OPINION

POGUE, Chief Judge:

This case returns to court following remand by Camau Frozen Seafood Processing Import Export Corp. v. United States, ___ CIT ___, 880 F.Supp.2d 1348 (2012) ("Camau I"). Camau I reviewed challenges to the final results of the fifth administrative review ("AR") of the antidumping duty order covering certain frozen warmwater shrimp from the Socialist Republic of Vietnam ("Vietnam").2 Id. at 1351. Specifically, Camau I rejected a facial challenge to Commerce's use, in the fifth AR, of its New Labor Methodology,3 but remanded the Final Results for Commerce to further explain or reconsider its determination to value labor solely on the basis of data from the Bangladesh Bureau of Statistics ("BBS") in light of Commerce's prior surrogate labor policy and the apparent discrepancy between the Bangladeshi labor data and the Philippine labor data on the record. Id. at 1358-61. In the Final Results of Redetermination Pursuant to Court Remand, A-552-802, ARP 09-10 (Apr. 12, 2013), ECF No. 90 ("Remand Results"), Commerce determined that it would continue to value labor solely on the basis of the BBS data.

For the reasons that follow, the court will order a second remand for Commerce to further explain or reconsider its determination to value labor in this case solely on the basis of the BBS data.

The court has jurisdiction pursuant to § 516A(a)(2)(B)(iii) of the Tariff Act of 1930, as amended, 19 U.S.C. § 1516a(a)(2)(B)(iii) (2006)4 and 28 U.S.C. § 1581(c) (2006).

STANDARD OF REVIEW

"The court will sustain the Department's determination upon remand if it complies with the court's remand order, is supported by substantial evidence on the record, and is otherwise in accordance with law." Jinan Yipin Corp. v. United States, ___ CIT ___, 637 F.Supp.2d 1183, 1185 (2009) (citing 19 U.S.C. § 1516a(b)(1)(B)(i)).

DISCUSSION5

Prior to adoption of the New Labor Methodology, Commerce used multi-country averaging to value labor because "wage data from a single surrogate country does not constitute the best available information for purposes of valuing the labor input due to the variability that exists between wages and GNI.... As a result, we find reliance on wage data from a single surrogate country to be unreliable and arbitrary."6 When Commerce adopted the New Labor Methodology, it did not repudiate this reasoning. Rather, Commerce acknowledged in the New Labor Methodology that "[d]ue to the variability in wage rates among economically comparable [market economy countries], the Department has tried to include wage data from as many countries as possible that were also economically comparable to the [nonmarket economy country ("NME")] and significant producers of comparable merchandise...." New Labor Methodology, 76 Fed.Reg. at 36,093; see also Camau I, 880 F.Supp.2d at 1358-59. But, based on its experience in light of Dorbest Ltd. v. United States, 604 F.3d 1363 (Fed.Cir.2010) ("Dorbest IV") and Shandong Rongxin Import & Export Co. v. United States, ___ CIT ___, 774 F.Supp.2d 1307 (2011),7 Commerce concluded that "the base for an average wage calculation would be so limited that there would be little, if any, benefit to relying on an average of wages from multiple countries for purposes of minimizing the variability that occurs in wages across countries." New Labor Methodology, 76 Fed. Reg. at 36,093. Camau I held this to be a reasonable basis for Commerce's change in policy, 880 F.Supp.2d at 1358; therefore, the decision to change the labor valuation policy is not before the court on review of the Remand Results. Nonetheless, insofar as Commerce maintains that (1) valuing labor based on a single surrogate country may be distortive given the variability in wage rates among countries that Commerce considers to be economically comparable and (2) the variability in wage rates corresponds to variability in GNI, the record in this case presents the possibility of just such a distortion.

As noted in Camau I, Commerce considered two wage rate values in the Final Results: one from Bangladesh, based on the BBS data, and one from the Philippines, based on Chapter 5B of the International Labor Organization Yearbook of Labour Statistics ("ILO Chapter 5B"). Id. at 1359-60 & n. 12. The wage rate value for the Philippines is several orders of magnitude larger than the wage rate value for Bangladesh. See Id. at 1360 (comparing GNI and wage rates of the Philippines and Bangladesh). In light of Commerce's prior policy and findings, it comes as no surprise that the Philippine GNI is also several times larger than the Bangladeshi GNI. Id. On these facts, Commerce's non-repudiated prior reasoning suggests that a single surrogate country value for labor could introduce distortion.8 While an averaging system that eliminates such distortion may not be possible, that fact alone is not a reasoned explanation for Commerce's choice between the two datasets. Therefore, Camau I remanded this issue for an explanation of why, in light of Commerce's prior reasoning and the record evidence in this case, valuing labor solely on the basis of the BBS data was reasonable and the best available information. Id.

Commerce justifies its decision in the Remand Results by invoking its policy of valuing all surrogate values from a single surrogate country when possible. Remand Results at 7-8. Commerce contends that using a single surrogate country to value all FOPs "better reflects the trade-off between labor costs and other factors' costs, including capital, based on their relative prices." Id. at 8. This is the only affirmative basis Commerce offers to support its choice of the Bangladeshi data. Thus, Commerce argues that its policy of favoring a single surrogate country to value all FOPs, and the reasoning supporting that policy, is sufficient to value labor solely on the basis of the BBS data in this case.

This basis alone, however, is not sufficient to address the remand order in Camau I. Commerce's policy of valuing all factors of production from a single surrogate country when possible, see 19 C.F.R. § 351.408(c)(2) (2011), may be reasonable because, among other reasons, it reduces surrogate value distortions introduced by out-of-market prices, see Clearon Corp. v. United States, Slip Op. 13-22, 2013 WL 646390, at *6 (CIT Feb. 20, 2013); nonetheless, Commerce has the statutory authority to use multiple surrogate countries, 19 U.S.C. § 1677b(c)(1), and has invoked that authority when it deemed such to be appropriate — specifically as part of its prior labor valuation methodology, see, e.g., Grobest & I-Mei Indus. (Viet.) Co. v. United States, ___ CIT ___, 815 F.Supp.2d 1342, 1356-60 (2012) (affirming Commerce's decision to use multi-country averaging for surrogate labor valuation); Peer Bearing Co.-Changshan v. United States, ___ CIT ___, 804 F.Supp.2d 1337, 1353 (2011) (noting Commerce's use of Indian and Thai data for different surrogate values in the same review). Therefore, it is not sufficient for Commerce to cite the policy of using a single surrogate country where, as here, there is reason to believe that the primary surrogate country may not provide the best available information for a particular FOP.

Case law repeatedly emphasizes that "use of a single surrogate country is justified when ... all other factors are fairly equal...." Clearon Corp., 2013 WL 646390, at *6 (internal quotation marks omitted); Peer Bearing, ___ CIT at ___, 804 F.Supp.2d at 1353 ("[T]he preference for use of data from a single surrogate country could support a choice of data as the best available information where the other available data `upon a fair comparison, are otherwise seen to be fairly equal....'") (quoting Peer Bearing Co.-Changshan v. United States, ___ CIT ___, 752 F.Supp.2d 1353, 1373 (2011)) (second alteration in original). In light of Commerce's prior reasoning with regard to labor values, however, the evidence on the record in this case cannot, without more, be considered fairly equal.9 Thus, because there is reason to doubt the primary surrogate country value, Commerce must address the conflicting evidence on the record that may counsel against the policy of valuing all FOPs from the primary surrogate country. Not addressing the conflicting evidence on the record, as noted in Camau I, fails the substantial evidence test because it does not take into account record evidence contrary to Commerce's determination. See Universal Camera Corp. v. NLRB, 340 U.S. 474, 488, 71 S.Ct. 456, 95 L.Ed. 456 (1951).

Commerce has not, however, addressed the conflicting evidence on the record in the Remand Results.10 While "the possibility of drawing two inconsistent conclusions from the evidence does not prevent an administrative agency's finding from being support by substantial evidence," Consolo v. Fed. Maritime Comm'n, 383 U.S. 607, 620, 86 S.Ct. 1018, 16 L.Ed.2d 131 (1966), Commerce must, nonetheless, provide a reasonable basis for its determination, see Nippon Steel Corp. v. United States, 458 F.3d 1345, 1351-52 (Fed.Cir.2006); Amanda Foods (Viet.) Ltd. v. United States, ___ CIT ___, 647 F.Supp.2d 1368, 1378-79 (2009). Instead, Commerce argues that the Bangladeshi data and the Philippine data are collected at different levels of aggregation; therefore, Commerce asserts that the two data sets are not comparable and a disparity in wage rates cannot be deduced from the data. Remand Results at 8-9. Commerce's argument is unpersuasive.

First, Commerce provides no explanation for why the different levels of aggregation render the data incomparable. Different levels of aggregation alone do not, necessarily, prevent two datasets from being compared. What is of consequence is the particular factors that make the datasets similar enough to compare or too different to compare — for example, the relative levels of aggregation, the relationship between the levels of aggregation, and the purpose of the comparison. In short, Commerce must provide some reason to justify its determination that the datasets are too different to compare, see Amanda Foods, ___ CIT at ___, 647 F.Supp.2d at 1378-79, and level of aggregation is a description not a reason.

Second, Commerce's treatment of the ILO data in other circumstances suggests that it may, in fact, be comparable with the BBS data. It is Commerce's default policy to use ILO data when valuing labor.11 Commerce considers data reported at an International Standard Industrial Classification ("ISIC") level representative of the industry in question to be industry specific. See New Labor Methodology, 76 Fed. Reg. at 36,094 & nn. 10, 11; Surrogate Values for the Preliminary Results, A-552-802, APR 09-10 (Feb. 28, 2011), Admin. R. Pt. 1 Pub. Doc. 144 ("Surrogate Value Mem.") at 7. Prior to adopting the standards from the New Labor Methodology in this case, Commerce determined that ISIC-Revision 3, sub-classification 15, described as "manufacture of food products and beverages," was industry specific because it included "processing and preservation of fish and fishery products." Id. The fact that Commerce considers the ILO data to be industry specific and would otherwise employ the ILO data but for the particular facts of this case — i.e., no ILO data for Bangladesh and an alternative industry-specific dataset — suggests that the ILO data and the BBS are comparable despite the different levels of aggregation. That is, the data sets are a least comparable enough in Commerce's view for them to be theoretically interchangeable for the purpose of valuing labor.12

Thus, Commerce's reasoning in the Remand Results remains an insufficient explanation, and the court remains unable to affirm Commerce's determination in the Final Results. Commerce's policy of valuing all surrogate values on the basis of the primary surrogate country is a reasonable choice insofar as there is no reason to believe that a value from the primary surrogate country would be distortive or inaccurate. Record evidence in this case continues to raise such a possibility, and Commerce has not addressed that evidence in the Remand Results. Furthermore, Commerce's attempt to avoid the troubling disparities between the surrogate values for labor by suggesting that the datasets are not comparable is unpersuasive. Commerce provides no justification for its conclusion of incomparability other than the different levels of aggregation — a distinction that, absent further explanation, is not a meaningful difference.

CONCLUSION

In light of the foregoing, the Final Results are again remanded to Commerce for further explanation or reconsideration of the surrogate value for labor consistent with this opinion and Camau I. Commerce shall have until September 30, 2013, to complete and file its remand redetermination. Plaintiffs and Defendant-Intervenors shall have until October 15, 2013, to file comments. Plaintiffs, Defendant, and Defendant-Intervenors shall have until October 29, 2013, to file any reply.

It is SO ORDERED.