Oral argument — American Public Gas Association v. DOE

Case number 22-1107 et al. American Public Gas Association petitioner versus United States Department of Energy. Mr. Day for the SPIRE petitioners. Mr. Neal for the petitioner Air Conditioning, Heating, and Refrigeration Institute. Mr. Starcher for the respondent. All right, Mr. Day, good morning. Good morning. May it please the court, Barton Day for petitioners. I'll be addressing the substantive reasons why the court should set aside the standards at issue. My co-counsel will then address our procedural arguments and the reasons why vacator is warranted. I'd like to reserve four minutes of my time for rebuttal. Petitioners challenge DOE's determination that the standards for commercial package boilers are economically justified. That determination must be supported by clear and convincing evidence, and it was based expressly on the results of life cycle cost analyses indicating that the standards would provide economic benefits for purchasers. Those claimed benefits were very small. For what is by far the largest category of products covered by the standards, DOE claimed average life cycle cost savings that barely exceed one-tenth of one percent. Moreover, for this category of products, DOE's numbers show that the majority of all investments in standards compliant boilers are bad economic investments that would leave a 2.8-year life of the product. In fact, DOE's numbers show that significantly more purchasers would suffer net costs as a result of the standard than would benefit from it. Nevertheless, DOE concluded that purchasers would benefit economically if standards forced them to make the 23 percent of potential investments in standards compliant boilers that they're currently declining in the absence of standards. DOE reached that surprising conclusion because it concluded that there's a small percentage, a relatively small percentage of potential investments that have very high benefits. Benefits that in some cases exceed $50,000 on products that have a total installed cost that average less than $30,000. Still, DOE's analysis only generated benefits because it assumed that the probability that purchasers would turn down these highly beneficial investments is exactly the same as the probability that they would decline the very worst investments, including investments that would take more than 900 years to pay off just in simple payback terms. There are three substantive issues concerning DOE's analysis. First, DOE based its analysis on overstated natural gas prices. Second, there are trial cases with unreasonably high operating hours, which indicates that the benefits in some of these very high benefit cases are artificially high. Third, DOE's analysis was based on a random assignment methodology that unreasonably assumes that in the absence of standards, purchasers of commercial boilers have literally no tendency at all to make good investments or turn down bad investments, no matter what the economic stakes are. On review of the rule, this court found that DOE failed to address serious concerns raised by all three of those issues. On remand, DOE again failed to address those issues. It still relies on overstated natural prices, most obviously because it failed to account for the lower prices paid by very large consumers. Second, its analysis is still distorted by artificially high benefits in cases with unrealistically high burner operating hours. Most importantly, DOE failed to address the basic concern that random assignment overstates the potential for standards to provide economic benefits for consumers, while understating their potential to impose net costs. All three of these issues warrant vacater. Well, I'll be happy to answer any questions concerning any of these issues. I'd like to focus on random assignment. Can you just address, I think, a kind of factual record type issue that I'm trying to get to the bottom of on this burner operating hour point? And that is the 30 BTUs per hour rule of thumb. And the respondent seems to claim in their brief, we'll hear more from them in a moment, that that number did not just come from the sources that were cited in the 2020 rule, I think it footnotes 41 and 42, but that it seems that they're implying that that number was validated by actual utility bill data. But it's not clear to me whether that's the case or not. Can you help elucidate that for me? Yeah, I don't think that the assumption is just an assumption. DOE claims it lines up with its numbers, but it claims a lot of things that it doesn't really explain a basis for. The 30 BTU per hour as a universal assumption was panned by commenters on the grounds that it makes no logical sense to have one assumption that applies in all cases. It's not a one size fits all kind of issue. The number is high. And most importantly, the fundamental concern was that when you use that assumption, you end up with trial cases with burner operating hours that are way too high to be realistic. And that indicates that in those trial cases, benefits are overstated. And the examples of that are cases, again, pointed out in the reply brief, where 40%, 40% of DOE's total benefits are in cases where they assume boilers are operating 24 hours a day, seven days a week for more than seven months out of the year. And the nature of these products is such that they have to be able to meet peak demand so that they would never be operating continuously for that long period of time. So regardless of where DOE says that that particular assumption came from, the assumption was challenged as unreasonable. And most importantly, it was the results that came out of that assumption. The DOE's work based on that assumption produced unreasonable hours, I'm sorry, unreasonable operating conditions and apparently inflated benefits. I mean, I'll hear from the agency about this, but I'm sorry, I missed the beginning. I'll ask the agency this question, but I didn't see any explanation about the 30 BTU per hour number and the supplemental submission by the agency. Was there anything said about it or not? No, they didn't even refer to it. Well, they may have mentioned it, but they didn't really say anything substantively about it. And that's the problem on Remap is that they talked about their methodology, but they didn't answer the questions. And in that case, the fundamental question was, is this, whatever you're doing here, you're generating these trial cases that appear to be artificially inflated. And of course, with random assignment, it's those high extreme economic outcomes that drive the results of the analysis. All right. I don't want to take up more of your time. Thank you. I would like to focus on random assignment. I think the principle issues there are that DOE's numbers show that more consumers, more purchasers would suffer net costs as a result of the standards than would benefit from. And the only reason they show benefits is because they assume that purchasers never consider the economics of potential investments, regardless of the stakes involved. That's an absurd assumption, and DOE has never even claimed that it's valid. Their explanation didn't even address the concerns about the absurd impacts of random assignment, particularly random assignment of these very high economic consequence trial cases, good or bad. Actually, the records don't counsel that. The agency on remand, consider. My colleague has a question. Go ahead. All right. So to say that they never even considered it is just inaccurate. You may disagree with its conclusion, but it did consider it. It addressed it. It never addressed the concern about the random assignment of very high consequence trial cases. It simply, it talked about market failures that don't even remotely justify the random assignment of trial cases, for example, where the more efficient product actually costs less to install. So there's no rationale to say a standard is necessary to force consumers to make those investments. And yet, despite the abstract arguments that those cases don't exist in this context, DOE's numbers show that they claim benefits for such cases. So they never, and they never even, that was an issue that was raised right from the get go. It's one of the most, it's been one of the most controversial issues in standards rulemaking for seven years now. And DOE didn't even mention it. They simply said, well, there are market failures and therefore purchasers don't make perfect economic decisions. And the conclusion that consumers don't always make perfect economic decisions based on a detailed economic analysis is no basis at all to conclude that purchasers literally never considered the economics of- Counsel, I think you're building a straw man to knock it down. I think that Judge Rogers' point is that, and it's in the briefing, is that they did present evidence that showed that there was not always correlation between the purchasing decisions and the efficiency kind of economic impact of those decisions. I think there were like three separate studies that were demonstrated that fact. Am I wrong about that? Well, there are a variety of things that I could say in response to that. One is that I don't think their studies say what DOE says they say, but putting that aside, even if you assume that all of DOE's factual arguments are valid, DOE's conclusion is consumers don't always make perfect economic decisions based on a detailed economic analysis. And that is no basis at all for random assignment of trial cases where the more efficient product costs less. There's no rationale for assigning cases like that randomly, and yet DOE did. If you want to understand this, I would refer you to the figure on page 32 of our opening brief, which shows the entire distribution of- and these are numbers. And what they show is, first, you have to ignore that long flat line in the middle, because those aren't real results. They're just zeros ostensibly representing the 77% of potential investments consumers would make on their own. But if you look at the extreme ends, which is where all the action is, you see that there are a lot of bad outcomes depicted on the left side of that figure. And there are a couple of cases there where the payback period in simple payback terms is over 900 years. So over 900 years just to break even. What's remarkable is not that there are two of those cases out there. What's remarkable is that DOE literally assumed that 77% of the purchasers that face those kinds of investments make the investment anyway. That's exactly the same percentage as they assume would make the very best investment. And on the other end of that scale, you can see that there are 84 individual trial cases that account for all of the benefits necessary to put the average above zero. And those 84 cases have benefits of the range goes up to over $70,000. And those cases are there because DOE assumed that the probability that purchasers acting on their own would turn down those highly beneficial investments is exactly the same as the probability that they would decline an investment with a 900-year payback period. That assumption is ludicrous. And there isn't any market failure that can explain that. And I can go through and if you'd like, I can give you a three-minute tour of why DOE's market failure arguments make no sense. But I think it's pretty clear on its face. When you look at the results of DOE's analysis, they're basically saying, okay, most of these investments are bad. In fact, over 60%, according to DOE's numbers, impose net costs with most of those imposing net costs in excess of $1,000, even after 24.8 years. And yet, because of this small percentage of highly beneficial investments, we're going to claim that there are net benefits. And that only works because DOE assumes that even in those very extreme cases, there's no statistical difference in the probability that purchasers will make or decline those investments, which is absolutely ridiculous. I see I'm out of time and I didn't reserve time for rebuttal. I want to respond to... Mr. Day, would you please wait? You keep interrupting us. Judge Wilkins... Oh, I'm sorry. I'm not hearing, perhaps. All right. Well, Judge Wilkins, have you had your question answered? Yes, I have. All right. Judge Rogers? I'm fine. Okay, you're fine. All right. Mr. Day, you have used your rebuttal time and we will perhaps give you two minutes, notwithstanding that. We'll now hear from Mr. Neal. Thank you. May it please the Court, Jason Neal for Petitioners. I'd like to make three main points about the procedural issue here and the appropriate remedy in this case. First, DOE's argument that the final rule was a mere supplement, not requiring notice and comment on the agency's first-ever evidence regarding actual market failures in this market is contrary to Chamber of Commerce against SEC. Second, DOE's backup position that complying with this Court's deadline on remand required DOE to skip notice and comment is also untenable. Third, the substantive issues and the procedural issues here require vacater. In APGA 1, the Court previously declined to vacate because DOE promised a full and sound explanation on remand. DOE's failure to provide such an explanation here demonstrates that it does not have one and that the rule should be vacated. DOE's primary justification in the final rule for its refusal of notice and comment was that this Court remanded without vacater and that the final rule is merely a supplement providing further explanation. The first part of that justification is irrelevant. Chamber of Commerce explains that notice and comment requirements still apply on remand. The second part of the answer distorts both the law and the facts here. As the Court explained in Chamber of Commerce, the exception to notice and comment is for information that merely supplements information or data already in the rulemaking record. The final rule says that it supplements DOE's conclusions that market failures apply, but supplementing data or evidence is very different from supplementing a conclusion that the Court already said was unsupported. DOE claims in its brief that the 2020 rule advanced a hypothesis and some supporting explanation, quoting in part from Building Industry Association, but DOE cuts off the quote. The Court and Building Industry Association said the agency's proposal, quote, advanced for comment a hypothesis and some supporting data. Here, the notice of proposed rulemaking did not justify random assignment based on market failures and the 2020 rule didn't either, except as an attack on the supposed alternative of assuming perfect rationality on behalf of all purchasers. Even there, the 2020 rule did so only in a supplement here. DOE also claims that it had good cause to skip notice and comment because of the Court's 90-day deadline. Again, DOE is wrong on both the facts and the law. On the facts, DOE's complaint that 90 days would be insufficient for notice and comment is wrong and beside the point. It is wrong because DOE already knew, at least as early as oral argument in APGA 1, that the Court was interested in evidence of market failures given DOE's invocation of those market failures in oral argument. DOE could easily have published a request for comment providing notice of its market failure evidence and supporting data in time to meet the Court's deadline. DOE's argument here is beside the point because the Court anticipated the potential need for more time by giving DOE 10 days to demonstrate such a need. DOE declined that opportunity and did not seek any other relief that would give it more time. I think it's important to take a step back and remember how we got here. DOE urged the Court not to vacate the rule by promising a full and sound explanation on remand. If DOE thought that it needed more time for a more substantive reconsideration of the rule, it shouldn't have made that representation to the Court. But it did. And if DOE knew that it should and could have done so in compliance with the APA, and we know from the certified index to the record, and we know from the final rule that as soon as two days after the Court released its opinion, DOE was collecting evidence of market failures for it to use. On the law here... Mr. Neal, what do you think about DOE naming this a supplement as opposed to an amended rule, a revised rule, and the title actually has a long description in response to the D.C. Circuit's direction? I don't know that I've seen that before. Have you? I'm not sure, Your Honor. I do believe in the Federal Register, it does still say final rule here. I could go look and check to be sure. I don't think it does. I think it's called a supplement, but... Okay. Because I think we looked at that. I will go back and see. I'm concerned with your... Well, I am concerned that your argument raises a point that this Court in the first case, in good faith, assumed that this would be not a ministerial remand, but that it would be simply to make a fuller record and that the explanation was in the record. It just wasn't... I mean, it was gathered, it just wasn't in the record. And your argument and the argument of your colleague is that they added all sorts of things, new evidence, new arguments, new claims, and so forth, and that you are entitled to notice and comment. And in that regard, can you expand a little bit on the prejudice that you suffer without the notice and comment? Because you've got that in your record. Sure, Your Honor. Let me take those in order. So I'll start with the importance of this being labeled a supplement. I think Chamber of Commerce says when a court remands without vacater, notice and comment requirements still apply. Second to your question about was this just an explanation, I think you only need to look at the final rule. DOE admits that it collected new data and new evidence to justify the market failures that it had invoked. APGA1 said there was no evidence of those market failures. I think the court was right there. Perhaps in theory, DOE could have just written a supplemental explanation to try to address the court's concerns. I don't think it could have done that in a sufficient way that would have addressed the substantive problems, but procedurally, perhaps it could have just written something. But that's not what it did. It provided explanation based on new economic literature that it had talked about at oral argument, and it collected new data that Judge Wilkins, you asked about. But it never provided an opportunity for us to comment on that in the record. And that brings me, I think, Judge Henderson, to your last question about prejudice. We did our level best in the March 2022 filing to anticipate the issues that we thought DOE was going to address on remand because it hadn't put anything out for public comment. I think the issues that we raised there were significant, and DOE never mentioned them in the final rule. I think that itself is sufficient to establish prejudice. But the bar for prejudice under Chamber of Commerce is quite low. I think it's that we have something useful to say, not that we would have changed the outcome of what the agency did. We had something useful to say in the March 2022 request, and we would have had more to say if DOE had made public its evidence and my rebuttal time. So, I'd like to stop there and save a minute if I can, but I'm happy to answer any further questions. All right. Judge Wilkins, do you have any questions? No. Okay. Judge Rogers? No, I'm okay. Okay. Then thank you, Mr. Neal. We'll give you some time in rebuttal. Mr. Starcher. Good morning, Your Honors, and may it please the Court. Jack Starcher on behalf of the United I'd like to start with just addressing a few representations made in connection with the random assignment, which has been sort of, I guess, the lead argument throughout this case. And I just wanted to address two specific representations before questionings take us away from it that are inaccurate and need to be addressed from the other side. First is this suggestion by opposing counsel that there's something suspect or surprising about the fact that the Department of Energy didn't specifically respond to their argument about the fact that there may be some boilers for which the cost of installing that boiler on day one, the more efficient boiler would have been the lowest cost option. There's a reason that the rule doesn't respond to that point, and the reason is that that argument was never raised during notice and comment rulemaking. As we note in our brief, that argument came up only one time in a footnote to a single submission, and it wasn't made as an argument. It was just a reference to a study made in a different rulemaking involving different kinds of boilers. And in the reply brief, petitioners identify under 50, I think it's 46 maybe, applications for one of the smaller subclasses of boilers, the large gas hot water boilers, for which the more efficient boiler would have been the low cost option at the date of install. But as even petitioners acknowledge in their reply brief, even if you lop off those 46 cases, that represents just about 10 percent of the total benefits even for that one set of boilers. And what petitioners don't acknowledge is that the data is available for all of these boiler sets. You can look at it and see that, in fact, for small gas hot water boilers, the largest set of boilers, that's the focus of all charts the petitioners include in their briefs, there are no cases for which the more efficient boiler would have been the cheapest result at install. So this idea that that's somehow concerning or surprising is just inaccurate and obfuscates the fact that that argument was never raised during notice and comment and can't be raised for the first time in a reply brief during the second set of petitions for review. And with that, I want to turn to this recurring theme, both at argument today and in petitioners' briefs, that somehow random assignment that the modeling decisions used in this model assumed random behavior or assumed that consumers are making certain shocking or surprising decisions at certain rates. As we explain in our brief, this random assignment is a methodological description of an assumption used in the model to try to best reflect complex consumer behavior. It has nothing to do with assuming that individual consumers are behaving randomly. It has nothing to do with assuming that certain outcomes are going to be, you know, shocking outcomes are going to be occurring at any specific rate. The whole reason why you use random distribution in a case like this, and this is explained at length, particularly in the supplement, but also referenced in the original rule, is that when you know a lot about a market, as we did here, the department had really good data about the kinds of boilers that were going to be shipped out absent the rule. It had really, really good data about the nature of the buildings that would be installing these and the department also knew a lot about how consumers in this space act, as noted in the supplement, as noted in the rule, and that explained in much greater detail in the supplement. Specifically in the realm of energy efficiency investments, we know for a fact, and there's lots of research out there about this, and there's also the studies cited in the supplement, we know for a fact that there are a huge number of factors that go into any individual purchasing decision, and while economic benefit is one of those factors, the ability for consumers to capture economic benefit in this space is complicated, and as we noted in our brief, and petitioners really had no reply to this, even accepting their position that, you know, the rule, yeah, I guess an opening thesis of their position that the rule under-accounted for economic benefit to consumers, that the problem with that is that even just taking that on its face, what does that mean? As we noted in the rule, you know, economic benefit is a conclusion that's the result of quite a complex modeling. Would it be economically beneficial for a consumer to buy a boiler that will produce a couple hundred dollars benefit over a 25 life of a boiler, 25-year life of a boiler, if that consumer thinks maybe they won't own and operate the boiler for the entire 25 years? If the purchaser of the boiler owns a residential building where the residents of that building are going to be paying the electricity costs? For all of the reasons that are given in the supplement and also discussed in our brief, petitioners' entire argument in this front really doesn't hold together, and I think the greatest proof of that is in their reply brief, petitioners acknowledge, I think, that really what they wanted the department to do here was to randomly pick some percentage of base cases. In their reply brief, they suggest 20 percent based on no evidence and no citation. It seems that petitioners just wanted the some subset of boilers and assign those and those alone based on consideration of a single fact, which is economic benefit. But again, as discussed at length in the supplement and in our briefing, consumers aren't making these decisions based on a single fact, and we know that, and we know that not only from economic data but from case studies that only confirm what the department said in its rulemaking. I also just want to briefly address, and certainly feel free to interrupt if you want to take me in a different direction, but I think another theme in this case is petitioners' suggestion that because of the heightened standard, the heightened evidentiary standard that's at issue with this statute, the clear and convincing evidence standard, that somehow all of this supports case law, and there's a lot of it, making the kind of APA review agencies are allowed to use models. They're allowed to engage in predictives. They're allowed to use data that is imperfect in certain ways as long as they explain why the decision to do so is reasonable. Petitioners' arguments all run contrary to that precedent, and the only way that they get there, I think, is by saying that somehow clear and convincing evidence is meaningfully different in that respect. I guess I would push back on that and direct you to the whole dispute over boiler operating hours. Part of our prior decision told the agency to explain the 30 BTU per hour assumption. Do you agree that we said for the agency to do that in our prior opinion? So, in the context of burner operating hours, that the thing that this court identified in its earlier decision, and again, I think it's worth noting that on this issue, the court said it wasn't even clear whether this issue alone was enough to warrant remand independent of the other issues, but this court made clear that the concern, burner operating hours is an output of other factors in the analysis, and this court made clear in its decision that the concern that had been raised by commenters was that certain, and this is still, and again, I think petitioners acknowledge that this is still their argument to this day, that certain very high extreme burner operating hours were being produced in a way that should have signaled to the agency that other assumptions that the agency had made were wrong. Respectfully counsel, that's not the way that I read our opinion. Our opinion on this burner operating hours issue noted concerns that had been raised by petitioners and it listed one of them is the adoption of the rule of thumb that for every square foot of heated area, a building uses 30 BTUs per hour, and then we conclude that by saying that section, we expect on remand a reasoned response to these concerns as well. So the response contained in the supplement is a reasoned response to all of the concerns raised by the court in that section of its opinion, and here's why. As explained in the supplement, all of these figures that we're talking about, so this is the burner operating hours, this is the heating load, this is the representative capacity that was used in estimating burner operating hours, as explained in the supplement, and this is I think most clear on JA 525 to 526, all of those figures don't meaningfully alter the actual conclusions that supported the outcome of the rule. And again, I guess to return for kind of a half second to the standard here, remember that the standard, the question for this court is, was it reasonable for the department to conclude that the heightened standards adopted are economically justified? So economic justification is the question here, and economic justification does not turn on every jot and tittle of every, you know, subcomponent of the model. As explained in the supplement and also as noted in our brief, there were a number of economic benefits at issue here. The statute didn't require any particular quantum of consumer savings to- It seems to me, counsel, with respect what you're saying is that even though the court said we wanted a response to the validity of this per hour heat load assumption, we didn't really have to give you a response to that because we said that that assumption wouldn't have really changed anything, that that assumption was really not relevant. And, you know, I don't know, you may or may not be right about that, but when the court tells you to explain something and you don't, and we say that that's part of what needs to be explained, then that seems to me to be a problem. So again, I guess I- Because I don't see any discussion at all of the 30 BTU per hour assumption in the supplement. I believe that's correct, but again, I guess- A two-part answer. Number one- How can you have a reasoned explanation about something that you don't discuss at all? In the following way, as this court noted, all of those figures that were- that had highlighted the 30 BTU and the assumptions about- and the other assumptions that fed into boiler operating hours. As petitioners acknowledge, those figures only affect burner operating hours, or at least for their argument are relevant to these burner operating hour figures. The answer is the burner operating hours are not concerning, even with those figures baked into them. And in any event, burner operating hours are not themselves something that feed into the conclusions that matter about the rule. Burner operating hours are primarily a result of the efficiency that was assigned to the building in the model, and the- all of the data we had about that building, and we had a lot of data. We knew the exact square footage of these buildings that we were assigning models to, and we knew- we even had energy consumption estimates for each building. The savings associated with the rule are not attributable to- and not attributable to burner operating hours, and the concerns that this court highlighted in connection- and yes, this court mentioned the 30 BTU per hour as part of the burner operating hour problem, but the agency did respond to the burner operating hour problem and explained, as this court directed it to, why those anomalous numbers and why the high burner operating hours that were highlighted by commenters during notice and comment rulemaking were, in fact, not surprising or concerning. I can get into more technical- I mean, this is all very technical, obviously. I guess just to give a high level, and this is noted in the supplement, these extreme cases that petitioners now really are focusing on, these handful of instances in which the rule produced burner operating hours of, you know, eight months or so for something like five percent of all cases, that's a direct result of the fact that the rule utilized representative capacities for each boiler class. Essentially, each boiler class covers quite a 300,000 BTUs per hour, all the way up to 2.5 million. As a sort of helpful way to simplify things, the rule utilized representative capacities within that range. That's essentially what's producing these really high burner operating hours, but those hours aren't concerning, and if you went into the model and changed things, you know, to the extent you were assigning higher output boilers in order to reduce burner operating hours for the high end, the fuel consumed by those boilers would be the same, and we know the amount of fuel being consumed for each building because we had actual data on that from CVEX and the residential survey data as well. So, I guess there's a long way of saying- So, help me understand, what work does the 30 BTU per hour assumption do? It's, as I understand it, it's a baseline figure, much like the sort of representative capacity that was used in order to calculate burner operating hours, and again, because burner operating hours weren't- So, you use the 30 BTU per hour number to calculate burner operating hours? Yes, that's my understanding. Is that both burner operating hours and, or sorry, both this 30 BTU figure- So, why isn't there some sort of a responsibility then if burner operating hours is relevant to explain one of the inputs into the calculation of how you derive the burner operating hour numbers? So, that's exactly the response. Burner operating hours are not relevant. They're not, they didn't feed into what the department, again, I think this is explained in the supplement on J525, 526. What mattered was savings to consumers, right? Savings to consumers in the form of lower total fuel consumption over the life of the boiler. Did the first panel believe that burner operating hours were irrelevant? I don't know if they believed that, but certainly, I mean, the comment- Did the opinion that they wrote conclude that burner operating hours were irrelevant? I mean, why would they remand it if they thought that it was irrelevant? So, and I think this is in the original opinion, although this wasn't a focus of earlier petitions, but the comment that was highlighted in the original decision, the specific comment about highlighting anomalous burner operating hours at the, you know, there was the, this is the one discussed in our brief, and that's discussed at length in the opinion, that, you know, a commenter commented that the median burner operating hours for all classes was over a thousand, etc., etc., etc. Because burner operating hours are not an input, the reason why that mattered is, and I think the panel noted this in its original decision, the reason why that mattered is because those high, the panels, the thesis of that comment, and the reason why this panel remanded on that issue was that there was a concern that some of those hours were so unrealistic that it should have flagged for the agency that certain inputs into that figure or certain other parts of the agency's model were wrong or should be surprising, and therefore there was no need to go in and reassess any of the figures that fed into burner operating hours, right? The burner operating hours, because they're not an input into economic benefit, could only matter to the extent they raise for the agency a red flag that would in turn cause the agency to go back and reconsider things that fed into that number, and the agency's response, and maybe we've gone full circle and I've given you a much cleaner answer to your question than I have prior, full circle, the agency's response in the supplement is burner operating hours are not surprising, there's nothing concerning here, and therefore there is no need for us to take the next step and to reevaluate figures that fed into burner operating hours. I think this is clear from the panel's decision, the earlier decision, that burner operating hours were relevant as sort of a red flag problem, right? That something about burner operating hours was so unrealistic or so surprising that the agency should have recognized other problems in its model, and the supplement does. Thank you. Thank you. Judge Shanderson, I'm just noting that the clock seems to be stuck at three minutes and 43 seconds. I don't know if the clerk can address that. Ah, there we go. There is a connection issue, but I am keeping track of the time. Okay. Sorry, I just got so engrossed in discussion of burner operating hours, I didn't even notice the time had stopped. Thank you for flagging that. I'm happy to respond to any other questions as my time runs out. I want to make sure we, there's a lot of, I know a lot of different moving parts in this case. I guess one point that I did just want to make, and this is returning us to the random assignment space, and I just want to make sure to get this out there. Petitioners, another, you know, and they said this, there's a set of argument today that somehow the rule was only economically justified because of some very, very small percentage of very high benefit cases. That simply, you can look at the data on this, DOE has looked at the data, you know, just as sort of a pressure test. For the small gas hot water heaters, the biggest group of boilers that were, that, you know, has been spoken about the most today, you can look at the data on this. DOE could have gone in and just ignored all of the highest benefit cases. So, for example, any benefit more than $15,000. You can go in and you can just lop off all of those highest benefit cases, and you still reach the same conclusion. You still show benefit to consumers, which is all the department needed to find, and again, it didn't, it's not even clear under the statute that it needed to find a benefit to consumers, but the department's conclusion here, which is that this was economically justified, is unchanged, even doing this totally unsupported exercise that petitioners ask the department to do, which is to arbitrarily go in and just lop off some certain number of base cases, even though it's reasonable to think those base cases exist for all of the reasons given in the rule. It's simply just, it isn't hard to envision, and I'm sorry, the clock. Yes, finish your sentence. Thank you. It's simply for all the reasons given in the supplement, it is not hard at all to believe or to imagine examples of leaving huge amounts of savings on the table for all sorts of reasons, and some of them are listed in the supplement. I think to me, the most intuitive and easy is owner of boiler does not pay gas bills because it's a rental space and someone else is paying the bills. Of course, that person is not going to internalize lower gas bills, and of course, that person is going to, in some cases at least, just buy the cheapest boiler because they don't. All right, we have your argument. Thank you, Mr. Starcher. Mr. Day, why don't you take two minutes? Thank you. I've got a lot to respond to here. DOE's arguments on burner operating hours are completely invalid. Burner operating hours are energy consumption, which is what drives benefits, divided by the capacity of the boiler. If your energy consumption numbers are wrong, your burner operating hour numbers are wrong, here the reverse is what we saw visibly, is that the burner operating hours were ridiculous. That means that either the capacity is wrong, in which case costs are low, so benefits are exaggerated, or energy consumption is high, and in that case, benefits are exaggerated. Either way, it's a problem. The fact that they don't use burner operating hours to calculate benefits is irrelevant. It's a clear indication of a problem with its analysis. The claim that owners of apartment buildings don't pay the bills is false. Boilers provide service to the building. You cannot meter them. There is a boiler in the basement. It's not like a furnace where everybody in their apartment has their own furnace. There is a boiler that serves the building. The bill goes to the building, and unless there's one tenant, that bill will never go to a tenant. That's in the record. They make these arguments, but they don't even refer to the fact that the record shows, and this is cited in our brief, that there are very, very few cases where even the possibility of split incentives exists. If DOE believed that there are 7% or whatever that they might be able to argue, and they assume that purchasers would go to the low-cost boiler in every one of those cases, what that means is you take 7% of the cases, and you assign whatever the low-cost product is, which may or may not be the low-efficiency product. The idea that that kind of a market failure even remotely justifies random assignments is absolutely invalid. The claim that these issues, particularly the concerns about the random assignment of the cases that drive the results of DOE's analysis, were not raised is just false. That issue was identified as a problem back in 2016. It was an issue I raised during the discussion with OMB when the final rule was in the works. DOE was on the line for that call. They can't claim they did not consider that. The record of that is in the administrative record cited in our brief. All right, Mr. Day, finish up your argument. Mr. Day, finish your sentence. The point is that the issue's been raised over and over again over the last seven years. It was raised in briefing below. I raised it in oral argument and in briefing. So, the idea that this is something that's a giant issue is just not correct. All right, thank you. Mr. Neal, why don't you take two minutes? Thank you, Your Honor. I'd just like to make two quick points. First, Judge Henderson, to your specific question about whether the final rule is a final rule, JA 517, the left-hand column, the bolded word action says final rule, semicolon, supplemental response to comments. That's what I had in mind in my answer to you. Second, I want to make sure that we discuss Bakatter and the appropriate remedy here. And I won't belabor the point, but I'm happy to answer questions. We've been through this process twice now. There's a long rulemaking process. We went up to this court in APGA 1. We identified fundamental defects with the rule that have come up today and that the court found that DOE had not adequately considered or explained. I think the final rule shows that the problem in APGA 1 was not that DOE didn't write enough about these issues. It's that its today that suggests that DOE is interested in some wholesale reconsideration of that. And so I would urge this court, if you agree with us, you need to vacate the rule, not just remand where the agency can write more words about these things. I think these are issues where if DOE seriously reconsidered the problems that we identified, its analysis would have to run so fundamentally differently that it would be a mistake to leave the rule in place. The rule is went into effect 10 days ago, and manufacturers were already having to make changes in anticipation of that effective date. Many of them have made significant changes to their operations. They'd love to have the previous rule back. And that's the remedy that we're asking this court to provide. And I'm happy to answer. If they've made these changes, how disruptive would it be to put them back where they were before January 10th? Your Honor, I can't speak to the circumstances of any particular manufacturer because I think it varies from one to the other. But I can say that for at least some manufacturers, they are capable, if this court orders relief fairly soon, of changing operations back, restarting lines of business that they've started to wind down. This court denied a stay. And so I there have been harms that have come to at least some of the manufacturers. But the industry would like to have the old rule back. And I think there would be a real benefit for that. As far as disruptive consequences go, this is just not like other cases this court has discussed where there's some effect of the rule that just can't be unwound. I'd be happy to talk about other cases, but that's not the point. Okay. Fine. Thank you. Any questions? Judge Wilkins or Judge Rogers? No, thank you. No. All right. Thank you then, Madam Clerk, if you would close us down.