American Seafood Distributors Association
February 7, 2007
Rep. Charles Rangel, Chairman
Committee on Ways & Means
U.S. House of Representatives
1102 Longworth House Office Building
Washington, D.C. 20515
Dear Mr. Chairman:
The American Seafood Distributors Association
(“ASDA”) is an association of U.S. importers of seafood, including importers of
shrimp from the six countries that are now subject to an antidumping duty order
issued by the Commerce Department on February 1, 2005. Shrimp imports are
absolutely vital to our members since domestic shrimp suppliers produce only 10
percent of the shrimp needed to supply fast growing demand for this valuable
and relatively inexpensive protein source.
On ASDA’s behalf, I hereby respond to your
January 31, 2007 request for comments from the public about what action the
Committee should take with respect to the Department of Commerce (DOC) proposal
concerning: (1) which methods of comparing normal values and U.S. prices can
or should be used in an investigation; and (2) whether the DOC should “zero”
negative antidumping margins when it uses any of the permissible comparison
methods. ASDA strongly believes that the Committee should accept ASDA’s view
that the DOC is required to adopt the following policy regarding the
calculation of the weighted average dumping margin in each antidumping
investigation:
1. Under
U.S. law, the Department must “normally” employ the average-to-average
comparison method. Moreover, as required by the consistent rulings of the
WTO’s Dispute Settlement Body, including very recent decisions issued in cases
brought by Japan and Ecuador, when it uses the average-to-average comparison
method in an investigation, the Department must provide “offsets” for
non-dumped comparisons, i.e., it must not zero negative dumping margins.
Furthermore, the Department must calculate average normal values and average U.S. prices over the entire period of investigation except in those very limited
circumstances where using POI averages would yield unfair comparisons. Zeroing
is not permitted in these limited circumstances, and offsets must be provided
here, as well.
2. The
Department may employ the transaction-to-transaction comparison method only in
the types of “unusual situations” contemplated by Congress. The Department
must also provide offsets for non-dumped comparisons when it uses
the
transaction-to-transaction methodology, as the WTO’s Appellate Body recently
held in Japan’s WTO case.
3. The
Department may employ the average-to-transaction comparison method only in
those comparatively rare situations where targeted dumping is alleged to have
occurred. The Department must also provide offsets for non-dumped comparisons
when it uses the average-to-transaction method. The WTO has affirmed that
zeroing is not permitted with this comparison methodology.
Disregard of the WTO’s rulings
through the Committee’s disapproval of the DOC proposal could have serious and
far-reaching consequences for American interests. On the other hand, approval
of the DOC proposal would fully comply with domestic and international law, as
we now explain in detail.
I. THE
DEPARTMENT MUST NORMALLY USE THE AVERAGE-TO-AVERAGE COMPARISON METHOD IN INVESTIGATIONS
AND MUST PROVIDE OFFSETS FOR NON-DUMPED COMPARISONS
A. Relevant
Legal Background
Article 2.4.2 of the WTO Antidumping Agreement
provides that:
the existence of margins of
dumping during the investigation phase shall normally be established on the
basis of a comparison of a weighted average normal value with a weighted
average of prices of all comparable export transactions or by a comparison
of normal value and export prices on a transaction-to-transaction basis.
A normal value established on a weighted average basis may be compared to
prices of individual export transactions if the authorities find a pattern of
export prices which differ significantly among different purchasers, regions or
time periods, and if an explanation is provided as to why such differences
cannot be taken into account appropriately by the use of a weighted
average-to-weighted average or transaction-to-transaction comparison.
(Emphasis added.) Section 777A(d)(1) of the Uruguay Round
Agreements Act (URAA), 19 U.S.C. § 1677f-1(d)(1), implements Article 2.4.2
as follows:
(A) In general. In
an investigation under part II of this subtitle, the administering authority
shall determine whether the subject merchandise is being sold in the United States at less than fair value --
(i) by comparing
the weighted average of the normal values to the weighted average of the export
prices (and constructed export prices) for comparable merchandise, or
(ii) by comparing
the normal values of individual transactions to the export prices (or
constructed export prices) of individual transactions for comparable
merchandise.
(B) Exception. The
administering authority may determine whether the subject merchandise is being
sold in the United States at less than fair value by comparing
the weighted
average of the normal values to the export prices (or constructed export
prices) of individual transactions for comparable merchandise, if—
(i) there is a
pattern of export prices (or constructed export prices) for comparable
merchandise that differ significantly among purchasers, regions, or periods of
time, and
(ii) the
administering authority explains why such differences cannot be taken into
account using a method described in paragraph (1)(A)(i) or (ii).
Thus, both the Antidumping Agreement and the
implementing provisions of the URAA established three separate comparison
methods for calculating dumping margins in investigations: (1)
average-to-average; (2) transaction-to-transaction; and (3) average-to-
transaction. As we now discuss, both the Antidumping
Agreement and the URAA established a strong preference for use of the
average-to-average method, stating that it would be the “normal” method.[1]
Thus, by definition, the transaction-to-transaction and average-to-transaction
methods are not the normal methods. Rather, they are exceptional methods, and
our review of the investigations that the Department has conducted since
passage of the URAA indicates use of the transaction-to-transaction method on
only three occasions in twelve years.
Moreover, the average-to-transaction method can be
used only in instances of “targeted dumping,” which the Antidumping Agreement
and the URAA define narrowly to mean a “pattern” of export prices that “differ
significantly among different purchasers, regions or time periods.” To the
best of the ASDA’s knowledge, the targeted dumping method has never been used
in an investigation since passage of the URAA.
ASDA urges the Committee to advise the Department
of its view that it should continue using its longstanding normal practice of
making average-to-average comparisons in antidumping investigations. Switching
to the transaction-to-transaction method as the normal method, for example,
would flatly contradict U.S. law and Article 2.4.2 of the Antidumping
Agreement. Moreover, regardless of the method it uses, the Department is
obligated to make offsets in all investigations for non-dumped comparisons. In
other words, to comply with the WTO rulings, it cannot “zero” any negative
dumping margins.
B. The
Legislative History Demonstrates the Strong Congressional Preference for
Average-to-Average Comparisons in Investigations
Although the text of 19 U.S.C. § 1677f-1(d)(1)
does not expressly state a preference among the three comparison methods, the
legislative history reveals the indisputable intention of Congress that the
Department use the average-to-average comparison method except in carefully
limited instances. Specifically, the SAA on the URAA states that:
Consistent with the
{WTO Antidumping} Agreement, new section 777A(d)(1)(A)(i) provides that in
an investigation, Commerce normally will establish and measure dumping margins
on the basis of a comparison of a weighted-average of normal values with a
weighted-average of export prices or constructed export prices. To ensure
that these averages are meaningful, Commerce will calculate averages for
comparable sales of subject merchandise to the U.S. and sales of foreign like
products. In determining the comparability of sales for purposes of inclusion
in a particular average, Commerce will consider factors it deems appropriate,
such as the physical characteristics of the merchandise, the
region of the country in which the merchandise is sold, the time period, and
the class of customer involved.
In addition to the
use of averages, section 777A(d)(1)(A)(ii) also permits the calculation of
dumping margins on a transaction-by-transaction basis. Such a methodology
would be appropriate in situations where there are very few sales and the
merchandise sold in each market is identical or very similar or is
custom-made. However, given past experience with this methodology and the
difficulty in selecting appropriate comparison transactions, the Administration
expects that Commerce will use this methodology far less frequently than the
average-to-average methodology.
H.R. Doc. No. 103-465 (1994), Vol. 1, at 842-43 (emphasis
added). Under 19 U.S.C. § 3512(d), the SAA constitutes the “authoritative expression
by the United States concerning the interpretation and application of the
Uruguay Round Agreements and this Act in any judicial proceeding in which a
question arises concerning such interpretation or application.” In addition:
As is the case with
earlier Statements of Administrative Action submitted to the Congress in
connection with fast-track trade bills, this Statement represents an
authoritative expression by the Administration concerning its views regarding
the interpretation and application of the Uruguay Round agreements, both for
purposes of U.S. international obligations and domestic law. Furthermore, the
Administration understands that it is the expectation of the Congress that
future Administrations will observe and apply the interpretations and
commitments set out in this Statement. Moreover, since this Statement will be
approved by the Congress at the time it implements the Uruguay Round
agreements, the interpretations of those agreements included in this Statement
carry particular authority.
SAA at 656.
After Congress expressed its strong preference for
use of the average-to-average method, as well as its view that the
transaction-to-transaction method would be used “far less frequently,” the
Department adopted 19 C.F.R. § 351.414(c)(1), which provides that: “In an
investigation, the Secretary normally will use the average-to-average method.
The Secretary will use the transaction-to-transaction method only in unusual
situations, such as when there are very few sales of subject merchandise and
the merchandise sold in each market is identical or very similar
or is custom-made.” The preamble to the Department’s proposed
regulation referred to the average-to-average method as the “preferred method
in an antidumping investigation.”[2]
Antidumping Duties; Countervailing Duties; Notice of Proposed Rulemaking and
Request for Public Comments, 61 Fed. Reg. 7308, 7348 (Feb. 27, 1996).
Nothing in the WTO rulings on
zeroing affects this “preference.” See also Antidumping
Duties; Countervailing Duties; Final Rule, 62 Fed. Reg. 27295, 27373 (May
19, 1997).
Thus, in light of unambiguous Congressional
intent, as implemented in the regulations, the Department must continue to make
average-to-average comparisons in investigations unless “unusual” facts warrant
the use of the transaction-to-transaction method.[3]
C. The
Department Must Continue Its Longstanding Practice of Calculating Average
Normal Values and Average U.S. Prices Over the Entire Period of Investigation
The Department must also continue its normal
practice of calculating the average normal value and the average U.S. price over
the entire period of investigation, rather than calculating them over
shorter time periods (e.g., on a monthly or semi-annual basis). This is
because 19 C.F.R. § 351.414(d)(3) provides that the Department will
calculate averages over shorter time periods in only exceptional
circumstances: “When applying the average-to-average method, the Secretary
normally will calculate weighted averages for the entire period of
investigation or review, as the case may be.” In fact, the Department may only
calculate average normal values or U.S. prices for shorter periods when they
“differ significantly” over the course of the POI. Id. As the
Department noted in the preamble to this regulation:
In the Department’s
view, price averaging means establishing an average price for all
comparable sales. In general, we believe it is appropriate to average prices
across the period of investigation, though we recognize that there are
circumstances in which other averaging periods are more appropriate.
Accordingly, the proposed rule is designed to ensure that the time periods over
which price averages and comparisons are made comports with the circumstances
of the case, while maintaining a preference for period-wide averaging.
62 Fed. Reg. at 27373 (May 19, 1997) (emphasis added).
In accordance with this statement of intent, the
Department’s longstanding and consistent practice has been to calculate average
normal values and U.S. prices for all sales made over the entire POI. The few
instances in which the Department has calculated averages over a shorter period
involved unique circumstances. For example, the Department has used shorter
averaging periods in cases involving “high inflation” economies. See, e.g.,
Certain Cold-Rolled Flat-Rolled Carbon-Quality Steel Products from Turkey,
65 Fed. Reg. 1127, 1132-33 (Jan. 7, 2000) (preliminary determination), aff’d,
65 Fed. Reg. 15123 (Mar. 21, 2000) (final determination) (the
Department calculated normal values and U.S. prices on a
monthly basis because Turkey experienced significant inflation during the POI).
The Department has also used shorter averaging
periods where the use of a POI averaging period would distort the margin
calculation because price levels changed significantly during the POI. For
example, in Stainless Steel Sheet and Strip in Coils from the Republic of
Korea, 64 Fed. Reg. 30664, 30676 (Jun. 8, 1999) (final determination), the
Department
determined that the significant and precipitous devaluation of
the Korean won during the Asian financial crisis would not yield appropriate
comparisons. Therefore, it used two averaging periods. In Static Random
Access Memory Semiconductors from Taiwan, 63 Fed. Reg. 8909, 8925 (Feb. 23,
1998) (final determination), the Department utilized quarterly averaging
periods because the comparison and U.S. markets had experienced significant and
consistent price declines during the POI. Similarly, in Dynamic Random
Access Memory Semiconductors of One Megabit and Above from the Republic of
Korea, 58 Fed. Reg. 15467, 15476 (Mar. 23, 1993) (final determination), the
Department found that monthly averaging periods were more representative
because prices in both markets had consistently declined throughout the POI.
In contrast, the Department has declined to use
shorter averaging periods where no evidence existed that distortions would
result from averaging prices over the entire POI. See, e.g., Live
Swine from Canada, 70 Fed. Reg. 12181 (Mar. 11, 2005), Issues and Decision
Memorandum at Comment 5 (the Department did not depart from its “normal
practice” of using annual averaging periods because respondent did not
demonstrate that distortions in the dumping calculations would result unless
the Department utilized shorter averaging periods); Certain Polyester Staple
Fiber from the Republic of Korea, 65 Fed. Reg. 16880 (Mar. 30, 2000),
Issues and Decision Memorandum at Comment 3 (the Department declined to use
shorter averaging periods because “we were seeking to examine whether there was
an overall general trend in prices of all subject merchandise sold by the
respondents” and “price changes during the POI were neither significant nor
consistent”).
All of these cases confirm that the Department’s
longstanding and consistent practice has been to calculate average prices for
the entire POI except in unusual situations. No legal basis exists to change
that practice for four reasons: (1) Article 2.4.2 of the Antidumping Agreement
states that the average-to-average method will “normally” be applied to “all
comparable export transactions;” (2) 19 U.S.C. § 1677f-1(d)(1)(A)(i) similarly
requires a calculation of “the weighted average of the normal values to the
weighted average of the export prices (and constructed export prices) for
comparable merchandise” without limitation of the time period for averaging;
(3) the SAA states that the statutory provision is intended to implement the
Antidumping Agreement, which as just noted, normally requires averaging of “all
comparable export transactions;” and (4) 19 C.F.R. § 351.414(d)(3) states that
the Department “normally will calculate weighted averages for the entire period
of investigation,” except where “prices differ significantly over the course of
the period of investigation.” Accordingly, the Department should continue to
make average-to-average comparisons in investigations over the entire POI,
except in unusual circumstances of the type discussed above.
D. Whenever
It Makes Average-To-Average Comparisons, the Department Must Provide Offsets
for Non-Dumped Comparisons
The Department cannot employ zeroing when it uses
any version of the average-to-average comparison method, according to the WTO
Appellate Body’s finding in United States – Final Dumping Determination on
Softwood Lumber from Canada, WT/DS264/AB/R, Aug. 11, 2004, at para. 108.[4]
The AB reached the same result in cases brought by the EU and Japan. See
WT/DS294/AB/R (17 January 2006) and WT/DS322/AB/R (9 January 2007).
As a result of the WTO’s definitive and repeated invalidation
of zeroing whenever it uses the average-to-average method, we agree with the
Department’s statement in its December 27, 2006 Federal Register Notice that it
is required to abandon the use of average-to-average comparisons without
offsets because it “is necessary to implement the recommendations of the World
Trade Organization Dispute Settlement Body.” 71 Fed. Reg. at 77722. In other
words, the Department must terminate its practice of “zeroing” in
investigations and, instead, grant offsets for non-dumped comparisons in order
to bring its practice into conformity with its obligations under Article 2.4.2
of the Antidumping Agreement. Moreover, because the Appellate Body’s holdings
expressly extend to all variations of the average-to-average method, offsets
for negative margins must be granted in investigations in those limited
situations in which it uses shorter averaging periods than the entire POI.
II. THE
DEPARTMENT SHOULD CONTINUE TO MAKE TRANSACTION-TO-TRANSACTION COMPARISONS ONLY
IN EXCEPTIONAL SITUATIONS, AND IT MUST ELIMINATE ZEROING AND GRANT OFFSETS WHEN
IT DOES
The Department has rarely exercised its authority
to make transaction-to-transaction comparisons in investigations. For example,
it did so in Large Newspaper Printing Presses because there were only a
few sales of subject merchandise and the merchandise itself was highly
customized, which is exactly the type of situation contemplated by the SAA and
19 C.F.R. § 351.414(c)(1).[5]
See Large Newspaper Printing Presses and Components Thereof, Whether
Assembled or Unassembled, from Germany, 61 FR 38166 (Jul. 23, 1996) (final
determination); Large Newspaper Printing Presses and Components Thereof,
Whether Assembled or Unassembled, from Japan, 61 Fed. Reg. 38139 (Jul. 23,
1996) (final determination).
More recently, the Department made
transaction-to-transaction comparisons in its Section 129 determination in Certain
Softwood Lumber Products from Canada in which the Department implemented
the WTO Appellate Body’s finding with regard to zeroing when making
average-to-average comparisons in investigations. See Softwood
Lumber – Section 129 Determination, 70 Fed. Reg. 22636 (May 2, 2005).
There, the Department explained that transaction-to-transaction comparisons
were necessary because of the “high level of price volatility” in the U.S. and Canadian markets that could distort the results of the
dumping calculations if average-to-average comparisons were made. Id.
at 22637-39; see also Preliminary Determination under
Section 129 of the Uruguay Round Agreements Act:
Antidumping Measures on Certain Softwood Lumber Products from Canada at 10,
available at http://www.ia.ita.doc.gov/download/ section129/Canada-Lumber-129-Prelim-013105.pdf
(“By applying the transaction-to-transaction analysis in this case, we are not
intending to implement a practice that applies to all antidumping
investigations. As discussed above, the use of this methodology is premised on
the combination of facts and circumstances that have led to and support this
determination.”) Thus, the Department premised its use of the
transaction-to-transaction methodology on a very “unusual situation.”[6]
When making these transaction-to-transaction
comparisons, the Department erroneously did not provide offsets for non-dumped
comparisons on the ground that the Appellate Body’s finding was limited to
average-to-average comparisons. However, the Appellate Body in Japan’s case expressly
found that the DOC’s use of zeroing in connection with the
transaction-to-transaction method in original investigations was illegal. See
AB Report at par. 190. And, even if it had not reached this result, the URAA
limits the use of the transaction-to-transaction method to exceptional
circumstances. As noted in the preamble to the final rule:
In the Department’s
view, the SAA makes clear that Congress did not contemplate broad application
of the transaction-to-transaction method. SAA at 842. Specifically, the SAA
recognizes the difficulties the agency has encountered in the past with respect
to this methodology and suggests that even in situation where there are very
few sales, the merchandise in both markets should also be identical or very
similar before the agency would make transaction-to-transaction comparisons.
Accordingly, we continue to maintain that the transaction-to-transaction
methodology should only be applied in unusual situations.
62 Fed. Reg. at
27374 (May 19, 1997) (emphasis added).
III. CONCLUSION
For the foregoing reasons, the ASDA strongly urges
the Committee to inform the Department that it supports discontinuation of
zeroing in all investigations.
Respectfully
submitted,
Wally
Stevens
President
[1]
Neither the WTO Antidumping Agreement nor the URAA defines the words “normal”
or “normally.” However, the dictionary meaning is “standard,” “common,” or
“typical.” Thus, any application of a different method than the
average-to-average method must necessarily be “non-standard,” “uncommon,” or
“atypical.” In other words, it is the exception, not the rule. In fact, the Statement
of Administrative Action (SAA) expressly refers to the
transaction-to-transaction method and the average-to-transaction method as
“exceptions.”
[2]
In contrast, this preamble stated that the “transaction-to-transaction method
will only be used in unusual circumstances.” 61 Fed. Reg. at 7349 (Feb. 27,
1996).
[3]
The average-to-transaction method was expressly intended for use in the rarest
of situations involving targeted dumping.
[4] See also
Notice of Determination under Section 129 of the
Uruguay Round Agreements Act: Antidumping Measures on Certain Softwood Lumber
Products from Canada, 70 Fed. Reg. 22636, 22640 (May 2, 2005) (Softwood
Lumber – Section 129 Determination).
[5]
As noted in Chapter 6 of the Antidumping Manual (at 7), “We may also establish
dumping margins by comparing NV and EP or CEP on a transaction-to-transaction
basis. This is normally done only for large capital goods made to order, such
as transformers. The difference between these custom-made products render average
prices meaningless.”
[6]
In Softwood Lumber – Section 129 Determination, the Department also
stated that its computer capabilities had improved since the passage of the
SAA, which meant that it could now perform transaction-to-transaction
comparisons on large databases. However, nothing in the SAA linked the strong
preference of Congress for average-to-average comparisons to the Department’s
computer capabilities. See 70 Fed. Reg. at 22641. Thus, improvements
in the Department’s technical capabilities are irrelevant to the choice of a
comparison method. |