OPINION

RESTANI, Judge.

This matter comes before the court following its decision in Qingdao Taifa Grp. Co. v. United States, 710 F.Supp.2d 1352 (CIT 2010) ("Taifa II"), in which the court remanded the Final Results of Redetermination Pursuant to Court Remand (Dep't Commerce Jan. 22, 2010) (Docket No. 100) ("First Remand Results") on Hand Trucks and Certain Parts Thereof from the People's Republic of China; Final Results of 2005-2006 Administrative Review, 73 Fed.Reg. 43,684 (Dep't Commerce July 28, 2008) to the United States Department of Commerce ("Commerce").

BACKGROUND

The facts and procedural history of this case are fully explained in the court's two prior opinions in this matter. See Taifa II, 710 F.Supp.2d at 1353-55; Qingdao Taifa Grp. Co. v. United States, 637 F.Supp.2d 1231, 1234-36 (CIT 2009) ("Taifa I").

In Taifa II, the court instructed Commerce "to determine, after proper investigation and analysis, whether a government entity exercised nonmarket control over" plaintiff Qingdao Taifa Group Co., Ltd. ("Taifa") sufficient to link the People's Republic of China ("PRC") entity-wide rate to Taifa. Taifa II, 710 F.Supp.2d at 1357. The court gave Commerce three options:

Id. at 1358 (footnote call numbers omitted).

Commerce asserts it chose the third option on remand. Final Results of Redetermination Pursuant to Court Remand 4, 23 (Dep't Commerce July 27, 2010) (Docket No. 118) ("Second Remand Results"). Commerce found that "Taifa failed verification with respect to its separate rate status," id. at 19, because Commerce found documents indicating that the Yinzhu Town Government owned a majority interest in Taifa, contrary to representations in Taifa's separate rate questionnaire responses, and because Taifa failed to register with the proper authorities documents indicating the transfer of the majority interest to certain individuals, six of whom were also members of Taifa's board of directors, which controls and manages the company, id. at 5-19. Commerce concluded that it could not determine whether those directors "actually operate under their own legitimate independent direction as Taifa claims, or whether the absence of proper documentation reflects an undisclosed continuation of governmental control over Taifa." Id. at 13. Commerce found that Taifa had not established a legitimate separation from the town government and applied a "presumption" that a respondent in a nonmarket economy ("NME") country such as the PRC is state-controlled. Id. at 13-19.

Following the remand determination, the court met with the parties in an attempt to learn how Commerce addresses these issues and the basis for its presumption of state control in this industry or for this company, which contrary to the court's order did not appear to be explained adequately. The parties were forthcoming about their views of these matters, but their approaches understandably differ. The court must address these underlying methodological issues in order for it to resolve the basic dispute of whether plaintiff should receive its own rate or the 383.60% PRC-entity rate.1

JURISDICTION AND STANDARD OF REVIEW

The court has jurisdiction pursuant to 28 U.S.C. § 1581(c). The court will not uphold Commerce's final determination in an antidumping review if it is "unsupported by substantial evidence on the record, or otherwise not in accordance with law." 19 U.S.C. § 1516a(b)(1)(B)(i).

DISCUSSION

The essence of Commerce's reasoning is that Taifa's assertion that it is controlled by an independent board of directors is not credible because Taifa impeded Commerce's review by withholding information relating to its town government ownership and provided information that could not be verified in records outside the company. See Second Remand Results at 15-19. Although in assigning an antidumping duty rate Commerce may draw an adverse inference against a respondent that withholds information, significantly impedes a proceeding, or provides information that cannot be verified if Commerce finds that the respondent "failed to cooperate by not acting to the best of its ability to comply with a request for information," 19 U.S.C. § 1677e(b); see id. § 1677e(a)(2), such an adverse rate should "be a reasonably accurate estimate of the respondent's actual rate, albeit with some built-in increase intended as a deterrent to non-compliance," not a punitive or unreasonably high rate "with no relationship to the respondent's actual dumping margin," F.lli De Cecco Di Filippo Fara S. Martino S.p.A. v. United States, 216 F.3d 1027, 1032 (Fed.Cir.2000) ("De Cecco"). Accordingly, Commerce may not apply the PRC-wide rate if substantial evidence does not support the finding that a government entity exercised nonmarket control over the respondent, and if there is no ultimate link between the respondent and the central PRC government.2 See Taifa II, 710 F.Supp.2d at 1356-57; Taifa I, 637 F.Supp.2d at 1240-44; see also Gerber Food (Yunnan) Co. v. United States, 387 F.Supp.2d 1270, 1287-88 (CIT 2005); Shandong Huarong Gen. Grp. Corp. v. United States, 27 CIT 1568, 1594-95, 2003 WL 22757937, *18-20 (2003).

In order to understand how this case arrived in its current posture and how it may be resolved, it is necessary to explain the development of the case, one hopes, in plain terms. First, what are no longer open issues before the court are 1) that plaintiffs may challenge, without the bar of failure to exhaust claims, its designation as part of China's state-owned enterprise structure, 2) that plaintiff may not challenge the calculation of the dumping rate applied to that structure, the PRC-wide entity rate, having waived that issue, and 3) that if plaintiff is to receive a rate separate from the PRC-wide entity rate, it will still be a rate based on facts available and an adverse inference ("AFA rate"), because of acts committed during Commerce's verification visit to its factory.

What remains at issue are the following: 1) Is plaintiff entitled to a separate AFA rate. 2) If so, is the separate AFA rate calculated by Commerce in its first remand determination supported by substantial evidence. The alternative PRC-wide entity rate plaintiff seeks to avoid is 383.60%. There is little likelihood that in any real world this could be an approximation of an actual rate. It is nearly four times the price of the subject merchandise, i.e., hand trucks and certain parts thereof.

As the court understands it, such rates derive from Commerce's decision that the PRC government, as the presumed owner of enterprises in a particular industry, which as a sovereign does not respond to questionnaires and does not cooperate in the investigation, receives a very high AFA rate based on information normally provided by the domestic industry in its petition. Furthermore, any entity which does not establish its status as separate from the PRC government receives this rate.3

In the preliminary determination, Commerce found that Taifa was sufficiently separate from PRC government-ownership and control, that it would receive its own rate. See Taifa I, 637 F.Supp.2d at 1235. Events intervened, notably the failed verification of Taifa's sales and production data, and, as described earlier, information also gathered at verification that cast doubt on Taifa's claim that a majority of its shares were registered in the name of independent private individuals or entities. See id.

All sides to this dispute emphasize that actual control of business decision-making is the key. Nonetheless, the briefing focused on the share registration issue and the somewhat conflicting views of Commerce, also reflected in the less than clear case law, as to the importance of de jure ownership.4 As, normally, who owns a company might tell one something about who controls its business decisions, one cannot fault the parties for at least considering the issue of de jure ownership. There is record evidence, however, that this company makes its own decisions, and there is no evidence of outside control of the relevant decisions. To be sure, Taifa has been shown to be less than honest, and Commerce is entitled to treat internal documents about who is running Taifa with skepticism. Skepticism, however, does not mean total disregard.

In any case, the court, in its review of Commerce's initial final determination, failed to see the connection between a PRC-wide entity rate and the way Taifa did business, so it directed Commerce to describe the evidence which would get the court from A to B to C on the issue or else give Taifa its own rate, adverse though it would be. Commerce, in its first remand results returned a determination to the court objecting to rejection of its "presumption" of state-ownership and the 383.60% rate. In that determination, Commerce stated that it has no choice under the court's decision but to assign Taifa a separate (and different) rate. It decided upon a rate of 227.73%. This rate is from data in the latest review in which Taifa was a mandatory respondent (and where its data was not rejected). It is from a less than fair value margin calculation for a group of sales representing 12% of Taifa's U.S. sales. First Remand Results at 3, 21. The court did not reach the issue of the validity of this rate but remanded the matter again because it viewed its decision, not as directing a result, but as requiring Commerce to explain a methodology that would link Taifa to the PRC-wide rate or make clear why it was not linked. See Taifa II, 710 F.Supp.2d at 1357-58.

In its second remand results, Commerce abandoned the 227.73% rate and returned to the 383.60% PRC-wide rate. It explained the methodology essentially as follows:

See Second Remand Results at 19. The second remand results, however, did more. Commerce finally came to grips with the fact that its "presumption" of state-control was a factual one, subject to change over time and based on Chinese laws and policies in effect in 1993, but apparently no longer in effect. See Second Remand Results at 28-31. Thus, Commerce looked at an updated assessment of state-control of Chinese businesses. Id. It did not supply the document or even a web citation to it, but the parties seem to know what it is. See id.; Antidumping Duty Investigation of Certain Lined Paper Products from the People's Republic of China ("China")— China's status as a non-market economy ("NME"), A-570-901 (Aug. 30, 2006) ("Certain Lined Paper"), available at http://ia.ita.doc.gov/download/prc-nme-status/prc-lined-paper-memo-08302006.pdf (last visited Oct. 29, 2010). Therefore the court will address it, even though technically it is not in the record. Whatever data underlie this document, however, are clearly not in the record and Commerce did not cite to any particular part of the document as explaining why companies such as Taifa must be presumed to be state-owned or controlled by the state in such a way that a separate rate is not warranted. See Second Remand Results at 28-31.

The court, however, has examined the document, particularly the portion relating to the manufacturing sector, beginning at page 36. It describes a movement over time from state-control to private ownership and control, particularly as to smaller businesses, or non-core industries. Certain Lined Paper at 36-40. It is unclear to the court what kind of private ownership exists in China, but it does appear from Commerce's document that private ownership exists in this section of the economy, so that businesses keep their profits and even foreign investment is permitted. Commerce seems to recognize this by allowing companies to demonstrate separate status. There is no indication that Taifa has foreign investors, but the court could not find the support Commerce alleges is in this document for a presumption that Taifa is state-owned or controlled to a degree that warrants application of the country-wide rate. That Commerce may make a judicially unreviewable decision that China is overall still a non-market economy,5 does not seem to answer the question of whether Commerce has supported a presumption that a manufacturing company such as Taifa is likely to be state-owned. Thus, to date there is no link to a rate other than Taifa's own rate.

If this were a case between the government and Taifa alone, the court would say "enough." After two remands and a failure to provide the information sought by the court to support Commerce's determination, the court might leave the government to litigate this when, presumably in another case, it is ready to support its view. In which case, the court would order that Taifa receive a separate rate, especially in view of Commerce's choice not to take the opportunity to reopen the record. There is no requirement in the statute that a PRC-wide rate be imposed when other adverse rates are available or may be constructed. This case, however, involves other parties, the domestic competitors. They remain entitled to a well-explained decision, as much as Taifa does.

So, we continue. While not a model of clarity, the second remand determination finding that the evidence of town ownership is at least in equipoise is supported. See Second Remand Results at 8-15. Share documents, which would indicate that a majority of stock is privately held, are not properly registered. Id. Certain stamps on "chops" are missing from other documents which might establish separation from town ownership. Id. at 9. Documents in Chinese government records did not match Taifa claims. Id. at 10. Thus, a finding of de jure town-ownership may stand.

Despite this, Commerce could find that an independent board of directors made Taifa's business decisions without contamination by government resource allocation or other non-market controls. However, some explanation of how and by whom the decision-making of town-owned manufacturing businesses generally is controlled in China is missing from the second remand results.

Thus, this matter must be remanded to Commerce. If Commerce cannot explain why substantial record evidence supports a finding of central government control that justifies imposition of the PRC-wide entity rate, Taifa must get the rate its own lack of verifiable production evidence warrants, without resort to an unconnected country-wide rate. Commerce is permitted to reopen the record on the issue of the link to the country-wide rate. The mere conclusions in the August 30, 2006, Memorandum, however, do not support a link to a PRC-wide rate, as explained.

Finally to avoid yet another remand, the court will review the selection of 227.73%, the separate rate from the first remand results, and the only separate rate Commerce has selected, to the extent it is able to do so on this record. Based on the results of the investigation and reviews cited by the parties, the court doubts that this AFA rate reflects reality. The issue is, is the AFA rate so far from what has been demonstrated by actual rates, that it must be rejected as in Gallant, or is it an adequately supported rate as in KYD. First of all, this is not an actual rate. It is not derived from an overall rate calculated for anyone by anyone. Unlike a petition rate, which although it is not from respondent's own data, is an overall rate, this rate is for a portion of Taifa's sales. Commerce sometimes uses a portion of sales to corroborate an overall rate based on facts available.6 7 It does not appear to have done that here.

The court does find support for Commerce's rejection of Taifa's 26.49% rate from the original investigation. See First Remand Results at 2-3. After receiving such a rate, Taifa failed to cooperate fully in a subsequent review. Commerce does not err by rejecting this rate. Thus, Commerce should calculate a supported rate, particularly one somehow grounded in the realities of this industry. Commerce is not necessarily confined to the rates of the investigation (up to 47%), but the rate must somehow be grounded in reality to avoid imposition of a punitive rate. See De Cecco, 216 F.3d at 1032.

The court does not presume to know how Chinese businesses in general, or Taifa in particular, operate. And the court appreciates that Commerce is sometimes stymied in trying to find answers to these questions. The court, however, can only review the record before it. The court does appreciate the assistance of the parties in explaining the applicable methodologies, but in the final analysis, there must be substantial evidence supporting a decision. A presumption based on nothing is not evidence; thin air is not evidence supporting a dumping margin, particularly one of almost 400%.

REMANDED.