Opinion & Order
CARMAN, Judge:
This action is before the Court following the final results of redetermination pursuant to remand (“Remand Results”), filed by the Department of Commerce (“Commerce”) on December 20, 2010. (ECF No. 43.) Plaintiff Shandong Rongxin Import & Export Company (“Shandong”) challenges
the Remand Results only with respect to the value Commerce assigned to the wage rate factor of production in constructing the normal value of the subject merchandise. (Pl.’s Cmts. on Final Results of Re-determination Pursuant to Remand (“Pl.’s Cmts.”) at 1.) For the reasons set forth below, the Remand Results are sustained in part and remanded in part.
Background
In
China First Pencil Co. v. United States,
34 CIT -, 721 F.Supp.2d 1369 (2010), the Court sustained-in-part and remanded-in-part Commerce’s final determination in the 2006-2007 administrative review of the antidumping duty order on Certain Cased Pencils from the People’s Republic of China (“PRC”). The above captioned case was consolidated -with
China First
shortly before the Court issued its opinion; after Commerce issued the Remand Results, Shandong was the only party seeking to challenge the Remand Results. Accordingly, the Court severed the cases and issued a judgment in
China First Pencil Co. v. United States,
Court No. 09-00325. Shandong now proceeds on its own.
Jurisdiction and Standard of Review
The Court has jurisdiction over this matter pursuant to 19 U.S.C. § 1516a(a)(2) and 28 U.S.C. § 1581(c).
By statute, when reviewing Commerce’s determination, the Court “shall hold unlawful any determination, finding, or conclusion found ... to be unsupported by substantial evidence on the record, or otherwise not in accordance with law.” 19 U.S.C. § 1516a(b)(l)(B)(i);
see also China First,
721 F.Supp.2d at 1372.
Discussion
In determining the dumping margin in this case, Commerce calculated the normal value of the subject merchandise on the basis of the values of certain factors of production in “a market economy country or countries considered to be appropriate by [Commerce].” 19 U.S.C. § 1677b(c)(1). The purpose of this statute is “to assess the price or costs of factors of production of [the subject merchandise in a surrogate market economy country,] in an attempt to construct a hypothetical market value of that product in [the nonmarket economy country].”
Nation Ford Chem. Co. v. United States,
166 F.3d 1373, 1375 (Fed.Cir.1999) (internal quotations omitted). In deciding how to appropriately value factors of production, Commerce is required to “utilize, to the extent possible, the prices or costs of factors of production in one or more market economy countries that are— (A) at a level of economic development comparable to that of the nonmarket economy country, and (B) significant producers of comparable merchandise.” 19 U.S.C. § 1677b(c)(4). When Commerce elects to calculate normal value pursuant to this method, one of the factors of production for which it must establish a value is labor, or the wage rate. 19 U.S.C. § 1677b(c)(3).
In
Dorbest Ltd. v. United States,
604 F.3d 1363, 1372 (Fed.Cir.2010), the Court of Appeals for the Federal Circuit (“CAFC”) invalidated the regulation in which Commerce had established its preferred method for valuing labor as a factor of production. The CAFC held that the agency’s regression-based methodology produced a value for labor that failed to utilize data only from countries that are at a level of economic development comparable to that of the nonmarket economy country, and that are significant producers of comparable merchandise, as required by 19 U.S.C. § 1677b(c)(4).
See id.
Because the final determination challenged in
China First
was issued prior to the CAFC
decision in
Dorbest,
the Court remanded this case to Commerce with instructions to “adjust the surrogate value for labor to conform with the statutory requirements [of 19 U.S.C. § 1677b(c)(4) ], as explained in
Dorbest.” China First,
721 F.Supp.2d at 1382. The value Commerce subsequently assigned to the wage rate is the only aspect of the Remand Results presently challenged by Shandong.
A. Commerce’s Wage Rate Determination on Remand
As a threshold matter, Commerce determined that when valuing labor as a factor of production, data from several countries is preferable to data from a single country. While the agency normally values most other factors of production by using figures from a single surrogate country, Commerce continues to find, as it did before the CAFC invalidated its regression-based analysis in
Dorbest,
that labor is different.
Remand Results
at 10-11;
see also
19 C.F.R. § 351.408(c)(2). According to Commerce, labor is distinguished from other factors of production because it demonstrates wide variability, even between countries with similar gross national income per capita (“GNI”).
Remand Results
at 10. Commerce justifies this view by citing “many socio-economic, political and institutional factors, such as labor laws and policies unrelated to the size or strength of an economy, that cause significant variances in wage levels between countries,” and by noting that “labor is not traded internationally.”
Id.
at 11
(citing International Labor Organization, Global Wage Report: 2009 Update,
at 5, 7 and 10 (2009),
available at
http://www.ilo.org/ wemsp5/groups/public/@dgreports/@ dcomm/documents/publication/wcms_ 116500.pdf (last visited April 21, 2011).) Accordingly, Commerce valued labor by calculating a simple average of certain specific wage rate data from the group of countries the agency determined to be both economically comparable to the PRC and significant producers of comparable merchandise.
Id.; see also Remand Determination of 2006-2007 Administrative Review of the Antidumping Duty Order on Certain Cased Pencils from the People’s Republic of China: Industry-Specific Wage Rate Selection
(“Wage Rate Memo”), October 13, 2010 (AR 2707) at 4-5.
1.
Economic Comparability
To identify which countries are at a level of economic development comparable to the PRC, Commerce began with the Surrogate Country Memo, in which it had identified five countries as potential surrogates from which the agency would be able to derive values for all the other factors of production, besides labor.
Wage Rate Memo
at 2. Commerce ranked the five countries in order of ascending GNI: India, the Philippines, Indonesia, Colombia, and Thailand.
Id.
Using the lowest-GNI country (India) and highest-GNI country (Thailand) as “bookends,” Commerce turned to the World Bank’s 2009 World Development Report to identify every country with a GNI between the two, thus establishing the first and broadest basket of countries—44 that the agency deemed to be economically comparable to the PRC.
Id.; see also id.
at Attach. 2.
2.
Significant Producers of Comparable Merchandise
Commerce then determined which of these 44 countries were significant producers of comparable merchandise by identifying every country that exported any quantity of comparable merchandise (defined as exports under HTS 9609.10.00) between 2005 and 2007.
Id.
at 3. This produced a narrower basket of countries-30 that, in the agency’s view, satisfy the
dual requirements of 19 U.S.C. § 1677b(c)(4).
Id.; see also id.
at Attach. 2.
3.
Selecting and Sorting Reliable Wage Data
The next step in Commerce’s process was to assess which of these 30 countries had made available adequate data that could be averaged together to produce a single surrogate value for labor. Deciding which data would suffice for this purpose required the agency to make a series of choices. First, Commerce decided that it would rely on earnings or wages data reported to the International Labor Organization (“ILO”).
Id.
at 3. Second, Commerce decided to prefer industry-specific data over data that did not reflect the different incomes associated with different types of work.
Id.
Thus, noting that industry-specific wages and earnings data were available in “Chapter 5B: Wages in Manufacturing” of the ILO Yearbook of Labor Statistics, Commerce decided to use Chapter 5B as the wages and earnings data source.
Remand Results
at 13;
see also LABORSTA, ILO Labor Statistics Database 1998-2010, available at
http:// laborsta.ilo.org/applv8/data/c5e.html (last visited April 21, 2011).
When countries report any type of industry-specific data to the ILO, including the wages and earnings data of interest in this proceeding, they do so according to a uniform code known as the ISIC.
But there are different revisions of the ISIC code, and not all countries report industry-specific data under the same revision. Thus, Commerce faced a third choice, to determine which revision of the ISIC code it preferred. No countries had reported data under the most recent revision of the code, ISIC-Rev.4, but many countries had reported data under ISIC-Rev.3.
Wage Rate Memo
at 3. Moreover, many countries reporting under ISIC-Rev.3 had provided specific wages and earnings data for Division 36, which covers “[m]anufaeture of furniture; manufacturing not elsewhere classified (“n.e.c.”),” and encompasses pencil manufacturing, which is explicitly included in Class 3699 “Other manufacturing n.e.c.” ISIC Rev.3 Class 3699,
available at
http://unstats.un.org/unsd/cr/registry/ regcs.asp?Cl=2&Lg=l&Co=3699 (last visited April 21, 2011).
Id.
Commerce did not address in the
Wage Rate Memo
or in the
Remand Results
whether any countries reported data under ISIC-Rev.2, or if so, whether those countries had reported data for a 2-digit Division under that revision that might also include pencil manufacturing. This series of decisions left Commerce with 10 countries that, in the agency’s view, satisfied the dual requirements of 19 U.S.C. § 1677b(c)(4), and that had reported adequate industry-specific
wages and earnings data for use in this case.
Id.
at 4.
Finally, Commerce filtered and, in some cases, adjusted the data provided by these 10 countries, in order “to arrive at a single earnings or wage rate for each country.”
Id.
at 4-5. First, the agency prioritized the use of earnings data when available, and wages data when not.
Id.
at 5. Second, the agency only considered data with combined male and female coverage.
Id.
Third, the agency used earnings or wages data most contemporaneous to 2007; if it had to use data from an earlier year, the agency inflated it with the use of the relevant Consumer Price Index to approximate 2007 price levels.
Id.
Fourth, Commerce “selected from the following categories [of workers], in the following hierarchy: 1) wage earners; 2) employees; 3) salaried employees; and 4) total employment.”
Id.
Finally, Commerce chose hourly data where available; if a country reported data only on a daily, weekly, or monthly basis, the agency converted that data to an hourly figure assuming “8 working hours per day, 5.5 working days a week, and 24 working days per month.”
Id.
After massaging the data in accordance to the foregoing parameters, Commerce calculated a final simple average industry-specific wage rate of $0.97 per hour.
Id.
Well, it was almost final. Before issuing the Remand Results, Commerce noticed that three of the countries it had included in the final group of 10 (because they met the dual requirements of § 1677b(c)(4), and had provided all relevant data in the right ISIC code revision) were nonmarket economies.
Analysis Memorandum,
December 6, 2010 (AR 2753) at 3. Because the surrogate value for labor in an NME proceeding cannot be based on labor values from other nonmarket economies, Commerce reduced the final group to seven countries, with a simple average wage rate of $1.07/hour.
B. Shandong’s Contentions
Shandong’s primary argument is that Commerce should have only used wage data from India, rather than from a basket of countries, to value labor as a factor of production. (Pl.’s Cmts. 4-20.) Plaintiff argues that by valuing labor from a group of countries, while valuing all other factors of production from a single surrogate country, “Commerce has acted in an arbitrary and capricious manner.”
(Id.
at 6.) Plaintiff cites 19 C.F.R. § 351.408(c)(2) and claims that “Commerce has a clear policy of obtaining as many surrogate values from
a single
country as possible.”
(Id.
at 12 (emphasis in original).) What Plaintiff does not acknowledge, however, is that this regulation explicitly identifies labor as an exception to the “normal” approach. 19 C.F.R. § 351.408(c)(2) (“Except for labor, [Commerce] normally will value all factors in a single surrogate country.”)
While Shandong’s only explicit argument is that Commerce was required to use labor data from one country (India) rather than a group of countries, there are portions of its brief which might be read to argue in the alternative that even if using
a basket of countries to value labor was permitted, the omission of Indian data from such a basket was unlawful. For instance, Plaintiff suggests that the only reason Indian wage and earnings data from the ILO were not chosen by Commerce is that India had reported its data under ISIC-Rev.2, rather than ISIC-Rev.3. (Pl.’s Cmts. at 6.) Plaintiff argues that Commerce did not identify any reason why the Indian data was “intrinsically unreliable and arbitrary,” and asserts that Indian ILO data is at least as reliable as the data chosen by Commerce. (Id. at 6-7.) Plaintiff highlights Commerce’s data standards of “quality, specificity, and contemporaneity,” arguing that the agency failed to apply these tests “to the seven countries and India equally.”
(Id.
at 9.) Shandong points out that the Indian ILO data is available for a 3-digit Section (390), which is arguably more specific to the pencil industry than the 2-digit Divisions utilized by Commerce from the other countries.
(Id.
at 13.) Also, Shandong points out that the Indian ILO data comes from 2006, nearly contemporaneous with the period of review, whereas data from several of the countries used by Commerce were several years older, and required the use of inflators to approximate period of review levels.
(Id.
at 14.)
Shandong’s other major dispute with the Remand Results is on the issue of which countries constitute “significant producers of comparable merchandise” within the meaning of 19 U.S.C. § 1677b(c)(4)(B).
(Id.
at 20.) Plaintiff argues that Commerce’s interpretation of the term “significant producers” to include any country with
any quantity
of exports in the relevant time period is an impermissible construction of the statute. (Id. at 21.) Plaintiff states that for Commerce “to conclude that any country that registers exports of a product, no matter how commercially inconsequential, represents a ‘significant producer’ of that product ... simply cannot withstand any real scrutiny.” (Id. at 21.) Plaintiff points to the fact that the share of world trade held by many of the countries Commerce deemed “significant producers” is astonishingly minuscule, including, for instance, .00266% for Ukraine, .00356% for Ecuador, and .00756% for Egypt. (Id. at 22.) Plaintiff also notes that according to the statute’s legislative history, “the term ‘significant producer’ includes any country that is a significant net exporter,”
(quoting
Conference Report to the 1988 Omnibus Trade & Competitiveness Act, H.R.Rep. No. 100-576, at 590, 1988 U.S.C.C.A.N. 1547, 1623 (1988)), and points out that Commerce “has previously relied on countries being a net exporter in order to qualify as a significant producer,”
(citing Certain Norn-Fro-zen Apple Juice Concentrate from the People’s Republic of China,
69 Fed.Reg. 65, 148 (Nov. 10, 2004) and accompanying-issues and Decision Memo).
(Id.
at 24.)
C. Defendant’s Response
Defendant responds to Shandong’s first argument by maintaining that Commerce’s reliance on data from multiple countries to produce a surrogate value for labor is supported by substantial evidence on the record and is otherwise in accordance with law. (Def.’s Resp. to Remand Cmts. by Pl. Shandong Rongxin Import & Export Co. Ltd. (“Def.’s Resp.”) 5-12.) Defendant points to the record evidence, which demonstrates wide variability of wages among countries with relatively comparable gross national income, and notes that Commerce cited several characteristics which distinguish labor from other production inputs.
(Id.
at 7.) Defendant highlights that Shan-dong’s position that Commerce must use a single country to value labor is explicitly contradicted by statute, regulation, and CAFC case law.
(Id.
at 8-9
(citing
19
U.S.C. § 1677b(c)(4), 19 C.F.R. § 351.408(c)(2), and
Dorbest Ltd. v. United States,
604 F.3d 1363, 1372 (Fed.Cir. 2010)).) Defendant concludes by dismantling Shandong’s remaining arguments as to why Commerce should have exclusively relied on Indian wage data.
(Id.
at 9-12.) Picking up on Shandong’s possible argument that Indian labor data should have been included with the basket of countries Commerce utilized, Defendant merely reiterates that Commerce did not want to “mix and match data that can cover different industry groupings from different ISIC Revisions.”
(Id.
at 15 (internal quotation and bracketing omitted).)
The United States also defends Commerce’s view that a country is a “significant producer” if it exports any quantity of comparable merchandise in the relevant period.
(Id.
at 12-16.) Arguing under a
Chevron
framework, Defendant asserts that the term “significant producer” is ambiguous and undefined in the statute.
(Id.
at 13
(citing Chevron U.S.A., Inc. v. Natural Resources Defense Council, Inc.,
467 U.S. 837, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984)).) Defendant’s view is that the legislative history is unhelpful, because it only indicates that the term “significant producer
includes
” a net exporter, but offers no specific metric or definition.
(Id.
(emphasis added and internal quotation omitted).) Thus, moving to the second prong of
Chevron,
Defendant asserts that Commerce is entitled to substantial deference. Defendant explains that Commerce’s motivation for defining “significant producer” in this way was “to maximize the size of the ultimate basket of countries for data purposes while still accounting for this criterion.”
(Id.
at 14.) But, pointedly, Defendant does not explain how treating a country with
any
quantity of exports actually accounts for this criterion.
D. Analysis
The Court finds groundless Shandong’s argument that Commerce was obligated to utilize data from a single country to value labor. This argument is untenable in the face of a statute, agency regulation, and CAFC case law which all explicitly permit the agency to utilize data from multiple countries. 19 U.S.C. § 1677b(c)(4) (instructing Commerce to “utilize, to the extent possible, the prices or costs of factors of production in
one or more market economy countries
”) (emphasis added); 19 C.F.R. § 351.408(c)(2) (explicitly excepting labor from Commerce’s normal approach of valuing factors of production in a single surrogate country);
and Dorbest,
604 F.3d at 1372 (describing a possible “subset of ... countries” that would satisfy the dual requirements of § 1677b(c)(4)). Additionally, Commerce is well justified in averaging wage rates to produce a single surrogate value, with record evidence establishing both the existence of variation in wages among similarly economically developed countries, and the reasons for it.
(See Remand Results
at 10-11
(citing International Labor Organization, Global Wage Report: 2009 Update,
(2009) at 5, 7, and 10.)) Commerce is therefore acting reasonably when it treats labor differently from other factors of production. Accordingly, Commerce’s decision to utilize data from multiple countries is supported by substantial evidence on the record, and is otherwise in accordance with law.
The Court is less sanguine, however, about the reasons Commerce cites for excluding Indian labor data, which was reported under ISIC-Rev.2, from the group of countries ultimately providing the labor rate, all of which reported data under ISIC-Rev.3. While the agency has made clear that it prefers “to use data from a single ISIC revision to ensure con-
sistency of the industry category,” the Court finds Commerce’s justification for this preference lacking and inconsistent. The Indian wages and earnings data reported to the ILO appears to meet all other criteria identified by the agency, including quality, specificity, and contemporaneity. Indian ILO labor data was reported for a year close to the period of review — 2006—and was reported at a more specific 3-digit level of the ISIC than the 2-digit-level data relied on by Commerce. Also, India reported a combined earnings figure for men and women, in accordance with Commerce’s preference, and the agency does not dispute that the ISIC-Rev.2 Indian labor data includes the pencil industry. To dismiss such apparently valuable data without further explanation is unjustified. Moreover, refusing to use ISIC-Rev.2 data contradicts what the agency has repeatedly identified as a paramount interest: generating the broadest basket of countries possible to value labor. Commerce has cited the need for a broad basket of countries to justify using less contemporaneous data,
Remand Results
at 28, and to attempt to justify the inclusion of labor data from countries with minuscule amounts of exports, (Def.’s Resp. at 14.). The inconsistency with which Commerce has asserted the need for a broad basket of countries warrants a remand.
Commerce has broad discretion to determine which criteria it will use to sort and prioritize the data it uses in making its determination. The Court’s role is to ensure that Commerce’s sorting and prioritizing decisions are reasonable and consistently applied. In this case, the Court finds that most of Commerce’s sorting and prioritizing decisions are well justified, such as the decision to use earnings data if available, and wages data if not, and the choice only to utilize data reported for both sexes. The decision to insist that data be reported under a common ISIC revision, however, is not supported by substantial evidence on the record. On remand, if Commerce still wishes to omit all labor data that a qualifying country reported under ISIC-Rev.2, it must explain why the need for consistency across ISIC revisions predominates over the need for a broad basket of countries to value labor. Alternatively, if Commerce determines that the chief value is to have the broadest feasible basket of countries, Commerce is instructed to review which qualifying countries have reported data under a prior ISIC revision which satisfy the agency’s other requirements.
The final issue of concern for the Court is Commerce’s construal of the term “significant producer” to mean any country with any level of exports under the relevant HTS subheading.
See Remand Results
at 20. Commerce’s justification for this is pure speculation: the agency claims that “a country’s ability to export comparable merchandise is indicative of substantial production because it is likely producing merchandise at a level that surpasses its internal consumption.”
Id.
There is nothing on the record to support this theo
ry — no evidence about domestic pencil manufacturing in these countries, nor even any evidence about the level of imports of comparable merchandise for comparison purposes. While the inference that exports implies significant domestic production is somewhat palatable in the case of a country such as Indonesia, which had between $40 million and $53 million of exports in each year between 2005 and 2007, it is increasingly absurd and implausible as the quantity of exports in any given country approaches zero. In fact, many of the economically comparable countries that Commerce identified as “significant producers” had little or no exports in the years in question. For instance, Maldives had $67 of exports in 2005, and no exports in 2006 or 2007, and yet was considered a significant producer.
Wage Rate Memo
at Attachment 2. Similarly, Cape Verde and Swaziland were considered significant producers despite never exporting more than $159 and $218 of pencils, respectively, in any year between 2005 and 2007.
Id.
The idea that Guyana, which exported a total of $43 of pencils between 2005 and 2007, is a “significant producer” that manufactured $43 of pencils more than domestic consumption for that period is imaginative, but wholly unsupported. Ultimately, wage rates from Maldives, Cape Verde, Swaziland, Guyana and other countries with similarly low levels of exports were not included in Commerce’s final labor calculation because these countries failed to report relevant labor data to the ILO. That coincidence, however, does not cure the ailment arising from Commerce’s absurd construction of the term “significant producer.”
The term “significant” in 19 U.S.C. § 1677b(c)(4) is not statutorily defined, and is inherently ambiguous. Significant has among its definitions “having meaning” and “having or likely to have influence or effect.”
Webster’s Third New International Dictionary
2116 (1993). Because the term is ambiguous, Commerce is entitled to broad deference in its construal of the term, and the Court’s question “is whether [Commerce’s] answer is based on a permissible construction of the statute.”
Chevron,
467 U.S. at 842-43, 104 S.Ct. 2778. Commerce’s interpretation of “significant” encompasses countries which almost certainly have no domestic production — at least not any meaningful production, capable of having influence or effect — and is therefore an impermissible construction of 19 U.S.C. § 1677b(c)(4). On remand, if Commerce wishes to continue to rely on export statistics as a proxy for significant production, it must include some additional mechanism to ensure that it does not propagate the fiction that countries with a few dollars of exports are engaged in significant production. Alternatively, Commerce is free to adopt an altogether different approach to identifying significant production.
Conclusion
In conclusion, for the reasons stated herein, the Court affirms Commerce’s decision to rely on data from multiple countries to value labor as a factor of production pursuant to 19 U.S.C. § 1677b(c), and it is hereby
ORDERED that this case is remanded to Commerce, and it is further
ORDERED that Commerce must reevaluate, in accordance with this opinion, the decision to omit labor data simply because it was reported under a previous revision of the ISIC, and it is further
ORDERED that Commerce must modify, in accordance with this opinion, the way in which it determines whether a country is a significant producer of comparable
merchandise within the meaning of 19 U.S.C. § 1677b(c)(4), and it is further
ORDERED that Commerce must provide the results of its redetermination pursuant to remand to the Court no later than June 21, 2011, and it is further
ORDERED that Plaintiff shall file its comments on the remand results no later than June 28, 2011, and it is further
ORDERED that Defendant shall file any responses to Plaintiffs comments no later than July 6, 2011.