Toscelik Profil ve Sac Endustrisi A.S. v. United States

Slip Op. 14 - 126

UNITED STATES COURT OF INTERNATIONAL TRADE

: TOSCELIK PROFIL VE SAC ENDUSTRISI A.S., : : Plaintiff, : : v. : Before: R. Kenton Musgrave, Senior Judge : UNITED STATES, : Court No. 13-00371 : Defendant, : : and : : WHEATLAND TUBE COMPANY and : UNITED STATES STEEL CORPORATION, : : Defendant-Intervenors. : :

OPINION AND ORDER

[On USCIT Rule 56.2 motion, administrative review determination remanded.]

Dated: October 29, 2014

David L. Simon, Law Offices of David L. Simon, of Washington DC, for the plaintiff.

L. Misha Preheim, Senior Trial Counsel, Commercial Litigation Branch, Civil Division, U.S. Department of Justice, of Washington DC, for the defendant. With him on the brief were Stuart Delery, Assistant Attorney General, Jeanne E. Davidson, Director, and Franklin E. White, Jr., Assistant Director. Of Counsel on the brief was David P. Lyons, Attorney, Office of the Chief Counsel for Import Administration, U.S. Department of Commerce.

Gilbert B. Kaplan and Jennifer D. Jones, King & Spaulding LLP, of Washington DC, for the defendant-intervenor Wheatland Tube Company.

Jeffrey D. Gerrish and Robert E. Lighthizer, Skadden Arps Slate Meagher & Flom, LLP, of Washington DC, for the defendant-intervenor United States Steel Corporation.

Court No. 13-00371 Page 2

Musgrave, Senior Judge: The plaintiff Toscelik Profil ve Sac Endustrisi A.S. (“Toscelik”), a producer of subject merchandise for the Turkish domestic and export markets, appeals from Circular Welded Carbon Steel Pipes And Tubes From Turkey: Final Results of Countervailing Duty Administrative Review; Calendar Year 2011, 78 Fed. Reg. 64916 (Oct. 30, 2013) (“Final Results”), see PDoc 201, including its issues and decision memorandum (Oct. 23, 2013) (“I&D Memo”), PDoc 202, and accompanying set of calculations (“Final Calcs”), PDoc 205, CDoc 212. Toscelik contests two benchmarks Commerce constructed to measure the benefits Toscelik received, respectively, from two parcels of land it acquired from a Turkish governmental entity. Toscelik received one parcel under a grant in 2008 and purchased the adjacent parcel in 2010. I&D Memo at 10, 12. For each parcel, the benefit is calculated as the difference between the price paid and the benchmark value. 19 C.F.R. §351.511(a)(2).

Background

Toscelik was a mandatory respondent for the calendar year 2011 administrative review at bar of the countervailing duty (“CVD”) order. See Initiation of Antidumping and Countervailing Duty Administrative Reviews and Request for Revocation in Part, 77 Fed. Reg. 25401, 25403 (Apr. 30, 2012). In its questionnaire response (“QR”) of July 29, 2012, PDocs 34, 35, CDocs 7-27, Toscelik reported, inter alia, that in 2008 it had received a grant of free land under Turkish Law 5084 in the Organized Industrial Zone (OIZ) of Osmaniye, Turkey. CDocs 29-30. Toscelik had previously reported this grant in the prior administrative review covering year 2010.

In its questionnaire responses, Toscelik also reported having purchased an adjacent parcel from the Osmaniye OIZ in 2010. Toscelik Supplemental Questionnaire Response (“SQR”) Court No. 13-00371 Page 3 (Nov. 8, 2012), PDoc137, CDocs142-146, at 1-2. Commerce initiated an inquiry to determine whether the price Toscelik paid for this 2010 parcel was less than adequate remuneration (“LTAR”). New Subsidy Allegation Memorandum (Oct. 9, 2012), PDoc 121, CDoc139.

In its SQR of November 8, Toscelik detailed the location, size, and price for Toscelik’s purchases of comparable properties from private-sector sellers and the site in Osmaniye of the 2008 and 2010 parcels. PDoc137, CDocs142-146, at 1-2 & Exhs. 2, 4, 5. The Government of Turkey (“GOT”) also submitted a QR. Response of GOT (Jul. 30, 2012), PDocs 36-85. Exhibits 24 and 25 thereto are respectively an English translation of Turkish Law 5084 (Feb. 6, 2004), the law under which Toscelik acquired the two parcels, and a list of the 49 least-developed provinces for which the regional benefits under Law 5084 are available, which includes Osmaniye Province. PDocs 63-64.

Wheatland also presented a “submission of factual information” (“SFI”) to Commerce. Wheatland SFI (Aug. 20, 2012), PDocs 100-102. An exhibit thereto contains advertisements for the sale of real estate in Turkey, showing the location, size, and value of properties in various locations as of August 18, 2012. See id.

Commerce conducted an on-site verification in February 2013, and its verification report for Toscelik included details of Toscelik’s acquisition of the 2008 and 2010 parcels. Verification Report (Mar. 15, 2013), PDoc 177, CDoc 203. Commerce also placed on the record a memo regarding the value of land in Turkey: “Placement of Land Price Information on Record of Review” (Mar. 26, 2012), PDoc 183. This was the same information that Commerce had used in the 2010 review. See Verification Report Exh. 18 (Feb. 20, 2013), CDoc 198. Court No. 13-00371 Page 4

In general, Commerce allocates non-recurring subsidies pursuant to 19 C.F.R. §351.524 over a period corresponding to the “average useful life” of the renewable physical assets used to produce the subject merchandise (“AUL”). Commerce in this proceeding (again) found the AUL to be 15 years pursuant to 19 C.F.R. §351.524(d)(2) and the U.S. Internal Revenue Service’s 1977 Class Life Asset Depreciation Range System.1 Commerce further explained that for non-recurring subsidies, it would apply the “0.5 percent expense test” described in 19 C.F.R. §351.524(b)(2) by comparing the amount of subsidies approved under a given program in a particular year to relevant sales (e.g., total sales or total export sales) for the same year and allocate the amount of subsidies “to the year of receipt rather than allocate[ ] over the AUL period” if the amount of the subsidies is less than 0.5 percent of the relevant sales. I&D Memo at 3.

For the preliminary results, Commerce again examined the relevant Turkish law and, as in the prior 2010 review, when considering the benefit from the 2008 parcel grant Commerce used the same benchmark from that prior review -- a weighted average of certain land values -- to value both the 2008 parcel as well as the 2010 parcel.2 Toscelik accepted the preliminary results as issued

1 I&D Memo at 3, referencing U.S. Internal Revenue Service Publication 946 (2008), “How to Depreciate Property” at Table B-2: Table of Class Lives and Recovery Periods. 2 Circular Welded Carbon Steel Pipes and Tubes from Turkey: Preliminary Results of Countervailing Duty Administrative Review; Calendar Year 2011, 78 Fed. Reg. 21107 (May 9, 2013), PDoc 185 (“Preliminary Results”), with its accompanying preliminary decision memorandum (“Prelim. Decision Memo”) dated Apr. 2, 2013, PDoc 180, and accompanying calculations, PDoc 186, CDoc 207. In the Final Results, Commerce describes Turkish Law 5084 as concerning the allocation of “free” land and the purchase of LTAR land for the purpose of encouraging the economic development of OIDs under Turkish control in order “to reduce inter-regional disparities and to increase employment in provinces where the development is relatively low”. I&D Memo at 10. Also in the Final Results, Commerce “continue[d] to find that the allocation of free land to Toscelik [in the “organized industrial zone” (“OID”) of Osmaniye] by the OIZ authority constitutes (continued...)

Court No. 13-00371 Page 5 and provided no administrative case brief. Wheatland did file, and its case brief argued Commerce should use additional land valuation information provided by Wheatland in its SFI. PDoc 191, CDoc 209. Toscelik’s rebuttal, directed at the issues raised by Wheatland’s case brief,3 opposed changing the benchmark. PDoc 192.

In the Final Results, Commerce stated that it “continue[d] to rely upon the land benchmark data used in the prior review.” I&D Memo at 11. But, Commerce relied on those 2010 prior review benchmark data only in part, apparently, because Commerce “decided . . . to build a more robust data set” by “add[ing] . . . twelve additional data points”. Id. at 28. Further towards revision, Commerce also agreed with the domestic industry that the benchmark price is “normally” derived from a simple average of the reference land prices available in the record,4 and that

[g]iven the lack of sufficient detail regarding the characteristics of the land involved in the transactions underlying the benchmark data -- in particular, the extent to which the composition of our reference data set reflect the broader market, e.g., whether the proportion of large/small tracts in the benchmark data compares to the proportion of large/small tracts throughout Turkey--we have no basis to assume that any one parcel

2 (...continued) a financial contribution” under 19 U.S.C. §1677(5)(D)(ii) and a benefit under 19 U.S.C. §1677(5)(E)(iv), and that Toscelik’s LTAR land purchase conferred a financial contribution within the meaning of 19 U.S.C. §1677(5)(D)(iii). I&D Memo at 10-12. Commerce also found that the subsidy program was “regionally specific” under 19 U.S.C. §1677(5A)(D)(iv) since it was “limited to companies located in the 49 eligible provinces”. These facts are not in dispute. 3 See 19 C.F.R. §351.309(d)(2) (“[t]he rebuttal brief may respond only to arguments raised in case briefs and should identify the arguments to which it is responding”). 4 I&D Memo at 28, referencing Certain Steel Wire Garment Hangers From the Socialist Republic of Vietnam: Preliminary Affirmative Countervailing Duty Determination and Alignment of Final Countervailing Duty Determination with Final Antidumping Duty Determination, 77 Fed. Reg. 32930 (Jun. 4, 2012), at “Land Benchmark,” unchanged in the Final Affirmative Countervailing Duty Determination and Final Affirmative Critical Circumstances Determination, 77 Fed. Reg. 75973 (Dec. 26, 2012), and accompanying issues and decision memorandum at 6.

Court No. 13-00371 Page 6

of land among the reference set is more representative than any other parcel for the purpose of deriving a market price by which to determine adequate remuneration for the land in question. Moreover, obtaining more detailed information beyond the general comparability factors such as land-use classification would be impracticable for the Department. Given these inherent limitations, for these final results, we are applying a simple average, which gives all benchmark data on the record equal weight rather than weight based on a factor (or factors) which, in this case, have not been demonstrated to be relevant for an appropriate benchmark price.

Id. at 28-29. Thus, for the Final Results Commerce used a “revised” (from the prior review) benchmark for the 2008 parcel and constructed a different benchmark for the 2010 parcel; both benchmarks used simple averaging in the calculation. Final Calcs, PDoc 205, CDoc 212.

After publication of the Final Results, Toscelik initiated suit, invoking the jurisdiction of 28 U.S.C. §1581(c) via 19 U.S.C. §1516a(a)(2).

Standard of Review

The court will not uphold an agency determination that is unsupported by substantial evidence or otherwise not in accordance with law. 19 U.S.C. §1516a(b)(1)(B)(i). An agency determination is presumed to be correct, and the burden of proving otherwise before the court rests upon the challenging party. 28 U.S.C. §2639(a)(1).

Discussion

In 1998, Commerce announced that it would amortize a non-recurring benefit over its “average useful life” (“AUL”), calculated by the formula provided in 19 C.F.R. §351.524(d). Countervailing Duties, 63 Fed. Reg. 65348, 65395-97 (Nov. 25, 1998). Amortization attributes “the” benefit over each successive administrative POR. See Norsk Hydro Canada, Inc. v. United States, 472 F.3d 1347, 1363 (Fed. Cir. 2006) (“[a]mortization of a non-recurring subsidy is not inconsistent with preserving the integrity of separate PORs; it simply reflects that a non-recurring Court No. 13-00371 Page 7 subsidy received in one POR may provide a ‘benefit’ in other PORs”). The variables of the regulatory formula are the AUL, the discount rate, and the amount of “the” benefit to be amortized pursuant to the formula. See, e.g., Final Calcs, CDoc 212.

I. The Benchmark for the 2008 Parcel

Pointing out that Commerce appears to have conducted “hundreds” of CVD administrative reviews and does not cite a single case in which it has ever amended a benefit-stream calculation in “mid-amortization” but has in fact denied a number of requests to do so,5 Pl’s Reply at 1-2, Toscelik argues Commerce’s revision to the benchmark for the 2008 parcel was unlawful and against the principle (not to mention the principal) of amortization. See generally Pl’s 56.2 Br.

The government cites Notice of Final Affirmative CVD Determination: Certain Cold-Rolled Carbon Steel Flat Products From the Republic of Korea, 67 Fed. Reg. 62102 (Oct. 3, 2002), and accompanying final issues and decision memorandum at comment 9, as an instance where Commerce changed a benchmark. Def’s Resp. at 13. The case does not appear analogous, however, since Commerce therein simply changed a benchmark between the preliminary determination and the final determination of the same segment of the proceeding. That is procedurally distinguishable from recalculating the value of a benefit upon which Commerce has already made a final determination including a final determination as to its amortization schedule.

Wheatland argues Toscelik “inappropriately conflates the legal standards governing the selection and establishment of the AUL with the legal standards governing Commerce’s

5 See, e.g., Magnola Metallurgy, Inc. v. United States, 508 F.3d 1349, 1356 (Fed. Cir. 2007); Norsk Hydro Canada, supra; Certain Carbon Steel Products from Sweden; Final Results of Countervailing Duty Administrative Review, 62 Fed. Reg. 16549, 1649-50 (Apr. 7, 1997) (“Steel from Sweden”).

Court No. 13-00371 Page 8 benchmark determination.” Def-Int’s Resp. at 21. Wheatland claims that the selection of an AUL is different from selection of other elements of the benefit calculation because the AUL involves time while the other elements determine the size of the benefit. Id. at 21-22. That seems to be a distinction without a difference, however: the benefit calculation is a formula; each element of the formula is determined in the initial analysis of the subsidy program; and there is nothing in the regulation to suggest that any of the terms of the formula, once determined, is ever subject to change, since a change in terms is antithetical to the principle of amortization. Steel from Sweden rather appears to uphold that principle. See 62 Fed. Reg. at 1649-50 (“if a subsidy has already been countervailed based on an allocation period established in an earlier segment of the proceeding, it does not appear reasonable or practicable to reallocate that subsidy over a different period of time”).

Referencing Iron-Metal Castings From India; Final Results of CVD Administrative Review, 62 Fed. Reg. 32297 (Jun. 13, 1997), Wheatland attempts to persuade that Commerce’s selection of a different discount rate in that POR than the rate used in earlier review(s) equates to a change to the calculation of a benefit stream. The argument is not persuasive. The Castings From India benefit pertains to short-term export credits that are expensed in the year received, not amortized over a period of time, and “[a]mortization of a non-recurring subsidy is not inconsistent with preserving the integrity of separate PORs; it simply reflects that a non-recurring subsidy received in one POR may provide a ‘benefit’ in other PORs.” Norsk, 508 F.3d at 1363.

Commerce and Wheatland acknowledge that the benefit of the 2008 parcel was conferred upon Toscelik at a certain and particular point in time,6 but they appear to contest the

6 See, e.g., Def’s Resp. at 2-3; Def-Int’s Resp. at 23.

Court No. 13-00371 Page 9 propriety of “fixing” its amortization over time. That is, they implicitly appear to argue for discrete and severable benefit periods and adjustable pro-ration or benefits within each period. Commerce, for example, justifies its determination to revise the benchmark for the 2008 parcel on the ground that each administrative review is a separate segment of proceedings, with its own unique facts,7 and it simply “corrected” its land benchmark averaging methodology to align with its predominant methodology in deriving land benchmarks. Def’s Resp. at 11, referencing I&D Memo at 28-29.

This contention amounts to post hoc rationale with respect to changing the benchmark for the 2008 parcel. Regardless, the selection of “correct” methodology for the allocation of “the” benefit is appropriate at the outset of its selection -- which is rather the whole point of amortization and 19 C.F.R. §351.524. The same is true with respect to valuing the LTAR benefit conferred by Toscelik’s purchase of the 2010 parcel: yearly revaluation of the benefit would be inconsistent with the regulatory definition of the benefit of “the” grant. 19 C.F.R. §351.505(a) provides: “In the case of a grant, a benefit exists in the amount of the grant.” The use of the singular means that a grant has only a singular benefit amount, and in the absence of clear indication to the contrary, a singular benefit amount is not variable and subject to recalculation in successive reviews. See also 19 C.F.R. §351.511(c) (“[i]n the case of the provision of infrastructure, the Secretary will normally treat the benefit as non-recurring and will allocate the benefit to a particular year in accordance with § 351.524(d)”). Had Commerce intended otherwise, there would have been no reason to create a multi-year benefit-stream formula.

7 See, e.g., Shandong Huarong Machinery Co. v. United States, 29 CIT 484, 491 (2005).

Court No. 13-00371 Page 10

Having thus established the benchmark for the 2008 parcel and the benchmark’s AUL amortization schedule in the prior review, use of that benchmark and schedule became final and unappealable.8 As a “tribunal” of this matter, Commerce thereby established the law of the case, which must be abided,9 unless, in this subsequent review, it can be demonstrated that Commerce’s prior determination is “clearly erroneous and would work a manifest injustice.” See Arizona v. California, 460 U.S. 605, 618 n.8 (1983) (citation omitted). No party has done so in this case. Consistent with its “obligation to examine the record in each individual segment of a proceeding”, e.g., Def’s Resp. at 9, Commerce does not argue, for example, that there has been any significant change in the “U.S. Internal Revenue Service’s 1977 Class Life Asset Depreciation Range System” since the prior administrative review that would require revision of the relevant AUL. And the fact that “Wheatland put twelve land parcel values on the record of this segment that were not on the record of the prior review”, Def’s Resp. at 11, does not render Commerce’s original calculation of the benchmark for the 2008 parcel inaccurate, let alone clearly erroneous or manifestly unjust. And while Wheatland’s administrative case brief argued that a weighted average can result in distortions

8 Cf. Magnola, 508 F.3d at 1353, quoting Commerce: “once a determination has been made regarding whether a non-recurring subsidy was specific (or not) at the time of bestowal, then that finding holds for the duration of the subsidy benefit barring any new facts or evidence pertaining to the circumstances of the subsidy’s bestowal.” 9 Indeed, “[g]iven that Commerce undertakes annual reviews, it would be duplicative and wasteful for later reviews to revisit matters subject to review in prior PORs. Revisiting issues that were resolved in prior review proceedings would impair the finality of any one annual review, potentially prolonging a CVD dispute far beyond the year to which it relates.” Norsk, 472 F.3d at 1361.

Court No. 13-00371 Page 11 and that Commerce’s “standard” practice is to use a simple average,10 Commerce’s consideration of the issue simply amounted to “agree[ment] with Petitioners that the Department normally derives the benchmark price from a simple average of the reference land prices available in the record”, I&D Memo at 28, not proclamation of clear error or manifest injustice.

Due to Commerce’s rationale for the acceptance of the petitioners’ figures in the revised benchmark for the 2008 parcel, Toscelik also complains Commerce ignored “abundant” evidence on the record of Toscelik’s actual purchase of comparable properties located reasonably close to Osmaniye province as well as record evidence of actual land-transaction values in Osmaniye (in contrast to mere offers for sale),11 and that was not in accordance with regulation. Further, Toscelik also complains that the parcels from Istanbul and Yalova are “outliers” that should not have been included in the revised 2008 parcel benchmark data set, not merely because of their “extreme”

10 Wheatland Case Brief, PDoc 191, at 3-6, referencing preliminary and final determinations in Drawn Stainless Steel Sinks From the People’s Republic of China: Final Affirmative Countervailing Duty Determination, 78 Fed. Reg. 13017 (Feb. 26, 2013) and accompanying issues and decision memorandum at “E. Benchmarks for Land” section; Certain Steel Wire Garment Hangers From the Socialist Republic of Vietnam: Final Affirmative Countervailing Duty Determination and Final Affirmative Critical Circumstances Determination, 77 Fed. Reg. 75973 (Dec. 26,2012) and accompanying issues and decision memorandum at 6-7; Crystalline Silicon Photovoltaic Cells, Whether or Not Assembled Into Modules, From the People’s Republic of China: Final Affirmative Countervailing Duty Determination and Final Affirmative Critical Circumstances Determination, 77 Fed. Reg. 63788 (Oct. 17, 2012) and accompanying issues and decision memorandum at 6, 13-14 and cmt. 11; Polyethylene Retail Carrier Bags from the Socialist Republic of Vietnam: Final Affirmative Countervailing Duty Determination, 75 Fed. Reg. 16428 (Apr. 1,2010) and accompanying issues and decision memorandum at footnote 23 and cmt. 9. 11 E.g., Pl’s 56.2 Br. at 11, 21-25, referencing, inter alia, Toscelik SQR, Nov. 8, 2012, at Exhs. 1-5, PDoc137, CDocs 142-146. On the issue of whether exhaustion of administrative remedy should apply with respect to administrative consideration of this data as argued by Wheatland (in which the government did not join), the court finds the doctrine inapplicable in this instance for the reasons stated in Toscelik’s reply brief, and as elucidated at oral argument. See, e.g., Pl’s Reply at 14-20.

Court No. 13-00371 Page 12 (as compared to the other benchmark data) values but because they are located in more highly developed areas of Turkey and not in the less developed areas covered by Turkish Law 5084.

Apart from the “propriety” of revising the 2008 parcel benchmark in the first instance, see supra, the court agrees that the record does not reflect Commerce’s consideration on whether these parcels are appropriate for inclusion therein even if revision would be considered proper. Cf. Zhaoqing New Zhongya Aluminum Co., Ltd. v. United States, 37 CIT ___, ___, 929 F. Supp. 2d 1324, 1328 (2013) (“Zhongya Aluminum”) (“the amenities currently advertised as available in the general region have absolutely no bearing on the condition of the specific plot as it existed when Zhongya assumed the land use rights in 2006”) (italics added). If Commerce reaches this issue on remand, Commerce must explain why the Turkish government’s schematization is an inappropriate basis on which to apply the comparability analysis required by 19 C.F.R. §351.511(a)(2), address the aspect of the Istanbul and Yalova properties’ listings as “agricultural land” in the appropriateness of their inclusion, see PDoc 99, and also address Toscelik’s proffered evidence and claim of “relatively nearby and manifestly similar properties.” See Pl’s Reply at 10, referencing CDocs 142-146.

For the above reasons,12 the matter must be remanded with instruction to either restore use of the benchmark and amortization schedule established for the 2008 parcel or explain why its use would be clearly erroneous and would work a manifest injustice. Any revision to the 2008 benchmark, if properly explained, must also be accompanied by adequate consideration of “similarity . . . and other factors affecting comparability” in accordance with 19 C.F.R. §351.511(a)(2).

12 The court has considered the parties’ other arguments on these issues but finds in the interest of brevity that their discussion would not further this opinion thereon.

Court No. 13-00371 Page 13

III. The Benchmark for the 2010 Parcel

Regarding the benchmark for the 2010 parcel, Toscelik claims a number of parcels are double-counted, resulting in an arbitrary benchmark unsupported by the record. See Final Calcs, PDoc 205, CDoc 212. Commerce seeks voluntary remand of this issue. The defendant-intervenors oppose, but Commerce’s reason therefor appears substantial and legitimate. Def’s Resp. at 25-26. Remand for that purpose is appropriate. See SKF USA Inc. v. United States, 254 F.3d 1022, 1029 (Fed. Cir. 2001).

Toscelik also argues the 2010 parcel’s benchmark construction was arbitrary and capricious. The rationale indicated in the I&D Memo is essentially that “more is better”; that the addition of twelve data points to the existing nine data points in the benchmark from the prior review makes for a “more robust data set.”13 Toscelik contends that since the 2008 and 2010 parcels are adjacent to one another and are of nearly identical square-meter size and features, they constitute “similar situations” that ought to merit the same benchmark, albeit indexed to the relevant difference in yearly value as in the preliminary results.

Commerce responds that despite their proximity several “material differences” distinguish the two parcels, namely the fact that the GOT conveyed the parcels to Toscelik at different times, and that one was conveyed for free and the other for LTAR. Apart from again being impermissibly post hoc, the temporality of the transactions appears to be of little relevance to land

13 Toscelik argues the addition should equate to 21 data points, whereas the Final Results only have 8 data points for the 2008 benchmark and 9 data points for the 2010 benchmark, certainly not the “more robust data set” (or sets) claimed. Pl’s 56.2 Br. at 20 n.7.

Court No. 13-00371 Page 14 benchmarking, which is supposed to be accomplished through proper “comparables.”14 See 19 C.F.R. §351.511(a)(2). And aside from the relevance of temporality, Commerce does not explain why the benchmarks used for the 2008 property (one 2009 parcel, five 2010 parcels, and one 2011 parcel) are temporally inappropriate to use in benchmarking a 2010 property, particularly since the 2009 and 2011 parcels are indexed to 2010 values. See Preliminary Calculations, PDoc 186. Furthermore, as Toscelik argues, the amount of the benefit conveyed in any particular transaction, while also post hoc, is irrelevant to what the value of the benchmark should be. The amount Toscelik paid would change the value of the benefit, not the benchmark.

On remand, Commerce is requested to address such concerns in its selection of the appropriate parcels for purposes of constructing the 2010 parcel benchmark. In addition, the fact that the 2008 parcel and the 2010 parcel are adjacent and of approximate size in square meters suggests, a fortiori, the propriety of comparing them to the same benchmark, which is what Commerce did in the preliminary determination, albeit temporally indexed for the 2010 parcel. That is not to imply that the use of different benchmarks for purposes of valuing the benefits conferred by the acquisition of the two properties (i.e., a “vertical” comparison) should be precluded, but if on remand different benchmarks are determined to be used, in the final results the reasonableness of the benchmarks selected will be adjudged by the same proximate relationship to each other, most relevantly in value (i.e., a “horizontal” comparison), as the adjacent 2008 parcel and 2010 parcel in realty reality bear to each other. See also supra & infra.

14 As Commerce’s own benchmarking method indicates, using “comparables” of proximate parcels (e.g., in location, terrain, size, features) is an accepted proposition for purposes of land and realty valuation. See, e.g., Morris v. C.I.R., 761 F.2d 1195 (6th Cir. 1985); Childers v. United States, 116 Fed. Cl. 486 (2013).

Court No. 13-00371 Page 15

III. Weighted Versus Simple Average, and Other Matters

In addition to its argument regarding reliance upon “cherry-picked” values above, Toscelik also complains that by switching from a weighted average to a simple arithmetic average Commerce effectively increased the 2008 benchmark tenfold as well as doubled the 2010 benchmark. Regarding the general explanation for selecting a simple average versus a weighted average quoted above, I&D Memo at 28-29, the impracticability of obtaining “more detailed information beyond the general comparability factors such as land-use classification” does not appear facially unreasonable to explain resorting to a simple average, but the provision of such information is an incumbence upon the parties to undertake, not Commerce. The papers presented do not otherwise provide sufficient commentary for the benefit of the court’s understanding.

Specifically, regarding Commerce’s first and more general point, it is unclear why it is necessary that the data of record must “reflect the broader market . . . throughout Turkey”. The issue goes beyond whether weighted or simple averaging is appropriate: why is it the case that the composition of the “reference data set” must constitute proportional representation of large/small tracts throughout Turkey before a determination can be made as to which parcels are the most comparable to “the land in question”? Commerce also does not elaborate on what more “sufficient detail regarding the characteristics of the land involved in the transactions underlying the benchmark data” would be necessary for that purpose, which also appears close to admitting that such data are inappropriate for use in benchmarking in the first place.

For Commerce to have taken the (proper) position elsewhere that land is unique, it makes little sense to obfuscate the uniqueness of “the land in question” by requiring that its Court No. 13-00371 Page 16 “comparables” be tied to or derived from (what appears to be) an “average” or “standard” set of data representative of the entirety of Turkish lands as opposed to an approximation of appropriately comparable parcels of land for “the land in question,” and the “necessity” of constructing a “broad market” representation of such entirety especially conflicts with the administrative finding that the benefits conferred by Turkish Law 5084 are limited to the 49 underdeveloped provinces that include Osmaniye, not the other and more-developed areas of the country. Cf. Zhongya Aluminum, supra.

Wheatland argues that use of a simple average for land benchmarking, not only with respect to the 2010 property, is consistent with policy. Whether that is true, it was not administratively offered as a reason for doing so with respect to the 2010 parcel benchmark. Remand being required in any event, further clarification on the advantages and disadvantages of simple versus weighted averages in the context of land benchmarking would assist the parties and the court, but Commerce shall not be precluded from reconsidering anew, in its discretion on remand, the benchmark that would be proper for the 2010 parcel.

Conclusion

For the foregoing reasons, this matter will be remanded to the International Trade Administration, U.S. Department of Commerce, for further proceedings not inconsistent with this opinion. A separate order to this effect will issue. So ordered.

/s/ R. Kenton Musgrave R. Kenton Musgrave, Senior Judge Dated: October 29, 2014 New York, New York