Oral argument — Timken Company v. United States

The next case is the Timken Company v. The United States et al., 2014-1454. Mr. Deprest, when you're ready. Good morning. Thank you. Your Honors, our briefs raised three legal errors that the United States and the United States and we believe that all of them are grounds for remand. There is one though that I would like to start with because it is so absolutely fundamental to the administration of the anti-dumping law. The law was, as you know, changed in 1994 as a result of international negotiations. And the change required that Commerce, instead of what it had always done, which is it looked at transaction-specific export prices, it was now required to look at average export prices. And this concerned Congress. So what did Congress do? It not only put in a section in the statute that said you have to look at average export prices, it also put in another section that said you, when these particular conditions are satisfied, are allowed, are permitted, to look at transaction export prices. Look at, use your old method. When? Well, when these conditions are satisfied. And the conditions are about masked and targeted dumping. And the irony is now that in this case, Commerce has used, has pointed to this provision, which was put in for the very purpose of addressing masked dumping. Commerce is pointing to this provision to provide it with authority to set aside the evidence itself had uncovered that there was masked dumping. Well, the statute says may, right? It does say may. As you pointed out, it permits Commerce to go with the A-to-T methodology. But it doesn't mandate that if these two conditions are met, that it now, without discretion, must apply the A-to-T methodology. The position of the Timken Company, which I think is the only one that finds support in the legislative history, is that what Congress intended to do is provide Commerce with discretion to choose the tools that it thought was most appropriate to address the issue of targeted dumping. And there are numerous cases with, you know, when this case first, not my case, of course, in the previous cases, it was a new practice. But there are now a lot of cases where Commerce in fact finds that, yes, we've identified targeting under the new test differentiated pricing, and we don't have to apply our old methodology because the average methodology also takes care of the issue. This happens far more often than I thought would be possible if you look at the various determinations. But that's one of the factors, right? If the A-to-A methodology can take care of the differences, then you're fine with that. But, you know, we're not in that situation because Commerce herself or itself has in the record is evidence that demonstrates that there is a mass dumping because that was Commerce's test. They did the test. They found that there was targeting, varying amounts for different companies, but targeting was found. They don't deny this. They're just not going to address it. And they also have on the record the results of the... ...kind of sufficiency inquiry. That's right, which to us makes no sense because when they first made their determination was presented to us in the form of post-preliminary calculation. So it was not, we were not graced with much of an explanation other than to say, well, it's not enough. Whatever this number is, it's not enough. And so we then pointed to several decisions of the Commerce Department that had occurred before where they said that, no, we're not going to do this type of analysis because our test doesn't really, you know, aim to establish a totality of targeted sales. It just aims to establish whether it's in fact targeting. And if it is in fact targeting, then we are going to look at what it does to the margins. And if we think we need to apply our remedial or our traditional methodology, then we will. If we think that it's not necessary, then we won't. That has been their approach. This is the 22nd annual review. Right. Does this, does the decision below represent a change from all the previous 21? Yes, it's the first time that, you know, we've been presented with a situation where, you know, we're good, you know, if I may speak colloquially. Commerce had, we were worried in the sense that we knew about the changes in the legislation. And we submitted, Commerce at one point submitted a proposed methodology. And then seeing that the writing was on the wall, we submitted a targeted dumping allegation in plenty of time. And then Commerce postponed its appeal in several places. Why don't they have discretion in determining their methodology? Well, they do have discretion to determine their methodology. What they do not have the discretion to do is to set aside the statutory goal. The premise for this provision is to address targeted dumping. It provides the agency with authority to pick a methodology, but it has to be a methodology that addresses the issue, mass dumping. You can't just choose, well, in this case, I mean… I think it's the same thing. I mean, I believe the SAA uses the term mass dumping, but I'm not, I believe it is the same thing. How many of the previous 21 annual reviews have been on appeal here? I'm very rarely the appellant. I'm usually on the other side. Isn't it because for most of them they used A&T because it was prior to this legislative change. Exactly, yeah. So your appeal to us is an appeal saying, we want you to continue to do what you did for the last 21 reviews. Yes, exactly. Not that we want you to do something different than what you did for those 21. Yes, but you recognize the change in the law. But, you know, I think any interpretation that uses the discretion granted for a particular purpose, i.e. to address targeted dumping, then uses that same discretion to say, oh, why, I don't have to address targeted dumping even if the conditions are satisfied, even if I myself determined that there was targeted dumping. That just cannot pass. Why not? It says the word may. It does provide discretion. Right. As we, as you open this argument, they're permitted to use that alternative tool, but it doesn't, the statute doesn't require that. I cite a number of cases, two trade cases were cited in my original brief, and then I cited a non-trade case, the McBride case, in the reply brief. And they all, the basic premise, which is shared with lots of other cases and other circuits, and it's also apparent from the dictionary definitions and so forth, that the word may in a statute, yes, it does suggest that there is discretion being provided, but that discretion must be tested against the legislative intent, against the structure of the statute, and what Congress intended to accomplish. Right, and then here, apparently Congress's determination, consistent with that Taiwan case, talked about, well, we have to figure out whether there's a sufficiency of what's we're targeting to determine whether in this instance we're going to go with the alternative methodology. And in their Federal Register notice, I can't remember, 2008 or I think it's 2012, they announced that they were going to do these kinds of cases really on a case-by-case basis, right? Right. But the explanation, the mere statement that we are proceeding on a case-by-case basis is basically, you know, every review is based on a unique record, so every case is case-by-case. Stating that Commerce does this on a case-by-case only means that it has to, that it will, or hopefully will, take account of the evidence that is in the record that is particular to the particular case. Do I understand one of your arguments in this case to be that in the Wood Memorandum, for example, in the Wood Memorandum of Decision it said, we reject the de minimis rule, and they go on and explain why they're going to reject the de minimis rule. And then in the UAE review where they likewise issue a decision and say, of course we reject the de minimis rule because targeting is meant to address situations where the overall average doesn't, isn't affected. Right. Is your view that that is a position that was adopted by Commerce because the Wood case came first, so the Taiwan case that Judge Chen cited came after that, and that basically they're flip-flopping around whichever way the wind blows them on this de minimis rule as they please, and that while they, even if they had discretion, say we don't go with you on your first argument. So if they have discretion, they don't have the discretion to be inconsistent. That's not permitted. Is that one of your arguments? Yes, that is one of my arguments. The, you know, obviously the Commerce determination, the agency is not constrained by a set of decisions and it can make different decisions, but... Not on identical facts. But... And here's the thing I don't understand. How are you ever supposed to know what the facts are? This is, what I don't understand is that in every other review, if they say this qualifies for de minimis or this doesn't qualify for de minimis, the numbers don't ever get published. Those are all confidential. Oh, yes. So how are you, or any person, not just situated like you, but situated like, you know, the foreign importers, how is anybody supposed to know what the confines of this de minimis rule are that Commerce has adopted? I agree with you on that. But Judge Rastani... It was a very... I'm sorry. Judge Rastani pointed out that you didn't make that specific argument below in front of Commerce and therefore she wasn't going to entertain that argument at the CIT level with respect to whether or not there was an adequacy of determining whether there was a sufficiency or insufficiency of some pattern of targeted dumping. Isn't that right? Well, I'd first like to say that I'm in my rebuttal time, but I raised before the agency all the same arguments that I raise now with the sole exception of the discretion argument because that was not put forward by the agency until the, you know, final IDM, so we didn't have access to that. Is there somewhere in the joint appendix that you can show us that points that after the post-preliminary results came out from Commerce in your comments, you specifically pointed out that the determination that there was a sufficiency was totally inadequate? What we did, we attacked it head on. What we said was... You attacked it on whether or not they can even consider some notion of sufficiency. No, that's right, but we also did more. We said if you are really going to do that, then you probably should not, you didn't want them, you know, but if you're going to do this, you should look at a very different number. What you should look at is the total of sales to those customers or regions or time periods or whatever it is that you found or confirmed were targeted. Take those sales because then you have a denominator that's, you know, more consistent with the numerator and it makes more sense. Instead of taking this number that, yes, it's available, but it's being used for a purpose it was never intended to be used for. Because, you know, the test, the NAILS test never was set up to determine what the total number of targeted sales is. That's not what it does. It just confirms what a particular customer is targeted or a particular region is targeted for an allegation. It first has to make an allegation. Commerce doesn't even do its own independent review. Mr. Deprest, if you want to save time for rebuttal, we'll give you three minutes for rebuttal. Thank you very much. Ms. Hogan, you're going to split your time with Mr. Falcone, 10 and 5. Good morning, Your Honors. May it please the Court. This accuse really presents a very simple question. Oh, I don't find this case very simple at all. The methodology aside, the math aside, I think the legal question is, Commerce announced in its regulation in 2012 that it would normally use an average-to-average comparison method in both investigations and reviews. It considered Timken's arguments to use an alternate comparison methodology, but in the end concluded that it would follow its regulations. Timken is not arguing that that's contrary to statute or contrary to regulation. The only legal question is then, was what Commerce did either an abuse of discretion or somehow arbitrary or capricious? The two basic arguments that Timken makes are, first, that Commerce had some established or long-standing practice in which it would always consider the A-to-T method any time there was a single instance of targeted dumping. This is simply not true, and I can go through the facts of the prior proceedings. It's also particularly not true because this NAILS methodology was relatively new. In fact, this was the first time that it was being applied in administrative reviews, so Commerce was still developing this practice. The second argument would be somehow that Commerce failed to provide sufficient guidance to the parties as to when Commerce would depart from its regulatory default method. As the trial court noted, Timken did not argue that the particular percentages presented in this case should be considered sufficient or that Commerce should establish some bright line rule. That argument was never presented. To the extent that that's a question that should concern anyone, I think it's worth noting that the NAILS test was subsequently abandoned by Commerce. This isn't even the test that Commerce uses now to test for targeted dumping. But to the extent that it's still there in some proceedings before the Court of International Trade and before this Court, Commerce subsequently did establish a number in the C.P. Kelko case on remand. To the extent that there was any question about clarity... In the Kelko case, did Commerce respond to that remand order already? Yes. Kelko, Judge Goldberg sent the case back to Commerce to more fully explain several different things. Right. And Commerce responded? Yes. And they established a number, an actual threshold? Right. And would that threshold result in these litigants receiving different treatment or the same treatment? I'm not sure I can answer that without revealing... You don't have to reveal what the threshold is. If you've established a threshold, and this is a definitive policy determination, then you can certainly reveal whether or not that threshold would result in a different... It would not have met the threshold. So this case would not have come out differently, even if decided after the threshold was established? Correct. Correct. This really, I guess, sort of puzzles why we're here at all. Because this is not an issue... This is not forward-looking at all? This is not forward-looking. This is not... Commerce has subsequently identified a differential pricing test that it's now using to identify targeted dumping. And so, again, within the very substantial discretion that Commerce has, using the default comparison method that it set forth in the regulation that it was going to use and did use here. And as the trial court noted, the percentages of targeted sales here were, quote, very small. And Commerce did not see that as a reason to depart from its normal default methodology. Your brief was really fighting against equating this sufficiency inquiry with a de minimis standard. And I need help. Where's the daylight between the two things? So I think there's two things. I mean, one is sometimes this de minimis terminology was really talking about the first step of the NAILS test. Is there a pattern? And there's no de minimis requirement there. If it meets the definition of that first standard deviation test, there is a pattern. I think why we really are pushing against that de minimis language here, as opposed to using a sufficiency language, is that, again, one of the reasons Commerce decided to approach this on a case-by-case basis was because it was trying to develop a methodology against real-world facts. And it didn't want to say there's a de minimis standard or there's a certain percentage. But there's a sufficiency standard. Because a sufficiency is really more of a stepping back. So we have the results of the NAILS test. There are targeted sales. There's no doubt about that. But let's step back and look at whether the percentage of these sales, by value and by volume, compared to the universe of sales, whether it makes sense to depart from – whether those small percentages of sales make sense to depart from what we've said we're going to use as a standard. And I know that there is some frustration that Commerce wasn't – Do you know why that, to me, sounds almost a little bit like a de minimis standard? Because it is one. I can certainly understand why many of the parties have called it a de minimis standard and why there's – Yeah, the China case, the UAE case, the respondents were saying, hey, you need to consider the pattern on whether it's significant enough, whether it meets some de minimis standard. And Commerce, each time, pushed back on that and said, no, we're not going to do a de minimis standard. And now we're looking at a case that says, ah, these are insufficient. What Commerce said was that we're not going to refuse to find a pattern, even if the pattern is small. So just as here, Commerce didn't refuse to find a pattern simply because the percentages were small. But notwithstanding that, let's look back at – now that we've established a pattern, we've established that there is – there are targeted sales, let's look at that in the big picture, in the overall scheme. And sufficiency is the term that Commerce has selected or had selected for the purposes of this test. I mean, but I feel like you're doing a little bit of rewriting history. And maybe it doesn't matter, and you can explain to me why it doesn't matter. But in the Wood Memorandum, Commerce said the Department will apply an average transaction comparison for all sales and calculate the weighted average dumping margin. In addition, the Department determines establishing a de minimis standard would not be appropriate because once the Department finds any instances of targeted dumping, the Department has determined that application of the average transaction methodology is necessary. Period. That's it. That's all there is. I mean, that's why – you know, so I don't think that you're right that, oh, to the extent that we rejected de minimis, it was insane about the pattern. I mean, this – that language clearly does not go to the pattern. You may tell me it's not an established position or a clear policy statement, not federal regulation. I mean, I'm looking to hear you tell me why that doesn't establish a policy by Commerce, but I don't think it's fair. What you said a minute ago isn't accurate. So – and my recollection from the Wood Flooring case is that that was a case in which Commerce had decided to use the AT method. And so the question was, do you apply the AT method to only targeted sales, or do you apply it to all the sales? That was – and I think the trial court noted that. Actually, here, let me clarify your understanding. The department disagrees with GOK, LAO, Wood, and sampling groups' suggestion to modify the current targeting. And number one – and they established the multitude of issues on appeal – number one is adopt a de minimis rule. Number two is apply average transactive methods only to the A to T sales. And then here you say, we're going to apply it to everything. In addition, the department rejects the notion of a de minimis. These were two discreetly argued issues, and you reached two different holdings with regard to them. So I guess – I definitely don't read Wood the way you're suggesting your recollection is. So tell me why, if I read Wood the way I do, nonetheless I shouldn't view it, because this is what I'm hoping and looking for you to do, quite frankly, is explain to me why this doesn't create a scenario in which Commerce has done something improper by flip-flopping around. Look, Commerce has the discretion. I'm with you. You got me. But discretion is bounded, and discretion doesn't allow you to flip-flop from one circumstance to another and thereby reach inconsistent results in otherwise identical cases. So if this is a policy statement by Commerce, we're not going to apply de minimis. If there's a single targeted dumping, boom, you're stuck with A and T. You understand where I'm going. So now please explain to me why this, if I interpret it the way I do – which you don't interpret it the same way, and that's fine – but if I interpret it the way I do, why it shouldn't cause me pause. If the court reads that language in the manner suggested, it certainly does not create what we would call an established or long-standing practice of Commerce always doing something one way. Well, it might not be long-standing, but when Commerce comes along in one case and says, we're never applying de minimis, we don't agree on what it says. Nah, nah, nah, nah, never, never, never, uh-uh-uh. And then the next case, they say, ah, de minimis, hey, looks good here, let's apply it. Wouldn't that be problematic? I mean, wouldn't that be the kind of decision-making that feels arbitrary and capricious? I don't have a problem with Commerce actually doing that. Then they have to explain why, right? They would say, well, you know, circumstances have changed, or this case involves different facts, or something. I mean, they just have some obligation to explain, because the executive is free to change its mind, you know, as the elections warrant. But, you know, they have to explain it when they do a 180-degree about phase. Sure. And I would say, even assuming that that's what Commerce did, there's a couple different reasons to treat it differently. Of course the facts are different, and I certainly understand the court's concern that we don't always know what the facts are, because they're often BPI. But this was also, these three administrative reviews were the first administrative reviews that Commerce was going to use the A-to-A method in a review, as opposed to an investigation. And I think there is some, although Commerce agreed that it was... I don't understand the difference between reviews and investigations, and I didn't find the arguments persuasive in the briefing on this point, so go to your next point. Okay. The bottom line is that Commerce announced in the post-preliminary memorandum that it didn't find those specific percentages to be sufficient to warrant consideration of the AT. The parties had the opportunity to argue and to convince Commerce otherwise, and Commerce declined to depart from what it said it was going to do in its regulations. And I'm sorry, I know we're cutting a little bit into your co-counsel's time, but can you tell me what the CALCO remand said? What did Commerce say as to this de minimis efficiency thing, and which part is it actually used in, and then how does it determine whether something is de minimis or sufficient or insufficient? Right. So in the CP CALCO case, on remand, Commerce set forth a 5% threshold for sales passing the NAILS test in order to apply AT. So to the extent that we needed now a benchmark for applying, for consideration of AT, that was the benchmark that Commerce set. So after you do the NAILS test and determine that there is targeted sales, you look to see what percentage of those sales compared to the value and volume of the university sales. So it's part of the May... It's part of the May. It's not part of the pattern. No. And now... And now we're using differential pricing. Using something else. Yes, yes. NAILS test is gone. It is no longer being used in ongoing, I mean, there's still, like I said, there's still cases that are still out there being litigated that this issue is still an issue. And if I could just make one quick point, I think to the point that Commerce is not where Commerce did apply the AT method in... After it applied the NAILS test to address targeted dumping. I assume both of those meet the 5% threshold. That I don't know. But to the extent... Are you going to end up in trouble if that's not the case? I hope not. Thank you, Ms. Hogan. We'll give Mr. Falcone his full five minutes. Thank you. And when Mr. Deprest comes up, we'll give him his full five minutes. May it please the Court, Chris Falcone on behalf of the SKF Companies. Just to add one thing to the whole de minimis Commerce potentially. Flip-flopping. I just wanted to point out that this Court has held in... You're not going to point out that we flip-flop, are you? That's not going to be much favor with us. No, no. When Commerce is making a discretionary decision on a case-by-case basis, it need only act within its statutory authority and does not need to provide a reasoned analysis. Yeah, but once you do provide a reason, you can't come along in the next case and provide the exact opposite reason. Right? I mean, if all other things are the same... I mean, you're right. Congress doesn't have to explain it all. But once they choose to, they don't get to go flip-flopping. Well, I believe that case, I believe the Court said, you don't have to provide a reasoned analysis to explain every departure from your prior approach. So in that case, they had one approach for determining what the best information available was and a party argued that they were doing it a different way this time and that they didn't provide an explanation for their shift in approach. And this Court held, because it was a discretionary case-by-case issue, that they didn't have to provide that kind of reasoned analysis as long as they were acting within their statutory authority. So I just wanted to add to that. And now I just want to go back to Timkin's first argument. They placed a lot of weight on the statute and Congress intended, by passing the statute, for Congress to address every instance of targeted dumping. But what they fail to mention is that the statute doesn't even apply to administrative reviews. The statute 1677F-1D1 only applies, that sets forth the standards for addressing targeted dumping, only applies to investigations. And as Congress noted in their final results, there is no statutory provision addressing targeted dumping in reviews. So does that mean, according to you, they can't use A&T in reviews at all, ever, even if identical circumstances are otherwise present? Well, one of our colleagues did argue that. We did not argue that. So you're going to let them do it, even though there isn't a statute that allows them to do it. But you nonetheless want me to recognize the distinction. Why, precisely? Because Congress didn't say anything about it. Their position is basically, Congress didn't limit how we do this in reviews, so we can do it how we want. And if the statute says nothing about it and we don't want to apply the alternative methodology here, there's nothing saying we need to. And they also suggest that there's these line of cases. I know you said you don't find it persuasive, but I just wanted to point out that none of these cases were reviews, and in the only review prior to this in which Congress addressed targeted dumping, passing the NAILS test, and decided, at least for one of the respondents, I believe that it wasn't sufficient. And in terms of the Divinity Test also, I just wanted to point out that it's not as if Congress was applying some bright-line rule in every case. They were approaching it on a case-by-case basis, and it wasn't a de minimis test in the sense that they were declining to do something they're generally required to do. That's how I normally think of a de minimis test. The statute says I have to do something, but the metric triggering that is at a very low level, so we don't need to do it. But here, the statute didn't require them to do anything. They were just looking at the volume of sales passing the NAILS test and using that as a factor to make a discretionary decision. And I see my time is almost up. I just want to note you guys, Your Honors, asked about whether this has ever happened in previous reviews in this proceeding. Well, it hasn't. It's because this is the first time that there was an allegation of targeted dumping ever made, so there's no reason for Commerce ever to address this issue in previous Bulgarian reviews. Thank you, Mr. Falcone. We'll give Mr. DePresse his full five minutes if he needs it. Thank you, Your Honor. First, I would like to provide the citations that you had asked for. The Joint Appendix contains our post-preliminary comments in their entirety, in fact. And so I've brought up only one volume here, but they contain the same arguments. So at Joint Appendix page 4973, you will find the argument that we made about the specific number. I think that's what you were concerned about. And what we said was we went through the test and what would make sense. And then we have some example that we lay out to try to make it more understandable, which got some billing later. And then we proposed that Commerce should use a numerator in the ratio that consisted of all the sales to the customers or regions or time periods that the department had found to be targeted or that were identified by the test. And so that is the argument that we raised about the specific number. After we obviously argued that there was no place for this at all. Then I would also like to address the statement that was just made about the authority of the Department of Commerce to even do this in reviews because the statutes talked about this in the context of investigations. That argument just holds no substance whatsoever because there is a provision in the statute about reviews, but it assumes that Commerce in reviews used transaction-specific prices. And it says, when comparing export prices of individual transactions to the weighted average price of sales of a foreign-like product, the administrating authority shall limit the averaging of prices to a period not exceeding a calendar month. So it assumes that Commerce would look at transaction-specific prices. Now, as a result of the WTO dispute, when the agency introduced changes to regulations, in February of 2012, it stated that, okay, so we're now going to do the same, by regulation, we are going to do the same thing that we did before in investigations, except for the business of limiting the averaging to one month. We're going to do the same thing, and we're going to apply the same criteria. And that addresses the question. They said they're going to do it on a case-by-case basis. Right. Yeah, I don't see anything in that statement that would somehow undermine our position. I applaud the agency's efforts to make case-by-case determinations. They have to be case-by-case. They have to reflect the record of each individual review. Anything more, Mr. DePrest? I have nothing more. Thank you. Thank you very much. The case will be taken under review. All rise. Our report will adjourn until tomorrow morning. It's at o'clock a.m.