. Timken contends that
the Department should recognize that this proceeding is sui generis, at
least with respect to Koyo and NSK, since the Department published a
decision that cash deposits would be imposed on entries subject to
existing findings as of the date of completion of the first
administrative review (45 FR 1084, January 4, 1980). Timken points out
that Koyo and NSK posted bonds for all of their entries during each of
the PORs.
Moreover, Timken notes that the Department's ruling with respect to
interest on Koyo and NSK entries was issued in June 1990 and is still
pending before the Federal Circuit. Based on this experience, Timken
states that it may take four years or more before assessment occurs. In
order to safeguard the revenue during the inevitable appeals of the
final results of these reviews, Timken urges the Department to require
conversion of bonds to cash deposits.
NSK argues that the Department has consistently followed its policy
of not requiring cash deposits on entries covered by pre-1980 findings
until the first administrative review has been completed under section
751 of the Tariff Act. NSK contends that Timken has cited no authority
to suggest that the Department's practice is contrary to law.
Koyo counters that neither the Department nor the Customs Service
possesses any authority to require the posting of cash deposits
retroactively. Koyo remarks, citing U.S. Customs Service, HQ 222367
(December 28, 1990), that the Customs Service has no authority to
require posting of cash deposits on entries that have been entered and
accepted.
Department's Position: In 1980, the Department of the Treasury
published a notice (45 FR 1084, 1980) stating that the 1979 Act
contained no provision for the immediate conversion of existing
antidumping findings from bonds to cash. Section 106(a) of the 1979 Act
merely provides that outstanding antidumping findings would remain in
effect and would be subject to review under section 751 of the Tariff
Act. Section 751(a)(2) of the Tariff Act provides that the
administrative review ``shall be the basis for * * * deposits of
estimated duties.'' Thus, the legal basis for requiring cash deposits
under the 1979 Act is a review under section 751 of the Tariff Act or
an order pursuant to section 736 of the Tariff Act. Because this case
is not governed by an order, the Department is without power to require
the imposition of a cash deposit until the completion of an
administrative review.
Under U.S. law, where an affirmative dumping finding has been made,
security must be provided upon entry of the merchandise into the
country, and if the form of security is cash deposits, it must be paid
no later than 30 days after entry (19 U.S.C. section 1505). Because the
merchandise in question has already entered the country, the form of
security cannot be changed retroactively. Similarly, since interest is
only collectible on cash deposits pursuant to sections 737 and 778 of
the Tariff Act, interest would accrue only from the time that the
Department required a cash deposit and would not be retroactive to the
time of entry under bond. The Department has followed this practice
consistently and will continue to follow it for merchandise already
entered into the United States.
Comment 4: Timken urges the Department to release detailed computer
printouts of the final results of review as the limited information
released in the preliminary results is insufficient to permit the
parties to identify all errors in the computer printout. Since the
Department has consistently refused to release ``pre-final''
calculations for review by the parties, Timken contends that the final
output often provides the only indication of clerical errors in the
Department's analysis. Therefore, at a minimum, Timken believes the
Department should release a representative sample of major datasets
(i.e., 500 observations for each annual period).
NSK agrees with Timken that the Department should release detailed
computer printouts of its final analysis. NSK also suggests that the
Department consider releasing the printouts in advance of issuance of
the final results, in order to expeditiously identify and correct any
clerical errors.
Koyo supports Timken's request for pre-disclosure of the final
results in this case due to the very complex nature of the preliminary
margin analysis and the number of clerical errors identified in the
preliminary results computer program.
Department's Position: We disagree with Koyo and NSK. The issuance
of pre-final programs and printouts would only serve to delay the final
results of these reviews. We believe that Koyo has misinterpreted
Timken's comments regarding release of the ``pre-final'' calculations.
In fact, Timken states that the Department's past consistency in
refusing to release pre-final results makes a detailed release of the
final margin program even more important. We agree with Timken in
principle, and intend to provide the parties to this proceeding with a
representative sample of the major datasets needed to identify any
clerical errors in the computer programs, if they request disclosure
after issuance of the final results to identify and comment on any
clerical errors.

Comments Regarding Terminations

Comment 5: Mazda states its concern over the Department's refusal
to terminate the review of Sumitomo Corp. with respect to Mazda
transactions. Mazda notes that in Timken's agreement to withdraw its
request to review Mazda, Timken consented to the termination of the
review with regard to resale transactions to Mazda's subsidiaries
through Sumitomo Corp. Mazda states its interest in this proceeding is
to retain its zero margin rate for all U.S. entries of Mazda
replacement-part TRBs, without regard to whether such TRBs are imported
directly from Mazda or indirectly from Mazda through resales by
Sumitomo Corp.
Mazda suggests three alternative proposals: (1) partial termination
of the review of those sales through Sumitomo which both originate and
end with Mazda; (2) termination of the review with regard to all sales
for export to the United States by Mazda, whether directly or
indirectly through Sumitomo Corp.; or (3) termination of the review
with regard to all sales for export to the United States by Mazda and
instructions to Customs to liquidate all TRBs sold by Mazda for export
to the United States, whether directly or indirectly through Sumitomo
Corp. Mazda contends that any of the three alternatives would leave
Sumitomo Corp. in the review while clarifying the Department's intent
to have all of Mazda's export sales of TRBs to the United States,
whether directly or indirectly through Sumitomo Corp. as a reshipper,
treated as exports by Mazda. Mazda maintains that such treatment is
consistent with both Mazda's and Timken's request to terminate the
review, as well as consistent with current and past treatment by
Customs of Mazda's exports to the United States.
Sumitomo Corp. maintains that it is not the importer of record, and
therefore has no liability for antidumping duties. Further, by
terminating the review of Mazda, the Department will effectively
terminate the review of Sumitomo Corp. with respect to Mazda's exports
during the POR.
Timken reaffirms its agreement to the termination of the 1985/86
review with respect to Mazda and, of the three alternatives proposed by
Mazda, prefers the third, namely, the termination of the review with
regard to all sales for export to the United States by Mazda, with a
clarification to Customs of the Department's intent that Mazda's
exporter rate should be applied to all TRBs sold by Mazda for export to
the United States, whether directly or indirectly through Sumitomo
Corp.
Department's Position: We agree with Timken that alternative three
offers the best solution with regard to termination of the review of
Mazda and Mazda transactions through Sumitomo Corp. It satisfies
Mazda's request and Timken's affirmation to terminate the review with
regard to all sales for export to the United States by Mazda, while
enabling us to appraise exports by Sumitomo Corp. for sale to parties
other than Mazda. Accordingly, in these final results, we are
terminating the review of Mazda with regard to all sales for export to
the United States by Mazda. In our instructions to Customs we will
clarify our intent that Mazda's rate should be applied to all TRBs sold
by Mazda for export to the United States, whether directly or
indirectly through any entity. For an explanation of our treatment of
Sumitomo Corp., please see the following comment.
Comment 6: Mazda argues that the Department erroneously treated
Sumitomo Corp. (also known as Sumitomo Shoji Kaisha, Ltd.) and Sumitomo
Yale Co., Ltd. as one company, when in fact they are two separate,
independent firms. Mazda also notes that only Sumitomo Yale was
included in the 1980/85 reviews (51 FR 24883) and only Sumitomo Corp.
was included in the 1985/86 review (51 FR 32817).
Sumitomo Corp. states that it should not have been included in
these reviews and should no longer be considered a party to this
proceeding because Sumitomo Corp. does not manufacture, import, or
further process TRBs. Sumitomo Corp.'s role, for which it receives a
fixed commission, in the sale and shipment of TRBs, has only been to
arrange for the transport of the merchandise from Japan to the United
States in transactions between related parties.
Because Sumitomo Corp. did not respond to the Department's
questionnaire, Timken argues that, for exports by Sumitomo Corp. for
sale to parties other than Mazda, the Department should apply BIA to
Sumitomo Corp.'s entries both for appraisement and to derive the cash
deposit rate for future entries.
Department's Position: Evidence on the record indicates that
Sumitomo Yale and Sumitomo Corp. are, in fact, two separate firms.
Evidence on the record also indicates, however, that we sent Sumitomo
Yale questionnaires for the 1980/85 periods to which the firm declined
to respond. Therefore, in accordance with section 776(c) of the Tariff
Act and the methodology outlined in the notice of preliminary results,
we have used first-tier BIA to determine the margins for Sumitomo Yale.
Although the Department initiated a review of Sumitomo Corp. only
for the 1985/86 POR, we inadvertently sent Sumitomo Corp. a
questionnaire for several periods preceding that POR. In addition, we
cannot confirm that we sent a questionnaire to Sumitomo Corp. for the
1985/86 review. Therefore, it is inappropriate to use BIA, since
Sumitomo Corp. did not fail to respond or otherwise impede the
proceeding.
We have considered several additional facts: (1) the age of the
proceeding does not allow for efficient retrieval from Customs of
information on suspended entries; and (2) Timken has withdrawn its
request for a review of Sumitomo Corp.'s shipments with respect to
Mazda for the 1985/86 POR. Therefore, we have accepted the portion of
Sumitomo Corp.'s case brief with comments regarding its activities for
the 1985/86 POR which provides evidence that Sumitomo Corp. did not
sell TRBs to the United States during the 1985/86 POR.
Therefore, for entries suspended as ``Sumitomo Corp.'' entries for
the 1985/86 POR, we will notify Customs to liquidate them at the
manufacturer's rate, either based on the final results of this review
or as provided for by 19 CFR 353.22(e). Specifically, in accordance
with 19 CFR 353.22(e), Mazda entries suspended as ``Sumitomo Corp.''
entries will be liquidated at the amount of cash deposit required at
the time of entry. We further note that we are currently conducting a
review of Sumitomo Corp. for the 1992/93 POR, and we will determine in
that review the role of Sumitomo Corp. in U.S. TRB transactions based
on the information on the record for that review.
For entries by all firms for which we are terminating these
reviews, in accordance with 19 CFR 353.22, we will instruct Customs to
assess antidumping duties in the amount of the cash deposit required at
time of entry. Existing cash deposit rates will remain in effect for
those firms until the next publication of final results of review.

Comments Regarding Annual Average FMV, Model Match, and Cost Test
Methodology

Comment 7: Both NSK and Timken argue that the Department's
calculations incorrectly excluded home-market models which lacked
physical criteria information from the model-match portion of the
analysis. NSK and Timken contend that, where possible, the Department
should match such home-market models with U.S. models with identical
nomenclature.
Department's Position: We agree. In the preliminary results of
review we did not intend to exclude from the model-match analysis any
home-market models lacking physical criteria information. We have made
the necessary correction to our calculations for these final results of
review to ensure that we consider all home-market models in determining
whether they are identical to models respondents sold in the United
States.
Comment 8: Koyo states that for the final results the Department
should follow the CIT's direction in Koyo Seiko Co., Ltd. and Koyo
Corporation of U.S.A. v. United States, 834 F.Supp. 431 (CIT 1993). In
that case, the CIT concluded that a 10-percent cap was required to
limit the permissible deviation of the criteria used to match TRB
models and agreed with Koyo that the use of a 10-percent cap on the
comparison of each physical criterion is necessary to avoid
``comparisons between products which differ so dramatically that they
simply cannot be considered commercially similar'' (Koyo Seiko, Co.,
834 F.Supp. 435).
Timken disagrees with Koyo, stating that the ``sum of the
deviations'' method of determining comparison merchandise, in
conjunction with the 20-percent difference-in-merchandise (difmer)
limit, implements the intent of section 771(16) of the Tariff Act that
the comparison merchandise be ``like'' the merchandise sold in the
United States, and of approximately equal commercial value. Section
771(16) of the Tariff Act, Timken argues, does not require that home-
market models be technically substitutable. Therefore, in Timken's
opinion, because this issue is currently under judicial review, the
Department should decline to modify its methodology, pending a
``final'' judicial decision on the issue.
Department's Position: We agree with Timken. As we stated in
Tapered Roller Bearings, Four Inches or Less in Outside Diameter, and
Certain Components Thereof, from Japan (58 FR 64721, December 9, 1993)
(1990/91 and 1991/92 TRBs), we are satisfied that use of the original
sum of the deviations methodology, without the 10-percent cap but
including the 20-percent difmer cap, accurately determines the most
similar model sold in the home market. We have, therefore, used the
original sum of the deviations methodology for model-match comparisons
in these final results of review.
Because we use the original sum of the deviations model-match
methodology, rather than the single greatest deviation methodology we
used in conjunction with the 10-percent deviation cap in the less-than-
fair-value (LTFV) investigation of TRBs over 4 inches, application of a
10-percent deviation cap to each physical criterion would mean that the
best overall match could be eliminated simply because a single physical
criterion deviated by more than 10 percent. By using the sum of the
deviations methodology, all five matching criteria have equal weight.
By contrast, the application of a 10-percent cap to each physical
criterion represents a distortion of the model-match methodology, in
that it establishes a randomly-selected hierarchy, rather than
evaluating each criterion on an equal basis. The 10-percent cap can
result in our inability to match some U.S. sales of TRBs which the
Department considers most similar overall, but which have a greater
than 10-percent deviation in only one criterion. This results in
reliance on CV to a much greater degree when the statutory preference
is for price comparisons.
Furthermore, if the Department were to apply both the 10-percent
cap and the 20-percent difmer cap, the methodology would become too
restrictive, since in some cases the only matches passing the 20-
percent difmer test may vary by more than 10 percent in one or more
physical criteria. Using both tests might eliminate matches of
otherwise comparable merchandise. Thus, this methodology would result
not in more precise matches but in fewer matches to such or similar
merchandise, and an overreliance on CV.
Finally, although the CIT held in NTN Bearing Corp. v. United
States, Slip Op. 94-25 (February 11, 1994), that the Department must
apply the 10-percent cap to the sum of the deviations methodology, the
Department respectfully disagrees with this decision. Because
application of the 10-percent cap in that case does not alter the
margins, the Department cannot appeal that case (see Zenith Electronic
Corp. v. United States, 895 F.2d 291 (Fed. Cir. 1989)). However, we
have appealed the next case involving this issue (Koyo Seiko Co., Ltd
and Koyo Corporation of U.S.A. v. United States, Slip Op. 93-185).
Therefore, for all of these reasons, and for the reasons explained in
the notices of prior TRB final results of review, we have not changed
our methodology for these reviews.
Comment 9: Koyo objects to the Department's use of set-splitting to
create artificial home-market sales of cups and cones for comparison to
sales of cups and cones sold separately in the United States. Koyo
states that there are sufficient actual sales of cups and cones in the
home market to allow the Department to avoid comparison with CV. In
addition, the use of annual average FMVs in these reviews permits the
Department to match a U.S. sale of a cup or cone to a home-market sale
of a such or similar cup or cone made at any time in the home market,
rather than within the six-month window the Department uses with
monthly weighted-average FMVs. Koyo asserts that this fabrication of
artificial sales by set-splitting violates the statute, skews the
margins, and creates unnecessary work and expense.
Koyo cites section 773(a)(1) of the Tariff Act as providing that
``[i]n the ascertainment of foreign market value, for the purpose of
this title, no pretended sale or offer for sale, and no sale or offer
for sale intended to establish a fictitious market, shall be taken into
account.'' Koyo contends that there may have been a ``sale'', but not
of the type of product to which it was matched in the U.S. market.
Acknowledging that the CIT upheld the Department's splitting of
sets in NTN Bearing Corp. of America v. United States, 747 F.Supp. 726
(CIT 1990)(NTN), Koyo contends that the facts of that LTFV
investigation are different from the facts of these reviews. Koyo
explains that in the LTFV investigation, to avoid over-reliance on CV,
the Department split home-market sales of TRBs into ``sales'' of cups
and cones. Koyo contends, however, that in these reviews there were
sufficient actual home-market sales of cups and cones to match with the
U.S. sales of cups and cones. In addition, Koyo states that reliance on
actual sales would simplify the calculations and lessen the
administrative burden. If the Department determines that it is
necessary to continue to split TRB sets, Koyo requests that the
Department at least combine the sales of cups and cones in order to
conduct a fair analysis, citing Timken Co. v. United States, 673
F.Supp. 495, 505 (CIT 1987) (Timken II).
Timken asserts that a TRB set is nothing more than a cup or cone
sold together as a unit. The practice of selling cups and cones as
individual components in some markets does not render the components
distinct articles of commerce. Timken challenges Koyo's contention that
cups and cones have different commercial values than complete sets,
when, in fact, Koyo points to no evidence in the administrative record
to support its contention.
Timken identifies several court cases where the CIT upheld the
Department's set-splitting. For example, in Timken II, Timken notes
that the CIT rejected NTN's argument that the Department was creating
pretended sales of cups and cones in the home market. Timken also cites
NTN as support for the Department's set-splitting methodology.
Moreover, Timken notes that the CIT has upheld the Department's use
of set-splitting in the context of using home-market annual average
prices in NTN Bearing Corp. of America v. United States, 835 F.Supp.
646 (CIT 1993)(NTN I). Although this issue was not specifically
addressed in this case, Timken notes that the Department's consistent
practice has been to split TRB sets, no matter how FMV was determined
(e.g., 1990/91 and 1991/92 TRBs, 58 FR 64720).
Finally, Timken sees no merit in Koyo's argument that the
Department and respondents would be spared the additional work
necessary to split home-market set sales into cups and cones.
Timken concludes that Koyo has provided no reason for the
Department to modify its present approach, and urges the Department to
continue to split home-market TRB sets into individual cup and cone
components for the final results of these reviews.
Department's Position: We agree with Timken. In its most recent
decision on this matter, the CIT again reaffirmed the Department's
practice of set splitting (NTN I). In Timken II the CIT pointed out
that these split sales are not ``fictitious'' sales, but real sales
made to real customers. The CIT upheld the Department's decision to
split sales of sets because, otherwise, respondents could have forced
the Department to use CV in its analysis by simply selling sets in one
market and cups and cones in the other: ``the Court declines to read
section 1677b(a)(1) to permit such control by foreign manufacturers of
the manner in which FMV is determined'' (Timken II, at 495, 504-505).
In addition, the CIT has also stated in NTN I that, if NTN's
interpretation of the statute were followed, ``such interpretations
would encourage importers to circumvent the antidumping laws by simply
using divergent invoicing methods'' (NTN I, 726, 741). The Department
considers set-splitting to be necessary for these reviews; cups and
cones split from sets are potentially the most similar merchandise to
the products sold in the United States. Because they may be the most
similar products, it is appropriate to include this merchandise in the
pool of home-market sales.

Comments Regarding Calculation of FMV

Comment 10: Timken argues that the adjustment to FMV for early-
payment discounts as direct selling expenses, which the Department
granted NSK in the preliminary results of the 1984/85 and 1985/86
reviews, is unsupported. Timken contends that the discounts were not
related to specific merchandise and notes that, under identical
circumstances, the Department has in the past denied a respondent's
claim for a direct adjustment.
NSK concedes that, using its present methodology, the Department
requires discounts to be reported on a transaction-specific basis.
However, NSK argues that in this case the Department's 1986
questionnaire did not require NSK to report early-payment discounts on
a transaction-specific basis, reflecting the Department's practice at
the time. Furthermore, NSK notes the company was unable to tie its cash
discounts to specific merchandise because the same discounts were
granted on all merchandise. NSK explains that its allocation of total
discounts over total sales was consistent with its record-keeping at
the time.
Department's Position: The Department's current practice is indeed
to require that discounts be reported on a transaction-specific basis
(see AFBs III, 39729, 39759). However, in our original questionnaires
for these reviews we did not require NSK to report discounts on such a
basis, and we did not issue supplementary instructions when our
practice changed. Therefore, because NSK's submitted discount
information represents the only information on the record, we have
continued to rely on these data for these final results of review.
Comment 11: Timken contends that the Department incorrectly treated
Koyo's post-sale price adjustments (PSPAs) and rebates as direct
selling expenses in the home market when, in fact, Koyo reported these
adjustments as aggregate amounts, attributable to all products, and,
therefore, did not report them on a transaction-specific basis. Timken
states that at verification the Department determined that Koyo
calculated (a) a lump-sum discount amount, and (b) a part-number-
specific debit/credit PSPA, and then reported a PSPA ratio based on the
aggregate of these two values. Timken notes that 19 CFR 353.56(a)
requires that billing adjustments bear ``a direct relationship to the
sales compared'' in order to be treated as a direct selling expense. In
addition, Timken notes that, according to several court decisions, the
burden of proving entitlement to adjustments rests with respondents.
Further, Timken cites Torrington Company v. United States, 818 F.Supp.
1563, 1579 (1993) (Torrington), as evidence that the CIT disallowed a
methodology which would include PSPAs and rebates on out-of-scope
merchandise in calculating an FMV.
Timken, however, believes that the Department should continue to
adjust for home-market price increases, by using Koyo's ``positive''
price adjustments, as these data constitute the best information on the
record concerning price increases during the periods of review.
Koyo states that the Department correctly accepted Koyo's home-
market billing adjustments as direct selling expenses. Koyo argues that
Timken's reliance on Torrington is misplaced because, although the CIT
stated that the Department should not include billing adjustments for
out-of-scope merchandise, the CIT proposed an alternative methodology
whereby billing adjustments for in-scope merchandise can be calculated
by identifying the ratio of in-scope merchandise to total sales to
which these billing adjustments apply, and applying that ratio to total
billing adjustments (Torrington, 1578-79). Koyo maintains that the
Department has verified and accepted Koyo's methodology in previously
completed TRB reviews (1990/91 and 1991/92 TRBs; Tapered Roller
Bearings and Parts Thereof, Finished or Unfinished, from Japan, 57 FR
4951 (February 11, 1992))(TRBs I).
Koyo concludes that the Department properly treated Koyo's home-
market PSPAs as direct selling expenses. Koyo adds that if the
Department contemplates rejecting Koyo's billing adjustments, it must,
at a minimum, re-open the administrative record to allow Koyo to
resubmit information that conforms to any new standard the Department
may use in lieu of the standard prevailing in 1991 when Koyo submitted
its consolidated response.
Department's Position: We disagree in part with Timken and Koyo.
For these review periods, Koyo pooled all debit and credit notes
associated with the PSPAs and rebates of each customer and allocated
the amounts over the total purchases (i.e., in-scope and out-of-scope
merchandise) of each customer. In our supplemental questionnaire we
asked Koyo to demonstrate how each price adjustment was linked to each
transaction. Koyo only provided a computer printout showing price
adjustments by customer code. The price adjustments were divided by the
customer's total purchases during each POR. Since Koyo's records during
the PORs did not reveal the exact nature of the price adjustments, we
have not treated these expenses as direct selling expenses. Nor have we
limited the adjustments to only positive values, as Timken suggests,
because the use of only positive values would distort the information
submitted by Koyo. Therefore, we have classified Koyo's PSPAs and
rebates as indirect, rather than direct, expenses.
Comment 12: Koyo states that the CIT held in Koyo Seiko Co., Ltd.
v. United States, 810 F.Supp. 1287, 1292 (CIT 1993) (Koyo Seiko), and
in numerous other cases that the antidumping statute requires the
Department to make a circumstance-of-sale (COS) adjustment to FMV for
U.S. direct selling expenses. Koyo cites as examples NSK Ltd. v. United
States, No. 92-03-00158, Slip Op. 93-216 (CIT Nov. 18, 1993), and NTN
Bearing Corp. of America v. United States, No. 91-08-00576, Slip Op.
93-56 (CIT April 21, 1993)). Given such clear and consistent
instruction from the CIT, Koyo argues that the Department, for these
final results, must adjust for U.S. direct selling expenses by adding
these expenses to FMV, rather than by subtracting them from USP.
NSK argues that the Department's treatment of U.S. direct selling
expenses as adjustments to USP was contrary to law and judicial
precedent. NSK argues that U.S. direct selling expenses are properly
treated as upward adjustments to FMV.
Timken counters that the law requires Customs to collect the full
amount of the difference between USP and FMV, and if an improper
denominator is used in calculating the ad valorem appraisement rate,
respondents will escape collection of the full amount of duties. Timken
asserts that in order for Customs to apply appraisement rates to the
entries subject to these reviews, the Department should calculate the
percentage rates on the basis of entered value. If this is done, Timken
claims there will be no difference in the rates calculated regardless
of whether the expenses are subtracted from USP or added to FMV.
Moreover, Timken notes that the Department has declined to follow
the decisions cited by NSK and Koyo (1990/91 and 1991/92 TRBs), on the
grounds that an appeal of the underlying issue is pending in other TRB
litigation. Therefore, Timken states that the Department should
continue to deduct direct selling expenses from USP rather than add
them to FMV.
Department's Position: It is our longstanding practice, pursuant to
section 772(e)(2) of the Tariff Act, to deduct all expenses incurred in
the United States, including direct selling expenses, in calculating
ESP. In calculating purchase price (PP), adjustments for differences in
COS pursuant to 19 C.F.R. Sec. 353.56, including direct selling
expenses, are made to FMV, and no deduction of direct selling expenses
is made from PP. This is necessary to avoid a systematic distortion in
the amounts of duty assessed which would result if the value on which
dumping margins were calculated were consistently different than the
entered value, to which Customs will apply the margin.
Entered value is most commonly based on the price to the United
States between the exporter and the importer. In contrast, the basis of
ESP is the resale price in the United States to the first unrelated
purchaser, which will approximate the entered value only after all
expenses incurred in the United States (including direct selling
expenses) are deducted. PP will approximate the entered value without
the deduction of any expenses because the direct selling expenses are
incurred in the exporting country and included in the price to the
United States used for both PP and entered value.
On September 30, 1994, the Federal Circuit reversed the decisions
of the CIT in Koyo Seiko Co., Ltd. and Koyo Corporation of U.S.A. and
Isuzu Motors, Ltd. and American Isuzu Motors, Inc., v. United States
and The Timken Company (93-1525, 93-1534), holding that ``(n)othing in
the plain language * * * of the Antidumping Act precludes Commerce's
approach of adjusting exporter's sales price by deducting therefrom
certain direct selling expenses incurred in the United States. Indeed,
Commerce's stated rationale for its approach is well within the bounds
of reasonableness.'' Therefore, we have maintained our practice for
these final results.
Comment 13: Timken maintains that Koyo cannot consolidate its sales
to a related distributor and then report the selling expenses of that
distributor (e.g., indirect selling expenses, credit), when it reported
the sale between Koyo and its related distributor and did not report
the distributor's re-sale prices to unrelated parties. In Timken's
view, Koyo's ad hoc treatment of related distributors' expenses and
sales information, which were consolidated for some purposes and not
for others, renders the entire indirect selling expense claim invalid.
Koyo states that Timken misinterpreted the verification report,
which makes clear that Koyo reported only selling expenses incurred by
Koyo at its headquarters and sales offices, and did not include any
expenses incurred by related companies.
Department's Position: We disagree with Timken. We have reexamined
the documents cited by Koyo, including the home-market verification
report, and have concluded that the information on the record regarding
Koyo's indirect selling expenses is accurate. Although we have adjusted
Koyo's home-market indirect selling expenses as a result of other
positions taken by the Department for these final results, we are
satisfied that Koyo submitted only those expenses it incurred at its
headquarters and sales offices.
Comment 14: Timken maintains that in the preliminary results of
review for Yamaha, the Department erroneously deducted an imputed
adjustment for home-market credit expense from FMV. Because Yamaha made
no claim for this adjustment, Timken asserts that the Department is
under no obligation to adjust for this expense, and that for the final
results the Department should not make the credit expense adjustment to
FMV.
Department's Position: We agree. The CIT has stated that the burden
of proving entitlement to adjustments rests with the respondent making
the claim (Timken II). Therefore, we have not made an adjustment to FMV
for Yamaha's imputed credit expenses.

Comments Regarding Calculation of USP

Comment 15: Timken asserts that section 772 of the Tariff Act
requires the Department to adjust USP for ``all costs, charges, and
expenses . . . incident to bringing the merchandise from the place of
shipment in the country of exportation to the place of delivery in the
United States. . . .'' Timken states that this includes pre-sale inland
freight. In contrast, Timken observes that there is no similar
provision for deduction of such expenses from FMV. Therefore, Timken
argues that the statute calls for distinct treatment of pre-sale inland
freight, depending upon whether USP or FMV is at issue.
Timken notes that the Federal Circuit's rejection of the
Department's practice of treating home-market pre-sale inland freight
as a direct deduction from FMV is categorical and not limited to the PP
context (Ad Hoc Committee of AZ-NM-TX-FL Producers of Gray Portland
Cement v. United States, Slip Op. 93-1239 (Fed. Cir., January 5, 1994)
(Ad Hoc Committee)). Timken asserts that the Department's
implementation of Ad Hoc Committee requires that any adjustment for
pre-sale inland freight be made pursuant to the ESP offset (as an
indirect selling expense), rather than pursuant to the COS provision of
section 773 of the Tariff Act.
Koyo points out that the Department treated pre-sale inland freight
as an indirect expense, which is contrary to the Department's
established practice in previous reviews of this case of treating pre-
sale inland freight as a direct selling expense. Koyo cites 1990/91 and
1991/92 TRBs.
Koyo states that in Ad Hoc Committee, which involved PP sales, the
Federal Circuit held that the Department may not make a deduction for
pre-sale inland freight in calculating FMV. Koyo notes, moreover, that
the Federal Circuit's decision explains that the case before it was a
departure from the Department's longstanding practice of deducting pre-
sale transportation expenses from FMV in ESP situations.
Therefore, Koyo maintains that the Department should continue its
practice of deducting pre-sale inland freight expenses as direct
selling expenses from FMV. Koyo also asserts that Timken's reliance on
Ad Hoc Committee is misplaced because the decision applied only to
situations involving PP transactions and does not apply to Koyo's
transactions in these reviews where USP is calculated using ESP.
NSK argues that the Ad Hoc Committee decision has no bearing on
NSK's sales. NSK contends that the Federal Circuit decision involved PP
sales, in which foreign inland freight was post-sale freight, whereas
NSK's sales are ESP sales, in which all inland freight is pre-sale
freight. NSK argues that, in an ESP analysis, the deduction of pre-sale
freight from USP but not from home-market price would result in an
unequal comparison.
Department's Position: The Ad Hoc Committee decision states that
the statute does not give the Department the authority to deduct home-
market movement expenses from FMV by invoking its inherent power to
fill in ``gaps'' in the antidumping statute (Ad Hoc Committee, Slip.
Op. 93-1239). Consistent with the rationale of this decision, the
Department applies this methodology to both PP and ESP situations. As a
result, we must now evaluate a claim for pre-sale inland freight within
the context of the COS adjustment provision of the Tariff Act. With
respect to Koyo's and NSK's claims of having incurred pre-sale freight
expenses, we find that neither claim is directly related to any
specific sale and no specific sale is anticipated at the time when
respondents incurred the pre-sale inland freight expense. Consistent
with our practice regarding non-sales-specific expenses, we have
treated pre-sale inland freight for NSK and Koyo as a home-market
indirect selling expense (19 CFR 353.56(b)(2)).
Comment 16: Timken notes that in the preliminary results the
Department excluded NSK's sales with prices of zero from the margin
analysis. Timken contends that these sales are likely sample or
promotional sales, and should be included in the margin calculation as
required by both the statute and past Department practice.
NSK argues that transactions showing zero prices are not ``sales''
because sales require the payment of money, and are thus more properly
viewed as selling expenses. NSK notes that, as verified by the
Department, sample sales are included in NSK's reported SG&A expenses,
and thus should not be included in the Department's analysis of sales.
Department's Position: We agree with Timken. Our general practice
has been to include zero-priced sample U.S. sales in margin
calculations. In AFBs II, where our policy is discussed at length, we
stated,

Sample sales fall outside the scope of the review when the
respondent can demonstrate that no transfer of ownership has
occurred between the exporter and the unrelated U.S. purchaser . . .
the statute and the regulations require the Department to analyze
all sales within the period of review. . . . Consequently, all U.S.
zero-price sales have been included for the final margin
calculation.

In Granular PTFE from Japan we excluded certain U.S. sample sales
from our analysis. In that case, however, sample goods were provided to
customers for testing. Because of the nature of the product, once
tested, the sample could not be returned. Although a transfer of
ownership had occurred, the product had not been used for commercial
consumption, and thus could not be said to have been ``sold'' (see
Granular Polytetrafluoroethylene Resin from Japan; Final Results of
Review, 58 FR 50345 (September 27, 1993)). In this case NSK states that
its sample sales involve transfer of ownership and makes no claim that
the samples are destroyed or rendered unusable, as in Granular PTFE
from Japan. Accordingly, we have included all U.S. sample sales in our
analysis for these final results of review.
Comment 17: Timken argues that NSK's selling, general, and
administrative expenses (SG&A) for U.S. sales are underreported. Timken
notes that NSK excluded SG&A expenses which it claimed to be related to
NSK's U.S. subsidiary's manufacturing operations. Timken argues that
NSK has not provided evidence to support such a claim, and that the
Department, in the 1974/80 reviews, rejected a similar claim. Timken
contends that the Department should recalculate NSK's SG&A expenses to
include the ``manufacturing-related'' expenses.
NSK argues that expenses specifically related to U.S. manufacturing
should not be included in SG&A expenses. NSK maintains that, in the
administrative reviews of the AFB orders, the Department recognized
that a portion of NSK's SG&A expenses are properly related to U.S.
manufacturing. NSK contends that it would be unreasonable to deny NSK a
similar allocation in these reviews.
Department's Position: We agree with Timken. NSK has failed to
support its claimed allocation in either these or previous segments of
this proceeding. At verification we noted that while total SG&A
expenses attributable to manufacturing-related expenses had been
verified, ``NSK did not provide evidence to substantiate the amounts to
be allocated to manufacturing only'' (see Verification Report, March
27, 1987, p. 10). Therefore, for these final results of review we have
recalculated NSK's SG&A expenses to be deducted from USP to include the
``manufacturing-related'' expenses.
Comment 18: Timken notes that NSK reported an export inspection fee
applicable to exports to the United States in 1985/86, and Timken
presumes that similar expenses were also incurred for other periods.
Timken contends that the Department has not adjusted USP for these
expenses in the 1980/85 reviews.
NSK agrees with Timken's contention.
Department's Position: In the preliminary results we did not adjust
USP for export inspection fees. We have made the necessary changes to
our calculations for these final results of review.
Comment 19: Timken argues that for 1980 through 1982, NSK claims
that it had no loans in the United States, and challenges the ``average
lending rate'' methodology NSK used to determine credit expenses and
inventory carrying costs. Timken claims that the ``average lending
rate'' used by NSK is the rate charged on intra-company loans. Timken
argues that the Department should recalculate the U.S. interest rate
based on the formula used by NSK for the other periods (i.e., the U.S.
prime lending rate plus a given factor).
NSK argues that its reported ``average lending rate'' was in fact
the average interest received on short-term deposits, and thus reflects
NSK's true credit costs. NSK argues that, as verified by the
Department, NSK had no loans for the period 1980 through 1982, and,
therefore, its cost of credit is properly the opportunity cost of late
payment, i.e., unrealized interest from short-term deposits.
Department's Position: We agree with NSK. We verified that NSK had
no loans for the period 1980 through 1982 and agree that NSK's credit
costs are the opportunity costs of late payment, and therefore,
properly captured using the average interest rates on short-term
deposits.
Comment 20: Timken maintains that Koyo's conclusion that 50 percent
of its inland freight costs are direct and 50 percent are indirect is
an unsupported estimate of pre-sale and post-sale inland freight
expense and, therefore, should be rejected in its entirety, or, at a
minimum, all of Koyo's inland freight expenses should be considered as
indirect selling expenses.
Koyo counters that there is no evidence that its reported freight
expenses are inaccurate or unreliable. Koyo states that the Department
verified the accuracy of its freight expense amount, and that the
Department's practice of treating pre-sale and post-sale inland freight
as direct selling expenses renders the amount of the allocation between
the two types of expenses immaterial.
Department's Position: Although Koyo requested that we deduct pre-
sale inland freight as direct selling expenses from FMV, it allocated
total freight expenses into direct and indirect categories in its
submission. Koyo contends that it did not distinguish post-sale and
pre-sale freight expenses in its records at that time and, therefore,
allocated half of the expenses to warehouse transfers and half to
direct inland freight. We verified the accuracy of Koyo's total freight
expenses and accept Koyo's allocation of these expenses as the best
estimate possible given the lack of specificity of Koyo's records at
the time. Therefore, we have continued to accept Koyo's allocation of
direct and indirect home-market freight expenses in these final
results, treating 50 percent of total freight expenses as pre-sale
freight (an indirect selling expense) and 50 percent as post-sale
freight (a direct selling expense deduction from FMV).
Comment 21: Timken contends that the Department should deduct a
reasonable profit amount from ESP. Timken concedes that the statute is
silent on this question, but states that there is broad international
understanding that reseller profits should be deducted in ESP-type
situations.
NSK argues that the Department has consistently rejected Timken's
argument, and has been sustained by the CIT. Therefore, the
Department's treatment of profit is in accordance with the law.
Koyo believes that Timken's argument should be rejected not only
because there is no statutory basis for deducting profit from ESP, but
Timken's interpretation of the statute would require a fundamental
change in a well-established Department practice. According to Koyo,
section 772(e) of the Tariff Act directs that, for purposes of
calculating ESP, the price shall be reduced by the amount of
``commissions for selling in the United States the particular
merchandise under consideration.'' Koyo asserts that the Department
(and, prior to 1980, the Department of the Treasury), interprets this
provision literally (i.e., that ``commission'' does not include
``profit''), citing Timken II and Timken Co. v. United States, 630
F.Supp. 1327, 17783-44 (CIT 1986). Koyo concludes that there is no
basis upon which to overturn the Department's decision not to deduct
profits from ESP.
Department's Position: We agree with Koyo and NSK. Sections 772(d)
and (e) of the Tariff Act do not include resale profits among the
detailed list of adjustments that the Department is authorized to make
to USP in ESP situations. Thus, there is no provision in the statute by
which we can make the adjustment that Timken requests.
Comment 22: Timken points out that the Department, in several
recent administrative reviews, rejected Koyo's U.S. discount data
because they were aggregated, rather than transaction-specific. Timken
states that aggregated discounts or price adjustments are unreliable
for purposes of antidumping analysis. Timken cites AFBs III, wherein
the Department rejected Koyo's U.S. discounts and sales allowances
because Koyo did not report them on a transaction-specific basis. In
that review, Timken asserts that the Department assigned to Koyo, as
BIA, the highest percentage discount or sales allowance of any U.S.
sale to all sales that received a discount or sales allowance. Timken
believes the Department should be consistent with its approach in AFBs
III.
Moreover, Timken maintains that the Department should not grant any
adjustment for upward PSPAs to USP given that Koyo's claim is based on
averages of price adjustments and discounts on in-scope and out-of-
scope merchandise rather than transaction-specific price adjustments
and discounts on in-scope merchandise.
Koyo contends that the application of BIA for U.S. discounts would
be inappropriate for these reviews because BIA is normally reserved for
situations in which a respondent fails to comply with an information
request from the Department. Koyo cites Olympic Adhesives v. United
States, 899 F.2d 1565, 1572 (Fed. Cir. 1990), in support of its
contention.
In this instance, according to Koyo, the Department never informed
Koyo that the methodology Koyo used in 1991 for reporting U.S.
discounts and sales allowances would no longer satisfy the Department's
new standard in the more recent final results of review the Department
published for TRBs and AFBs in 1993. Thus, Koyo concludes that it would
be extremely unfair for the Department to reject Koyo's claimed U.S.
discounts at this late stage of these long-delayed reviews.
Department's Position: We agree with Timken. In these reviews of
TRBs it has been the consistent practice of the Department to require
that discounts be reported on a transaction-specific basis. As early as
June 6, 1991, in our final results of review for the 1987/88 POR
(Tapered Roller Bearings from Japan; Final Results of Review, 56 FR
26059) (TRBs), we stated that Koyo's methodology of calculating one
discount factor for all bearing sales and assigning it to a group of
customers is not representative of actual TRB sales experience, since
it fails to tie the actual discounts granted to the sales to which they
apply (56 FR 26059). Thus, we did not allow discounts, rebates, or
price adjustments based on broad allocations in that review. Moreover,
in our 1993 supplemental questionnaire for these administrative
reviews, we requested documentation of the discount rates in effect
during each of the PORs. We also asked Koyo to report discounts and
sales allowances on a transaction-specific or, if that were not
possible, on a customer-specific basis. However, since Koyo did not
report its U.S. discounts and sales allowances on a transaction-
specific basis for the reviews in question, we assigned, as BIA, the
highest percentage discount reported for any U.S. sale to all sales
that received a discount during each POR.
Comment 23: Timken contends that Koyo's ESP offset cap is inflated
due to the Department's inclusion of freight-out expense and
commissions which the Department should re-classify as direct selling
expenses, not indirect. Timken points out that Koyo reported its U.S.
sales commissions on a transaction-by-transaction basis as direct
expenses. Similarly, Timken states that there is no basis for including
freight-out in indirect selling expenses since freight, in general, is
considered a movement charge when calculating USP.
Koyo agrees that the Department should define the ESP offset cap in
the same manner as in its other recent TRB reviews.
Department's Position: We agree with Timken that freight incurred
in the United States should be treated as a direct deduction from USP
pursuant to 19 CFR 353.41(d) and have done so for these final results.
However, the Department rejected Koyo's claim for a deduction for home-
market commissions because Koyo was unable to tie the total commission
amount to sales of covered merchandise at verification. We have,
therefore, included U.S. commissions in the ESP cap, pursuant to 19 CFR
353.56(b).
Comment 24: Timken objects to Koyo's reporting of U.S. freight
expenses based on sales value. Timken notes that the Department
requested that Koyo report its freight-in expense based on the cost of
goods sold, but Koyo indicated that it was unable to comply. Timken
cites TRBs to assert that the Department's consistent position is that
``allocation of freight costs by volume, weight, distance, or a
combination of these, is preferable to allocations based on sales
value'' (TRBs 56 FR, 41508, Comment 17). Timken requests that the
Department re-allocate freight expenses based on the weight of the
merchandise, since the net weight of individual bearings is available
in the administrative record of these reviews.
Koyo claims that its method of allocating freight expenses over the
sales value of the subject merchandise remains the most accurate method
available to account for these expenses. Koyo asserts that the
Department accepted this method as reasonable in prior reviews and
should continue to do so for these final results.
Department's Position: We agree with Koyo. While we have stated, as
Timken notes, that allocations of freight costs by volume, weight,
distance, or a combination of these elements are preferable to
allocations based on sales value, we have also recognized that
individual bearing firms do not maintain records of freight expenses
based on weight (see TRBs). Therefore, we determined that Koyo's
allocation of freight expense based on value is a reasonable method and
does not produce distorted results.
Comment 25: Timken states that the Department is faced with
competing data for export sales expenses reported in May 1991 (revised
November 1993) and export department expenses verified in 1987 for the
period 1974/85. Since both the March 1987 and November 1993
verification reports indicate that total export expenses were
``unsubstantiated'' or an aspect of the expense was ``not verified,''
the Department should use the highest figure on the record for
determining export selling expenses.
Koyo counters that Timken has misconstrued language in the
Department's verification report that Koyo failed verification
regarding its selling expenses. Koyo asserts that the statement that
the breakdown of employees into various categories ``was not verified''
means that the verification team did not address that issue in detail,
and not, as Timken has interpreted it, that Koyo failed verification
regarding its selling expenses.
Department's Position: We agree with Koyo that Timken
misinterpreted the verifiers' explanation of Koyo's home-market
indirect selling expenses, specifically selling expenses incurred in
Japan with respect to U.S. sales. Since export selling expenses were,
indeed, satisfactorily verified in November 1993, we have used the
export selling expenses reported in Exhibit C-16 of Koyo's November 10,
1993, submission.
Comment 26: Timken asserts that Koyo, in its calculation of U.S.
inventory carrying costs, failed to include the imputed interest
expense for time in transit from Japan to its U.S. subsidiary, American
Koyo Company (AKC). Timken cites Silver Reed America, Inc. v. United
States, 12 CIT 250, 683 F.Supp. 1393 (1988), as support for its
position that such an interest expense can be deducted from ESP.
Although Koyo contends that the U.S. inventory period includes time in
transit, Timken maintains that there is no support for that claim in
the administrative record. Timken cites the Department's U.S.
verification report (March 27, 1987), covering the PORs from 1974/85
wherein the Department determined that inventory carrying costs, which
were reported on a similar basis as the present period, did not include
time in transit from Japan.
Timken urges the Department to increase the reported inventory time
by a reasonable time for shipment from Japan to the United States,
using data supplied by another producer, or by assuming at least
thirty-five days in transit.
Koyo states that Timken is incorrect in its assertion that Koyo's
U.S. inventory carrying cost does not include ``time on the water''.
Koyo explains that AKC takes title to merchandise when it leaves Koyo's
warehouse in Japan, and it is booked into AKC's inventory at that time.
Therefore, Koyo contends that AKC's inventory values used to determine
the average time in inventory include the value of the inventory in
transit.
Department's Position: We disagree with Timken. Koyo submitted U.S.
inventory carrying costs in two separate stages, the expenses incurred
in Japan for exported goods and the inventory expenses associated with
U.S. sales. Koyo states that these costs include ``time on the water''.
Because its sales were satisfactorily verified, and we have no reason
to believe that this expense is inaccurate, we have continued to use
Koyo's U.S. inventory carrying costs as reported.

Comments Regarding COP and CV

Comment 27: Timken argues that NSK reported interest expenses for
COP net of interest income, but NSK did not demonstrate that this
income stemmed solely from short-term deposits or normal operations.
Timken contends that the Department should recalculate NSK's COP based
on gross interest expenses, or in the alternative, recalculate the
interest expense factor to exclude long-term interest income based on a
supplementary cost response in which NSK segregated short-term and
long-term interest revenue.
NSK agrees with Timken's contention that the submitted COP figures
incorrectly included an interest expense net of all interest income.
NSK suggests that the Department should recalculate the interest
expense factor based on the supplementary cost response figures.
Department's Position: The interest-expense factor should reflect
only short-term interest income attributable to the normal production
of the merchandise within the scope of the order. We have reviewed
NSK's data for this claimed adjustment and are satisfied that the firm
attributed the reported interest income only to the production of
merchandise within the scope of the order. We did not find evidence on
the record to suggest that some portion of the claimed interest income
is attributable to the production of other merchandise or associated
with long-term deposits.
Therefore, for these final results we have recalculated NSK's
interest expenses based on a NSK's supplementary COP response, wherein
NSK segregated short-term and long-term interest income.
Comment 28: Timken argues that the Department's deduction from CV
for home-market selling expenses is overstated for NSK. Timken contends
that the Department calculated selling expenses by applying reported
factors to CV to determine the amount of selling expenses to deduct
from CV, which already includes selling expenses, rather than applying
those factors to COP.
NSK argues that CV is a proxy for price, and that since the selling
expense factors reported were calculated as a percentage of total
sales, these factors are correctly applied to CV.
Department's Position: We agree with Timken in part. To avoid
overstating the home-market selling expense adjustments, and given that
we have the home-market unit prices, we have applied the reported
selling expense factors, expressed as a percentage of unit price, to
obtain a yen amount for the expense. We then deducted the resulting
amounts from CV.
Comment 29: According to Timken, the Department should exclude
below-cost sales in calculating profit for CV. Timken cites section
773(e) of the Tariff Act to argue that CV includes profit earned on
sales ``in the ordinary course of trade''. Since CV is merely an
alternative basis for determining FMV, and the Department disregards
below-cost sales when sales form the basis of FMV, Timken asserts that
the Department should not include any below-cost sales in its
calculation of profit for CV.
Timken notes that the Department recognized the need for this
balance in Timken II and in Asociacion Colombiana de Exportadores de
Flores v. United States, 13 CIT 25, 704 F.Supp. 1114, 1124, (1989),
aff'd, 901 F.2d 1089, cert. denied, 498 U.S. 848 (1990). Timken points
out that in these cases the question was whether CV is properly subject
to COS adjustments pursuant to 19 CFR 353.56, which provides for
adjustments to home-market price for certain types of selling expenses.
Timken notes that the CIT held that adjustments were proper, given the
functional equivalent of CV and FMV.
Timken states that in Certain Stainless Steel Wire Rods from
France, 58 FR 68865 (December 29, 1993), the Department recognized that
non-arm's-length sales to related parties should not be included in the
profit calculation for CV. Timken concludes that the same reasoning
applies to below-cost sales in the home market and that the statute's
reference to sales in the ordinary course of trade does not include
below-cost sales when those sales are made in substantial quantities
over an extended period of time.
Citing section 773(e) of the Tariff Act, NSK argues that the
statute does not require that sales below cost be excluded in the
calculation of profit of the class or kind of merchandise under
consideration.
Koyo asserts that Timken's arguments run contrary to the statute
and represent unprecedented and radical departure from the Department's
past administrative practice. Koyo notes that the Department has
recently rejected substantially identical arguments made by Timken in
the most recent TRB reviews and the two most recent AFB reviews (1990/
91 and 1991/92 TRBs, 58 FR 64728; AFBs III, 58 FR 39751; AFBs II, 57 FR
28374). Koyo concludes that the Department should reject Timken's
arguments and continue its longstanding practice of including all home-
market sales in the calculation of profit for CV.
Department's Position: As noted by Koyo, the Department has
rejected Timken's arguments that below-cost sales should not be
included in the calculation of profit because (a) the statute does not
explicitly provide that below-cost sales should be disregarded in the
calculation of profit, (b) the definition of ``ordinary course of
trade'' (section 771(15) of the Tariff Act) does not exclude or even
mention sales below cost, and (c) the provision requiring the
Department to disregard certain sales below COP in the calculation of
FMV suggests that below-cost sales are not, per se, outside the
ordinary course of trade. Therefore, in these reviews we have not
excluded below-cost sales in calculating profit for CV (see, e.g., AFBs
III).
Comment 30: Koyo objects to the Department's initiation of a sales-
below-cost investigation on the grounds that the Department has
acknowledged, and the CIT has held, that Timken's below-cost allegation
in 1983 for the 1974 through 1979 reviews formed an inadequate basis
for a below-cost investigation. Because the 1989 allegation for the
1979/86 PORs was based on the 1983 allegation, Koyo asserts that the
later allegation is also inadequate. In addition, Koyo maintains that
both allegations were untimely.
Koyo points to section 773(b) of the Tariff Act, which states that
the Department may initiate a below-cost investigation whenever it has
``reasonable grounds'' to believe sales of subject merchandise in the
home market are being made at prices which represent less than the cost
of producing the merchandise. Koyo maintains that Timken's 1983 and
1989 cost allegations were not company-specific and instead relied
entirely upon references to public sources and, therefore, should be
rejected. Koyo cites Al Tech Specialty Steel Corp. v. United States,
575 F.Supp. 1277 (CIT 1983), in support of its position that COP
allegations must be company-specific. Koyo states that despite the
CIT's decision that Timken's allegations were deficient with respect to
the 1974/79 PORs, Timken never attempted to remedy the deficiencies in
its cost allegations regarding the periods covered by the 1979/86
administrative reviews.
Koyo states that prior to the promulgation of 19 CFR 353.31, which
requires all below-cost allegations to be filed within 120 days of the
initiation of the administrative review, the Department examined the
span of time between when Timken possessed sufficient information to
make the allegation and when the below-cost allegation was actually
made. Koyo cites Color Television Receivers, Except for Video Monitors,
from Taiwan (51 FR 46895, December 29, 1986), which stated that
possessing information 4 months prior to making a COP allegation was
enough time to make the allegation untimely. Koyo states that the
Department relied on these parameters in dismissing two below-cost
allegations, one 10 months after the release of cost-of-manufacture
data, the other 12 months later, submitted by Timken against NSK in the
1974/80 reviews.
In contrast to the NSK situation, Koyo notes that the Department
initiated a below-cost investigation of Koyo for the 1974/79 reviews
based on Timken's 1983 allegation despite the fact that it was based on
information that was available to Timken nearly two years before it
filed the allegation. Koyo notes that Timken's 1989 allegation for the
1979/86 reviews was filed three years after the end of the final period
covered by these reviews, and nine years after the end of the first
period covered by these reviews.
Koyo concludes that the below-cost allegation on which the
Department predicated its cost investigation in the 1979/86 periods of
review was both factually deficient and untimely and, accordingly, the
Department should not investigate whether Koyo sold such or similar
merchandise in the home market at prices below the COP.
Timken disagrees with Koyo's assessment of Timken's 1983 below-cost
allegation for the 1974/79 administrative reviews, stating that Koyo's
argument contains several key errors and omissions, namely its failure
to mention Timken's supplemental COP allegation of April 28, 1992, or
its submission of August 19, 1988.
Timken notes that although the Department initiated a sales-below-
cost investigation for the 1974/79 periods in September 1983, that
investigation was subsequently interpreted to cover only the 1978/79
period, and was to encompass subsequent periods only if the Department
found below-cost sales in the 1978/79 POR. Timken states that when the
Department found that Koyo made below-cost sales during the 1978/79
POR, it issued a COP questionnaire for the 1979/86 PORs in July 1989.
Timken states that, upon completion of the 1974/79 reviews, Koyo
filed a CIT action challenging the Department's decision to conduct a
COP investigation of Koyo. Timken elaborates on the history of that
litigation: 1) the court ruled that Timken's original 1983 submission
did not meet the Al Tech standard for initiation of a cost
investigation and remanded the case for re-calculation of the margin
without a COP analysis (Koyo Seiko Co., Ltd. v. United States, 796
F.Supp. 517 (CIT 1992); 2) on Timken's motion for rehearing, the CIT
modified the remand order, permitting Timken to supplement its COP
allegation without recourse to information obtained during the
Department's initial COP investigation (Koyo Seiko Co., Ltd. v. United
States, 16 CIT 92-139 at 11 (August 21, 1992)); 3) the Department
determined and the CIT affirmed that Timken's supplemental allegation
was a sufficient basis for the below-cost sales investigation (Koyo
Seiko Co., Ltd., et al. v. United States v. United States, 819 F.Supp.
1093 (CIT 1993)).
With regard to the 1979/86 reviews, Timken asserts that it
supplemented its prior allegations for the 1979 through 1986 reviews in
April 1992, based on the COP data Koyo submitted in response to the
July 1989 questionnaire and the variable cost data that would have been
on the administrative record in any event. Timken notes that Koyo
failed to address Timken's April 1992 submission.
Department's Position: If we find sales below cost in the
immediately preceding segment of a proceeding, we have reasonable
grounds to believe or suspect below-cost sales in the subsequent
segment of the proceeding pursuant to section 773(b) of the Tariff Act.
Therefore, we require a response to a cost questionnaire in an
administrative review subsequent to finding sales below cost in a
previous segment of the proceeding. We decided to proceed in the 1979/
86 reviews with a cost investigation since, upon remand in the 1974/79
litigation, the Department conducted a cost investigation for 1978/79
and found sales below cost. The Department's finding of sales below
cost in the 1978/79 review is an adequate foundation to conduct a cost
investigation in the subsequent administrative reviews.
Comment 31: Timken maintains that Koyo's COP information should be
rejected in its entirety as unreliable because Koyo admitted that its
primary records for COP prior to 1984 were destroyed. According to
Timken, the information submitted by Koyo is simply an extrapolation of
data from the 1985/86 POR. Moreover, the data have never been verified.
Timken points out that the Department rejected Koyo's COP data for
the 1978/79 review as inconsistent and unreliable and consequently used
BIA. According to Timken, Koyo has not demonstrated that the cost
information submitted for these reviews is any more accurate or
reliable than the cost data rejected by the Department in the 1978/79
review. Moreover, Timken notes that corporate losses sustained by Koyo
in the late 1970's and early 1980's makes it inherently unreasonable to
extrapolate 1985/86 costs to this earlier period.
Timken identifies the following specific deficiencies in Koyo's COP
submissions:
From 1980-86 Koyo used three separate and irreconcilable
COP systems.
Post-1983 standards and variances were projected backward
to yield costs for the 1979/83 PORs.
``Actual'' and ``reconstructed'' costs for a limited
sample of bearings differed substantially, calling into question the
reliability of Koyo's methodology.
Koyo failed to provide information on transfers of
production equipment from related parties, although Koyo is a major
manufacturer of bearing-production equipment.
Koyo failed to provide part-number-specific data regarding
material usage and prices during the PORs.
The responses contain no detailed information on
calculation of research and development expenses.
Timken asserts that the Department should reject Koyo's COP data
unless it can establish that the newly-submitted data are consistent
with previously-verified information on the record. If the Department
determines to accept the data as submitted, Timken argues that the
Department should make the following modifications:
Timken states that, under the Department's precedent in
AFBs III and Minivans from Japan (57 FR 21933, May 26, 1992), interest
income must be related to production of the subject merchandise to be
allowed as an offset to interest expense. Since Koyo was unable to
report separately long- and short-term interest income for the 1979/83
periods, and instead used net interest expense based on its experience
during the 1983/86 PORs, Timken urges the Department to use the
interest expense contained in Koyo's consolidated financial statements.
Timken maintains that the Department should ensure that
all adjustments to CV are consistent with the price-based adjustments
Koyo reported. If it is not possible to adjust the reported expenses in
this manner, the Department should decline to make any adjustment to CV
for selling expenses.
Timken claims that the Department misconstrued Koyo's
reporting of direct and indirect selling expenses, G&A expenses, and
interest expenses as a fixed amount per unit rather than as a
percentage of the cost of manufacturing (COM). Timken urges the
Department to revise its calculations accordingly.
Koyo reiterates its contention that the Department's initiation of
a below-cost investigation in these reviews was improper (see Comment
30). In response to Timken's arguments Koyo makes the following points:
Timken's assertion that Koyo's costs for the 1979/85
periods are merely an extrapolation of Koyo's 1985/86 costs is
incorrect. Koyo submits that the costs for all the periods covered by
these reviews can be reconciled to total cost of goods sold as reported
in the annual financial statements.
Contrary to Timken's conclusion that essential records
were destroyed, Koyo states that it did retain a complete historical
record of total standard costs, or COM, by bearing model and year.
Koyo argues that it is normal for a firm to change its
cost accounting procedures in the ordinary course of business over a
ten-year period.
Koyo maintains that nothing in its submission suggests
that the ratio of interest expenses for the later period was applied to
the earlier period, as Timken implies.
Koyo did not project post-1983 standards and variances
backwards, but applied standard costs and variances for each review
year to obtain the actual costs for that year.
Contrary to Timken's assertion that Koyo failed to provide
information on transfers of production equipment from related parties,
Koyo maintains that it fully disclosed all consolidated subsidiaries
and subcontractors.
Koyo maintains that differences in how it accounted for various
expenses in different years do not, in themselves, suggest that Koyo's
system is inaccurate or unreliable. Therefore, Koyo concludes that the
Department should continue to use these data if it decides to pursue
the below-cost investigation.
Moreover, Koyo contends that Timken fails to produce new evidence
of any deficiencies in Koyo's cost data, relying only on references to
Koyo's cost submissions in the 1974/79 reviews. Koyo argues that
references to those reviews are irrelevant, and Koyo is under no burden
to draw comparisons between its current submissions and submissions in
previous reviews. Koyo adds that the Department has had Timken's
comments for over two and a half years and has not seen fit to request
additional or clarifying data from Koyo. For the Department to decide
that these data are inadequate would be grossly unfair and prejudicial
to Koyo's interests.
Department's Position: We disagree with Timken's assertion that
Koyo's cost response should be rejected in its entirety. Timken does
not explain why or how Koyo's cost accounting systems are
irreconcilable. Moreover, there is no evidence on the record that Koyo
extrapolated data from the 1985/86 review period and applied a deflator
for previous years included in these reviews. Koyo stated in its
responses to our COP questionnaires that all cost data for these
reviews can be traced to its financial statements. While we did not
verify these specific cost data from the October 1989 and April 1990
submissions, we did verify a significant portion of Koyo's data from
these review periods and found the reported information in the vast
majority of cases to be accurate.
We agree with Koyo that Timken is incorrect in its assertion that
Koyo used net interest expense for the 1979/83 PORs extrapolated from
its experience in the 1983/86 PORs. In its 1990 supplemental response
Koyo explained that it does not separate long- and short-term interest
expense in the normal course of business. However, Koyo stated that it
was able to separate these expenses for the periods 1983/1986. For the
periods 1979/83 Koyo estimated the short-term element based on the
ratio of short-term to long-term liabilities actually paid rather than
accrued each year. Koyo deducted short-term interest income to derive a
net interest expense ratio which was then applied to each model's basic
cost. Therefore, Timken is incorrect in its contention that 1979/83 net
interest expenses were based on Koyo's experience in the 1983/86 review
periods.
We agree, however, with Timken's contention that not all of the
expenses reported by Koyo as direct expenses should be deducted from
CV. For these final results we have not used the direct selling expense
variable as submitted in the CV database because this variable
represented commissions, which we disregarded (see our response to
Comment 23), and rebate expenses, which we did not treat as a direct
adjustment to Koyo's home-market unit prices (see our response to
Comment 11) in these final results. We have used the weighted-average
home-market credit expense of all sales reported in each POR as the
only direct deduction from CV.
We believe that the cost information submitted by Koyo provides the
necessary data to conduct a cost test for each of these periods of
review. Therefore, for these final results of review, we have continued
to use Koyo's cost data as submitted, with certain changes explained in
this notice which we deemed appropriate (see Comments 33, 38, 41, and
46).
As for Koyo's contention that initiation of a below-cost
investigation is improper, see our position on Comment 30.

Comments Regarding Use of BIA

Comment 32: Timken asserts that in all instances BIA for missing
data should be based on adverse inferences. Timken points out that in
Rhone Poulenc Inc. v. United States (899 F.2d 1185, 1190 (Fed. Cir.
1990)) the Federal Circuit upheld the Department's use of the highest
prior margin as BIA. In this instance, however, Timken notes that for
those home-market models lacking variable cost of manufacture (VCOM)
information, the Department set the difference in merchandise (difmer)
to 20 percent, the maximum difference allowed for matching bearings
under the Department's model-match methodology. Timken claims that this
approach infers that the match is otherwise valid, and, in practice,
provides an incentive for respondents to manipulate the model-match
process. Timken concludes that the Department should apply the highest
rate in any previous administrative review to any U.S. sales that are
missing cost data. In Timken's view, such a choice would reflect ``a
common sense inference that the highest prior margin is the most
probative evidence of current margins because, if it were not so, the
[respondent], knowing of the rule, would have produced current
information showing the margin to be less'' (Rhone Poulenc).
NSK argues that at the time that the VCOM data were submitted in
this case the Department did not use a 20-percent difmer as BIA, and,
therefore, respondents had no incentive to selectively report VCOM
information.
Department's Position: We disagree with Timken's contention that
setting the difmer equal to 20 percent when home-market sales lack VCOM
data provides an incentive for respondents to manipulate the model-
match process. We have no reason to believe that such manipulation is
taking place. As a result, we do not agree with Timken that the highest
rate from a previous administrative review should be applied to U.S.
sales which match to home-market sales for which respondents did not
provide VCOM data. Rather, for these final results we have set the
difmer equal to 20 percent. We have used this approach in previous TRB
reviews (see TRBs, 56 FR 26057 (June 6, 1991), TRBs I, 57 FR 4986
(February 11, 1992), 1990/91 and 1991/92 TRBs 58 FR 64731 (December 9,
1993)).
Comment 33: NSK disagrees with the Department's application of
NSK's rate for the April 1, 1978 through July 31, 1978, period as BIA
for these reviews. NSK argues that the Department should not have used
as BIA a weighted-average rate resulting from a four-month period. NSK
contends that the Department has never relied on a rate from a prior
administrative review that covered such a brief period of time.
Timken notes that the Department frequently applies as BIA margins
from LTFV investigations, which generally cover six-month periods.
Timken argues that NSK has not offered any evidence that a BIA rate
from a four-month period should be deemed unreliable, particularly
where the underlying determination has been through judicial review.
Department's Position: We agree with Timken. As stated in our
preliminary results, the Department sought to use, as BIA, the highest
rate NSK received in a previous POR, which in this instance was a rate
for a four-month POR. In the final results of review covering the 1974
through 1980 period the Department found a rate of 23.43 percent for
the period August 1, 1977 through July 31, 1978 (55 FR 369, June 1
1990). NSK challenged those results and the CIT remanded them to the
Department with instructions to use Treasury master lists for sales
made from August 1, 1977, through March 31, 1978, and to recalculate
the margin for the remaining (non-master list) period in accordance
with the CIT's instructions. Thus, the margin from the previous review
covered four months.
We note that NSK has not cited any instance where the Department
deemed a rate from a four-month period to be unsuitable as BIA.
Furthermore, NSK has not provided evidence that the rate in question is
not representative of the rate that we would have obtained from a
longer POR. Therefore, for U.S. sales for which we have relied on BIA,
we have applied the margin from the four-month POR, since it is the
highest margin for NSK in a previous segment of the proceeding.
Comment 34: Timken states that as a result of Koyo's model-match
exercise it became clear that Koyo failed to provide VCOM data for a
number of U.S. and home-market part numbers. In these instances the
Department applied a 20 percent variable-cost differential as BIA.
Timken suggests that splitting the variable costs of the sets would
provide VCOMs for those cups and cones which lacked variable costs in
the preliminary results.
Koyo acknowledges that there may still be some part numbers for
which no VCOM can be found after correction of the VCOM error. However,
Koyo submits that the Department's present BIA approach is more
efficient and reliable than the alternative of set-splitting advocated
by Timken.
Department's Position: In our calculations of FMV we did split the
variable costs of the home-market sets. There were four TRB sets sold
in the home market with no reported cost data which resulted in four
cup and cone models with no reported VCOM after set splitting. For
these sales we have continued to calculate a 20-percent VCOM
differential as BIA. There were no models of cups and cones sold in the
home market for which Koyo did not report a variable cost.
Where the variable cost of models sold in the U.S. was missing, we
did not use BIA to calculate a difmer adjustment to FMV. For U.S. sales
of those models with no variable cost information, we either matched
them to identical home-market models or we have used CV as the basis
for FMV pursuant to Sec. 773(a)(1) and Sec. 773(a)(2) of the Tariff
Act.
Comment 35: Timken argues that the Department should base NSK's
margin on total BIA. Timken contends that NSK failed to submit complete
home-market sales data on a sale-by-sale basis, despite a specific
request from the Department to do so. Timken notes that NSK reported
only home-market sales with quantities equal to five percent or more of
the corresponding U.S. sales, and generally utilized its own model-
match criteria, which deviated substantially from those currently
applied by the Department.
Timken further argues that the Department should determine that NSK
was ``non-cooperative'' and should apply first-tier BIA to NSK for all
periods (i.e., use the highest margin calculated for any respondent in
either prior or concurrent periods of review).
NSK argues that there are no grounds for the Department to base
NSK's margin on total BIA. NSK contends that the Department verified
that the home-market sales list reported by NSK was complete. NSK
contends that the Department, after verification, did not challenge or
further investigate the model-matching method employed by NSK.
NSK further argues that the Department's request for a sale-by-sale
listing of home-market sales was made in the context of a cost
investigation, and came two years after the submission of the original
home-market sales listing.
Finally, NSK argues that to the extent that NSK responses may have
been deficient, the omissions were minor and not intended to impede the
reviews, and first-tier BIA should not be applied.
Department's Position: We agree with NSK that the Department
verified the general completeness of NSK's reported home-market sales
listing. We also agree with NSK that the Department's request for a
sale-by-sale listing of home-market sales was pursuant to a cost
investigation.
However, as NSK has conceded both in its rebuttal brief and at the
hearing, NSK never responded to our July 1, 1988, request for a
complete home-market sales listing. This failure to report a sale-by-
sale listing has major implications with respect to the cost test and
the model match. By selectively reporting home-market sales, NSK could
have arguably contrived a result where U.S. models would be matched
with low-cost, low-price ``similar'' models. Therefore, while we
appreciate NSK's candor in admitting that it never provided the
complete sales listing, we cannot ignore the fact that the cost test
and model match were compromised.
We note that this conclusion applies only to U.S. models which do
not have ``identical'' matches in the home market. Throughout the TRB
reviews, all parties have operated on the premise that bearings with
identical nomenclature (i.e., product or identification number) are
identical in all physical aspects. Thus, models with identical
nomenclature will in every instance be matched to each other, and NSK's
incomplete sales listing does not compromise such a comparison.
Therefore, for these final results of review, we have relied on BIA
for sales of U.S. models for which NSK did not make sales of identical
merchandise in the home market. Where identical bearings were sold in
both markets, we have used NSK's reported data.
With respect to the choice of BIA, we note that, throughout the
many delays that have arisen in the course of the 1980/86 reviews, NSK
has generally been a cooperative respondent. Accordingly, for those
U.S. sales where no identical matches are possible, we have relied on a
second-tier BIA rate, which is the highest margin for NSK from any
preceding review period.
Comment 36: NSK takes issue with the Department's application of
BIA for U.S. sales lacking VCOM information. NSK contends that its COM
records are tied to the year of production, rather than the year of
sale and, therefore, the Department should use cost data from another
year within the 1980/86 PORs for any U.S. sale lacking difmer cost
data.
NSK further notes that in the 1974/80 segment of this proceeding,
the Department used 1980/85 cost data, if available, as BIA when cost
data for the 1974/80 period were missing.
Timken argues that while it may be reasonable for the Department to
search the annual period immediately prior to the period in which the
sale took place, any search beyond one period would be affected by
fluctuations in steel prices and changes in labor efficiency over the
course of the six-year review period.
Timken further notes that NSK's citation of the 1974/80 results is
misleading. In that case, the Department relied on NSK's 1980/85 cost
data as BIA because NSK had not retained cost information for the 1976/
80 PORs. Timken contends that for the 1980/86 PORs, NSK simply failed
to provide complete review-specific VCOM data despite the Department's
request for the information.
Timken counters that the Department specifically requested VCOM
information in both markets for the entire POR, and NSK failed to
provide it. Under these circumstances Timken concludes that application
of a BIA rate for sales with missing data is reasonable.
Department's Position: In our preliminary results of review we
applied BIA to sales of U.S. models lacking VCOM data which were
matched with similar home-market models. For these final results of
review, we applied BIA to sales of all U.S. models when we did not find
an identical match in the home market. Since we are not attempting to
find matches for U.S. models with no identicals in the home market,
VCOM data is not relevant to these results. Therefore, the issue of
missing VCOM data is moot.
Comment 37: Koyo contends that the Department should not apply BIA
to Koyo's U.S. sample sales because the Department did not specifically
request this information until September 17, 1993, 13 years after the
end of the first period and 7 years after the end of the final period
covered by these reviews. Koyo asserts that it was not the Department's
practice to require the submission of information regarding such sales
during the periods under review (1979/86). Koyo argues that it may have
been able to provide information regarding its U.S. sample sales had
the Department completed these reviews in a timely manner. In Koyo's
view, the application of BIA to these sales, in effect, would punish
Koyo for the Department's own delay. In the interest of fairness, Koyo
states that the Department should not apply any BIA to these sales.
Timken agrees with the Department's proposal to use BIA for U.S.
sample sales, based on the value of sample sales in the home market, as
obtained at verification. Timken also states that the Department should
guard against increasing the value of U.S. sales (the denominator of
the weighted-average margin) and increasing the value of the duties due
(the numerator of the weighted-average margin) by a BIA factor which
would result in the same amount of duties collected absent the
application of BIA.
Department's Position: In its response to our supplemental
questionnaire, Koyo failed to comply with our request for a list of
U.S. sample sales, stating that ``to produce such a sales listing, AKC
would have to review manually all of its sales invoices and records for
the covered periods'' (November 1, 1993 submission, p. 6). In that
letter Koyo stated that time restraints were the only obstacles to
producing a list of U.S. sample sales. However, in its case brief Koyo
stated that ``it is unable, at such a removed point in time, to
identify its U.S. sample sales, because (it) did not maintain a
database of its U.S. sales that identified its sample sales.''
Section 776(c) of the Tariff Act requires that BIA be applied when
a party does not produce requested information whether the non-
compliance is due to refusal or mere inability to produce the desired
data (Olympic Adhesives, Inc., v. United States, 899 F.2d 1565, 1574
(1990)). Koyo failed to reply to our request for a list of U.S. sample
sales, saying the effort would be too great. In our preliminary
analysis memorandum for these reviews we identified the methodology we
intended to use as BIA for the missing U.S. sample sales in our final
results of review. Therefore, we are applying BIA to these sales. We
only have information on Koyo's home market sample sales. Therefore, we
used the data for home market sample sales from the 1985/86 period. We
used the relationship of home market sample sales to total home market
sales to represent the relationship of U.S. sample sales to total U.S.
sales. With this information, we determined a value for those U.S.
sample sales and applied a BIA margin to that value. We added both the
resulting duties due amount and the calculated value of the sample U.S.
sales to our respective margin and value totals in deriving our
weighted-average margin.
Comment 38: Timken asserts that Koyo's failure to report home-
market sample sales in its original submission, and then, at the
Department's request, Koyo's submission of a list rather than a
computer tape, should not be countenanced. Timken claims that the
printout of home-market sample sales Koyo provided in its supplemental
response is clearly from a computer file.
Therefore, Timken concludes that submission of a computer tape of
home-market sample sales could not be as difficult as Koyo professes.
Moreover, Timken chides the Department for not requesting such data
despite Timken's timely protest of Koyo's failure to report home-market
sample sales. Absent sample sales, the home-market sales listing is
incomplete and unreliable. In Timken's view, the only proper solution,
short of rejecting the entire response, is to require immediate
submission of a computer tape listing of sample sales.
Timken maintains that Koyo elected not to comply because the
Department has only limited options regarding ``best information''
under these circumstances. Timken states that in 1990/91 and 1991/92
TRBs the Department addressed Koyo's failure to support its claim that
particular home-market sales were samples.
Koyo argues that the Department properly excluded its home-market
sample sales because they are de minimis in volume and value and have
been verified by the Department as sample sales outside the ordinary
course of trade. Koyo characterizes Timken's suggestion that Koyo be
required to submit an entire new computer tape of the home-market
sales, including these sales, as ``ridiculous'' and contrary to
Timken's own stated desire not to ``increase the cost of bringing this
proceeding to a final conclusion.''
Department's Position: In our October 1993 supplemental
questionnaire we requested that Koyo submit a listing of its home-
market sample sales. In response to our supplemental questionnaire Koyo
submitted a hard copy computer printout of its home-market sample sales
(November 1, 1993, Exhibit 1). At verification we confirmed the
accuracy of the nature of these sales (i.e., samples) and the value and
volume reported. We determined that the volume and value of these sales
were indeed minuscule and the fact that we did not have them
electronically could not have any measurable or significant impact on
our ability to conduct these administrative reviews. We are confident
that the home-market data base we used for comparison purposes was
indeed sufficient and complete. Requiring a computer tape of the home-
market sample sales would be, in this instance, superfluous and
unnecessarily time-consuming.
Comment 39: Timken notes that the Department relied upon the
revised data submitted by Koyo in 1991 for its analysis and did not
consider other information contained in the administrative record. In
view of Koyo's decision to revise its data, Timken asserts that the
Department should draw adverse inferences in any instance where data
are deficient or missing. According to Timken, the Department's failure
to draw any adverse inference based on the omission of various data
(e.g., home-market sample sales, some variable costs) allowed Koyo to
control in part the Department's calculation of margins.
Koyo asserts that it is absurd for Timken to suggest that Koyo's
co-operation in the resubmission of its data for the 1979 through 1986
PORs should be used against Koyo, when Koyo is in no way responsible
for the delays in this proceeding. Koyo states that the CIT made it
clear that respondents must not be prejudiced by the government's undue
delays in the completion of administrative reviews, and, similarly,
respondents cannot be punished for intervening changes in the
Department's antidumping methodologies (Koyo Seiko).
Koyo also objects to Timken's conclusion that Koyo's cost data
should be rejected in their entirety. In Koyo's opinion the Department
should not even conduct a below-cost investigation (see Comment 30).
However, should the Department do so, Koyo states that the Department
has had several years in which to request supplemental cost data or
other germane information, and the Department cannot now decide that
Koyo's data are inadequate without unduly and illegally prejudicing
Koyo's rights in these proceedings.
Department's Position: We disagree with Timken's contention that we
should draw the most adverse inferences in applying BIA to Koyo's
revised submission where data are deficient or missing. The Department
agreed to accept a consolidated response from Koyo to facilitate and
expedite the completion of these administrative reviews. We verified
Koyo's consolidated response and found that, for the most part, Koyo's
data were reasonable and accurate (see Verification Report, November
22, 1993, and Comment 31 regarding the adequacy of Koyo's cost data).
We used BIA when we did not agree with Koyo's methodology in
calculating an expense, or when data were missing from the submission.
Therefore, we do not believe that we allowed Koyo to control the
calculation of the margins in any way.

Comments Regarding Clerical and Ministerial Errors

Comment 40: Both Timken and Koyo addressed the following clerical
errors:
1. The preliminary program erroneously sets the VCOM and other
variables equal to zero (see Comment 45).
2. The denominator for recalculating the home-market credit expense
in 1979/80 must be added to the program.
3. The ratio of inventory carrying costs to total U.S. sales for
the 1980/81 POR should be multiplied by the landed cost, consistent
with the other PORs.
4. The Department failed to adjust Koyo's reported home-market unit
price for post-sale freight expense in calculating FMV. However, Timken
believes that no change is necessary because Koyo failed to report its
freight expenses correctly (see Comment 20).
5. The Department should delete certain duplicate observations in
the home-market database.
6. Timken maintains and Koyo confirms that Koyo reported direct and
indirect selling expenses, G&A expense, and interest expense as a
percentage of COM, not fixed amounts per unit as interpreted by the
Department.
Department's Position: We agree with Koyo and Timken and have made
each of these changes in the calculations for these final results of
review.
Comment 41: Timken states that the Department inadvertently
reversed signs in its preliminary program for Koyo by applying the 20-
percent BIA rate for models with no reported home-market VCOM.
Department's Position: We agree with Timken and have corrected the
program for these final results.
Comment 42: Timken asserts that the Department must ensure that
selling-expense adjustments to CV are consistent in all respects with
price-based adjustments. Timken contends that not all of the expenses,
originally reported as direct selling expenses, were ultimately
considered direct by the Department (e.g., freight, discounts).
Department's Position: We agree with Timken's assertion that
deducting the expenses reported as ``direct'' selling expenses in
Koyo's CV database was incorrect. For these final results we have not
used this variable in our CV calculation because it represented
commission and rebate expenses, both of which we did not treat as
direct selling expenses for Koyo in these final results (see our
responses to Comments 11 and 23, respectively). Therefore, the only
valid direct deduction from CV is the weighted-average credit expense
for home market sales from the sales database for each POR.
Comment 43: Koyo maintains that the Department's constant practice
in its TRB determinations has been to break ties between two equally
similar matches by first comparing the sum of the deviations for the
two models, then comparing the levels of trade for the two models, then
comparing the cost deviations for the two models, and finally by
ranking the two models alphanumerically. According to Koyo, the
Department inadvertently omitted the level-of-trade (LOT) tie-breaker
in conducting the model match for these PORs. Koyo notes that this is
not the same issue regarding LOTs that Koyo has raised unsuccessfully
in previous reviews--that the Department should not cross levels of
trade in comparing identical models before looking for similar models
at the same level of trade.
Timken agrees, stating that where two matches with equivalent
``sums of the deviations'' are found, each at a different LOT, the
Department should select the match at the same level of trade.
Department's Position: Koyo's statement that we have consistently
selected the best model match based on, in order, ``sum of the
deviations,'' LOT, and cost differences in all of its recent TRB
determinations, is correct in the context of our computer programs for
some previous reviews of this finding and the 1987 order of Koyo's U.S.
sales of TRBs. However, in those previously completed reviews, the
Department's stated methodology was to rank similar merchandise by
minimizing first the ``sum of the deviations'', second, the difference
in the VCOM of the U.S. and similar home-market product, third, the
differences in LOT, and lastly, alphanumerically. When the ``sum of the
deviations'' was the same for two or more bearings, the Department
ranked them according to the similarity in costs, and, if the costs
were the same, according to the similarity in LOT (TRBs I, 57 FR 4978
(February 11, 1992)). In our computer programs for these administrative
reviews, we have followed this methodology, ranking similar bearings by
``sum of the deviations'', differences in cost, LOT, and, lastly,
alphanumerically.
Comment 44: Koyo states that the load rating for some sales of one
model was reported incorrectly on Koyo's submitted computer tape.
Although the cup and cone of each model should have the same basic load
rating as the assembled model, in this instance a computer input error
on Koyo's part resulted in an inconsistency in reporting the load
rating of the cup, cone, and set. Koyo requests that the Department
correct this error for these final results and cites AFBs III as
precedent.
Timken contends that the Department should make the requested
correction only if the error is discernible based on the evidence of
record at the time the error was identified. Timken notes that the
documents cited by Koyo that contain corroborating evidence were all
submitted in administrative reviews for periods other than 1979/86.
Department's Position: The load rating of individual bearings for
the periods under review is available in any Koyo catalogue and, as
such, is part of the public domain. Therefore, we have made the
appropriate corrections in the load rating of the model designated by
Koyo.
Comment 45: Timken points out that the preliminary program
erroneously sets certain home-market VCOMs equal to zero. Timken states
that for certain set-split bearings only VCOM is zero. For others the
VCOM, net unit price, and indirect selling expenses are zero, resulting
in erroneous calculations for FMV and the weighted-average home-market
indirect selling expenses.
Timken notes that this problem has been discussed with Department
personnel and revision of the program instructions will apparently
result in assignment of correct values.
Koyo affirms Timken's observation, stating that the Department
erred in its calculation of the cup and cone ratios during the set-
splitting exercise, resulting in the progressive reduction of the cup
and cone ratio variables and, consequently, the VCOM ratio.
Department's Position: We agree with Koyo and Timken. We
inappropriately altered VCOM when we merged the home-market sales and
cost data files and split the home-market sales of TRB sets into sales
of cups and cones. However, Koyo's explanation that the cup and cone
ratios are progressively reduced toward zero is wrong, as evidenced
from the computation of the home-market net unit prices and indirect
selling expenses for sales of split sets. We have determined that the
value for the VCOM itself was reduced toward zero, and have corrected
the calculations accordingly.
We also found a similar error in the calculations where we split
home-market sales of TRB sets with no reported cup or cone ratios using
the weighted-average of the reported cup and cone ratios as BIA for
each POR. We have also corrected this error.
Comment 46: Timken believes that a discrepancy in the U.S. sales
values reported by Koyo and

[Text truncated at 120,000 characters. The full text is on the page linked above.]

---

Source: Frix Law Library, https://www.frixlaw.com/law-library/documents/fr%3A94-27783. Public record. Not legal advice.