Oral argument — Michigan v. EPA

We'll hear argument this morning in Case 1446, Michigan v. Environmental Protection Agency and the Consolidated Cases. Mr. Lindstrom? Mr. Chief Justice, and may it please the Court, EPA's view that it can decide whether to regulate electric utilities without considering costs is contrary to the text and structure of Section 7412. The text sets out two distinct terms and thus directs EPA to consider whether it is appropriate to regulate and whether it is necessary to regulate. EPA found that it is necessary to regulate because of the existence of public health harms, and it found it appropriate to regulate for the exact same reason, the existence of public health harms. I'm not sure that that's quite what EPA said. My understanding of what EPA said is that it's necessary because of public health harms, and it's appropriate because there are technologies that can address or remedy those public health harms. So, on the one hand, it said that the phrase appropriate went to the nature of the harms, the phrase necessary went to the nature of the harms, the phrase appropriate went to the existence of technologies. Justice Kagan, when they relied on the availability of controls, they did that only after having already said, we must find it's appropriate if a health hazard exists. So they've already determined that the health hazard is a necessary and sufficient condition, so the existence of controls is something that's not to be— Did you give us a citation to their opinion? I mean, the language speaks for itself, I assume. What are you referring to? For example, if you look, this is in our reply brief, but if you're looking at— In the reply brief? Yes, Your Honor. I think the — if I point you to the actual final rule, that's volume one of the U.R. petition appendix. So, the language, let me find it in the reply brief, I have the U.R., I have the volume. I'm sorry, could you repeat it? Sure. If you turn to the U.R. petition appendix, volume one. Good Lord. Page 196, this is the text of the final rule. Page 206A, I misspoke the first time. 206 says, we must find at the top of the page, we must find it appropriate to regulate EGUs under Clean Air Act Section 112 if we determine that a single HAP emitted from EGUs poses a hazard to public health or environment. They've said we must regulate. And the phrase must regulate means that when you get back to the availability of controls, then there's nothing left to be done. You've already said that we have to do it, so the availability of controls isn't doing any additional work. The overall— Is the government going to say that if the predicate for regulation exists, i.e., their emissions, then it's appropriate to regulate? That's what the government will say. I mean, that's—appropriate is a capacious term. And it was in the end of the government's—it's appropriate to regulate if there's an emission. It is a capacious term, but I think that cuts against the government because one of the things that's encompassed within the term appropriate is that it looks at all of the circumstances in the context of determining whether or not you're going to regulate costs is a relevant circumstance. So the very fact that it's capacious cuts against them. But the fact that they've said we must find it's appropriate to regulate means that this other— Well, are you saying that they didn't look at the availability of technologies? Is that what you're saying? That they thought that the availability of technologies was itself irrelevant to the determination? Is that your argument? We're not saying they thought it was irrelevant. They thought—well, I guess logically irrelevant. They thought that it's something they did look at. It's something that when they did the utility study, they examined the availability of controls. But then they said regardless of whether or not controls are available, if a health hazard exists, we have to regulate. So if I could give an example. I'm sorry. I thought they said only if it was necessary. Congress was motivated in not listing these sources because it didn't know whether the technology that was going to be put in place to control acid rain would reduce the hacks sufficiently so that regulation wasn't necessary or listing wasn't necessary. I had a different understanding of appropriate and necessary. Appropriate if there were hacks, but necessary only if those hacks were not sufficiently controlled by the other technology. The necessary—both of them looked at whether or not there was going to be an ongoing harm because both necessary and appropriate turned on the utility study. The utility study was something that examined what health hazards would remain after all the other regulation that's already been— The health hazard could have been low enough so that no standards were necessary. Well, they determined how severe—the severity of the health hazard. The severity went into determining whether or not a public health hazard existed at all. So they looked at the effects, and the only place they looked at severity in the final rule was determining whether or not a public hazard existed. Once there were enough health effects that there was a public health hazard, then they said we must regulate. And the fact that they said we must regulate, it's necessary to regulate, is exactly the same as what they said with appropriate, that we must regulate. Can I step back for a minute, General Lindstrom? Because it seems to me that this quest for very particular meanings attached to each one of these adjectives, appropriate means X and necessary means Y, if we step back a little bit, I mean, that kind of language is used all over the U.S. Code, and indeed that kind of language is used in our Constitution, the Necessary and Proper Clause. And as I understand what courts have done with that kind of language, that they haven't insisted that there be separate defined meanings for each of those words. When John Marshall was doing this in McCulloch, in fact, he starts off with the word necessary, and then he says, no, this is a phrase, and we have to understand what the phrase as a whole means. And why shouldn't we similarly say this idea that, you know, you can catch them in a redundancy or a superfluity, it's just not right because it's a complete phrase? I think there's two responses to that. First of all, under the Necessary and Proper Clause, if you look at what this Court did in Prince v. United States, it recognized something might be necessary and not necessarily proper. The commandeering of State legislation might be necessary to accomplish what you're trying to do. Exactly. We have separated out the two words and said something could be necessary and not proper. And what Marshall said was that necessary doesn't mean absolutely indispensable. It just means useful. That's quite different from saying that proper has no role to play. It can be necessary, that is useful to the federal government, and yet not proper. So why do you get to pick what it means? I mean, I thought in our agency law we repeatedly say if a term is ambiguous and there's no legal definition of appropriate, it's contextual, yes, but by definition if you're saying a word that's not self-defined, you have to look at in context and it's ambiguous. I don't think it's ambiguous in context. You can use the word appropriate in such a way that everyone understands what you're meaning. If I said we're going to take a group of people and we're going to go someplace and I want you to behave in an appropriate manner, and then I told you we're going to the library, everyone would know that that needs to be classified. So I look at this statute and I see them doing the first part, the part at issue, and the very next provision says in four years instead of three, do a mercury study that includes cost. I'm looking at it. I can very safely say one study doesn't use the word cost, the other does. The first one doesn't necessarily intend cost to be looked at. What is irrational or not plausible about that reading? Because all we have to find is a plausible reading to uphold the EPA's interpretation. It's irrational because they're taking the key statutory word and treating it as surplusage. The language I should have pointed you to earlier is on page four of our reply brief. So by treating them as doing exactly the same work, they're reading a word out of the statute, and Chevron deference doesn't extend so far as to say we can violate an ordinary rule of statutory construction, which is that independent words have to be used. But the word appropriate often is a signal that this question is what's fitting, and you have an expert agency, so the word appropriate, I think, is commonly used to indicate that the expert agency will do what it finds fit based on its expertise. So how you, well, you are saying that appropriate necessarily embodies a cost calculation, and yet this is a statute that uses cost to rec EPA to consider cost multiple, multiple times. Is there any case in all of our decisions where we have said even though there was no instruction to consider cost, EPA is required to consider cost? Is there any such decision? I don't think this issue, no, I don't think this issue has arisen the same way, where Congress has given broad discretion to an agency, told them to look at all of the circumstances, and the agency has said we're going to ignore what is an important part of the problem, and that's why in the way that Judge Kavanaugh looked at it, this is a problem under Chevron step one, two, or under the State Farm Doctrine, because agencies are supposed to not ignore an essential part of the problem as they engage in reasoned decision making. But I think what Justice Ginsburg is getting at, General Lindstrom, is sometimes what we've done is we've looked at silence, and we've said that given that silence, cost considerations are precluded, right? So that's the example in Whitman. Sometimes we've said silence still allows agency discretion. They can do what they want with it. But it's so far from our most closely analogous case, which is Whitman, to say not only is cost considerations, you know, not precluded, it's required when there's silence as to that subject. Now, if Congress wanted to require something, and clearly Congress required this in other places, Congress knows how to require consideration of costs, to get from silence to this notion of a requirement seems to be a pretty big jump. And, Justice Kagan, I don't think it is silent when it tells the agency to look at all of the circumstances, and the material circumstance in the context of the question that the agency has to answer is should we regulate under this section? Costs are part of the relevant material circumstance. I'm not even sure I agree with the premise that when Congress says nothing about cost, the agency is entitled to disregard cost. I would think it's classic arbitrary and capricious agency action for an agency to command something that is outrageously expensive, and in which the expense vastly exceeds whatever public benefit can be achieved. I would think that's a violation of the Administrative Procedure Act. Even without the word appropriate. And I think that's where there's overlap between what the state... Sorry. The study at issue that Congress commanded was simply a study, the administrator shall perform a study of the hazards to public health reasonably anticipated to occur as a result of emissions by electric, utility, steam generating units. So the study that was directed to be made was only of public health hazards. And then it says the administrator shall regulate these entities under this section if it finds regulation is appropriate and necessary after considering the results of the study. So if the study is directed only at public health hazards, it doesn't talk at all about cost. Just public health hazards. Why in the world would one assume that Congress was thinking about cost? Why didn't it do as it did with mercury? Make sure the study tells us how much control is going to cost. But it didn't do that. It just said tell us if there are public health hazards. And, Your Honor, it didn't limit the considerations that EPA was supposed to look at, only the study. But it says only the study. It doesn't say only the study. It says the administrator shall regulate if the administrator finds such regulation is appropriate and necessary after considering the results of the study. After considering the results of the study. The only thing that the study requires is an evaluation of hazards to public health. I'm not sure how you get to them having to do another step when the only step that's a prerequisite to registration is studying public health hazards. Well, first of all, even EPA doesn't think it's limited solely to the things that were studied in that utility study. They rely on environmental harms to justify. Does it say after considering only the results of the study? No, Your Honor. It doesn't say that, does it? No, Your Honor, that's correct. They have to consider the results of the study. It doesn't say they can't consider everything else. And the word appropriate seems to suggest that they may consider other stuff. Correct. So there's a study they're supposed to look at, but that's not the end of the analysis. They're supposed to do something else. That second step is to figure out whether it's also appropriate and necessary to regulate. So it didn't stop at just the study. And, again, EPA agrees they can look beyond the results of the study. They look at environmental harms, which is not particularly mentioned here and is mentioned in M1B. It seems to me that a very salient feature of the statutes that we have to interpret, maybe the most salient feature, is that Congress chose to treat power plants differently from other sources. It could have treated them the same way. And if it had done that, then the listing decision would not have taken into account costs. It would have been based on emissions, right? Or if it was an area source, it would have been based on effect of health alone. So what, if anything, can we infer from that, from the fact that Congress pointedly decided to treat power plants differently? I think we can tell that they're trying to create a different regime. They're trying to do something different here than they did elsewhere. Well, they were trying to create a different regime, but the reason is pretty clear on its face. They were trying to create a different regime because they thought that the acid rain program might have a real impact on what these electric utilities were doing. So they said, wait and see, and let's see how the acid rain program works, and let's see if we still have a problem to solve. And that's the reason why they put the electric utilities in a different category, isn't it? And that highlights why costs are significant. The acid rain program, in particular, was an economically based approach that was determined to regulate in a cost-effective manner. But the point is that the acid rain program didn't do what Congress thought it might have done, and it was still left with this issue of continuing harm from the electric utilities. And then, once that happened, it seems to me that it's natural to take a look at the rest of the statute and to say, let's regulate in a similar way to the way all other industries are being regulated. But if they wanted to do it in the same way, there would have been no need to use the phrase necessary and appropriate. They could have just simply gone to the 10-ton threshold emissions that apply to major sources and to the risk-based analysis that goes to area sources. So the fact that they use different criteria here in N1 as opposed to criteria that's listed in 7412C. They could have, but they might have thought, you know, let's take a look at the acid rain program, let's take a look at the problem that still remains, if any, and let's give the discretion to the agency at that point, because it will be years down the road in a different set of circumstances. But the discretion includes looking at the entire problem? I mean, again, the language in the circumstances requires looking at the material circumstances. And this ties into the State Farm Test. You have to look at all of the relevant circumstances if you're going to engage in a reasoned decision-making. You can't ignore an important part of the problem. If the reason for the separate treatment was the belief that the acid rain program would be sufficient at some point in time to bring emissions from power plants below a level that would result in their being listed if they were other sources, why would it be necessary to enact this separate provision, asking whether it's necessary and appropriate to regulate them? I don't see how that can be the explanation. Yes, Your Honor. They could have just had a three-year delay if that's all they were trying to do, as opposed to, and then go through the ordinary system? No, because they didn't know. They thought it might. They thought it might not. They were going to wait and see. It depended on how the industry responded to the regulatory requirements of the acid rain program. That still doesn't explain, Your Honor, why they chose to use different criteria, as opposed to just reiterating the criteria that are under 7412C, under the ordinary thing that applies to every other source. They're still trying to treat electric utilities differently. And I would like to return to one point about the acid rain program, which is that, again, if you're addressing emissions from electric utilities in a program that's specifically targeting electric utilities, as they did in the acid rain program, and that was entirely based on cost effectiveness, it makes little sense to look at what's remaining after you've already done that, and then to say in this area of diminishing marginal utility, we're going to say costs are irrelevant. That's backwards. Costs would be especially relevant when you're in the area of what's left over. When the statute refers to the emissions standards for the 12 percent of the best-performing plants, will the government say that that implicitly is a cost consideration? Is that their position? And if so, how do you answer? I expect they will. The way I would answer that is to say that that looks at plants across the range of how old they are. So plants that were built in 2005, for example, might have been built in such a way that they have technology where it was cost-effective to include certain control measures. But if you're looking at a plant that was built in 1960, imposing those same control measures on an older plant is something that would be a lot more expensive. It's the difference between renovating your house and building it, certainly, in the first two or four. It's a mandated base from which the government must operate. It is. And it seems to me like there's an implicit cost consideration there. Do you still say that's insufficient because? Yes, Your Honor, that's insufficient because it's not. I'm explaining why it doesn't necessarily take costs into consideration. The fact that some utilities were able to impose things doesn't mean it would be cost-effective for other ones to do it. Why isn't that taken care of? You have to take into account costs somewhere. And so the other side says there's room for that. Suppose that 25% of all electricity generators are near waterfalls. This is easy for them. But 75%, it's impossible, and they'll all go out of business, and we'll have no electricity. Imagine that were so. Could the EPA, under their current theory, take account of that? Well, I guess the answer you want to say is no, but they say yes they can. What about this 12% rule? A little bit earlier in the statute, it says the administrator may distinguish among classes, types, and sizes of sources. So if you really had this situation, you could say, look, 75% of the generators in the United States have this really old technology, and they'll all go out of business. And EPA could say, fine, that's a different class. Okay? I mean, if that were really true. So don't they have, through that provision and the 12% and the next one, the ability to take into account at least serious cost problems? Well, assuming they have the ability to take that into account, do they? Yes or no? Do they? Yes or no? No, I think. No, they don't. Why not? The reason that costs are not directly relevant to the first one is what I was explaining about the 12%. So in other words, the example you gave shows that some might be able to have the lower cost effective approach just because they're near a waterfall. So relying on the fact that 12% were able to meet this doesn't show it's cost effective. Fine. My point was, I wanted an answer to this. My point was, if in the imaginary situation that I've imagined, 20% of the generators, for whatever reason, can meet this pretty easily, the next 80% will require the entire gross national product to meet. Suppose that were the situation. You, I guess, could go to EPA and say, create of that second group a separate class, a separate type. For that's the reason it's so expensive. And therefore, the 12% doesn't apply to them because they're in a separate class. Now, my question is, can you legally make that argument? And they will take it into account. And that's what I want a yes or no answer to. And I think the answer might be yes in the future. But now we can't do it. Did you make the argument here now? My point is, if you're saying yes in the future, let's now go to this case and say, do you make this argument? I don't believe either side has made that argument. Well, has the agency made this obvious argument? The agency has not. Has the agency said, we're going to take costs into account. We're going to use this provision that Justice Breyer discovered. And that's what we're going to do. Is that how the agency issued its rule? It didn't. I never heard of this argument. Now, wait, because I still want to know a fact. The fact that I want to know, and you're familiar with this record, is did anyone on your side of the issue ask the agency to take costs into account brutally, roughly, crudely? Or did they all say, we want a cost-benefit analysis? I would like your characterization of the record on that point. Because reading what they said, it's about cost-benefit analysis, that paragraph. So that gave me the idea that maybe everyone interested in costs asked for a cost-benefit analysis. And I think the answer is that we asked them to consider costs. We thought a cost-benefit analysis is the ordinary way that a reasoned agency decision-making happens, not through some vague sense of what the costs are, but by doing an analysis. And they have said their entire position here is that we don't need to do that because costs are irrelevant. That's not something we have to consider under the statute at all. As I understand what happens, listing and standards are the only thing that you can generally appeal from. It's only a final agency action when the standards are issued. And I thought it was at the issuance of the standards that the government sometimes breaks up the sources and the amount of emissions that each type of source that Justice Breyer is talking about can have. So I think the listing is just of a broad category because we've had plenty of cases in this court where we've looked at the agency saying this type of source meets these standards. That type of source meets another standard. Isn't that the way it works? Well, that highlights why they're not treating this as a separate listing versus regulatory decision. They did both at the exact same time here. At the same time they made the necessary and appropriate finding, they also promulgated emission standards. And that shows this isn't simply a typical listing standard. Once they're listed, they are subject to minimum standards, aren't they? That's absolutely correct. And it's not up to the agency to calibrate the standards. Once they're listed, minimum standards apply, right? And the agency can have discretion as to whether to lift the standards further, but the minimums apply, right? That's EPA's position is we must regulate as soon as that has been made. If I could reserve my time for a vote. Thank you, Mr. Lindstrom. Mr. Brownell. Mr. Chief Justice, may it please the Court, I would like to make three points to supplement my colleague's arguments. First, regarding the nature of power plant regulation under the Clean Air Act. Second, the language of Subsection N1A. And third, the broader Clean Air Act context. To begin, power plants are the most regulated source category under the Clean Air Act, both before 1990 and after the 1990 amendments. The Court has talked about some of the programs. It's not only the Title IV acid deposition program, but a visibility best available retrofit technology, pollution transport programs targeted at power plants, and a variety of other control programs, both control and air quality. Mr. Brownell, I would think that that cuts the other way, that every other significant industry in the United States is subject to this program except for electric power plants. Your Honor, what it shows is that for the other industries, EPA estimated in 2010 that for all other industries, this air toxics program would impose compliance costs about $840 million. All of these other programs for power plants would impose compliance costs, EPA estimated, in 2011 about $10.4 billion. This single regulation now on air toxics imposes annual costs of $9.6 billion. And what does one get for it? There are three standards at issue here, and I think this is important to understand some of the questions that have been asked. There's a regulation for mercury, there's a regulation for nonmercury metals, and there's a regulation for acid gases. Most of the costs here, the majority, about $5 billion annually, are associated with the acid gas regulation, which the agency has concluded presents no public health risk, no public health concern. The agency said that our modeling has consistently showed that power plant-related exposures are at least an order of magnitude below the conservatively determined safe level. Now, in the Title IV program, Congress addressed pollution with acidification potential and required reductions of 9 million tons a year at about a cost of $1 to $1.4 billion. The acid gas program is projected to result in reductions of acid gases about 40,000 to 50,000 tons per year at a cost of $5 billion. What that background shows, Your Honor, is why Congress treated power plants differently. It asked whether it is appropriate to impose further regulation of a specific type, whether it's appropriate to impose regulation under this section on the most aggressively regulated industry under the Clean Air Act. Now, what the statute was- Can I take you back to Justice Breyer's first question? And the first question was about the way these categories work and how the categories enable the EPA to mitigate certain dramatic or onerous costs on certain segments of the industry. Because that's not an unknown provision of any kind, and indeed it seems to me that the provision very much cuts against your argument because EPA in some ways can't even figure out the costs until it makes those categorization decisions. But the aggregate costs, not just within each category, but the aggregate costs obviously depend on how EPA categorizes and subcategorizes. So you would have the EPA make the cost calculation before it really can, given the structure of the statute. Your Honor, the cost does factor into a variety of determinations that are made as part of the regulatory process. When EPA issued its notice of regulatory finding in December of 2010, it said this is non-final. EPA confirmed again, and this is at page 555A of the U.R. Petitioner's Appendix, that there is no final N1A determination or listing, and we are going to take comment on that as part of the rulemaking to examine Section 7412B emission standards. So as part of that, the agency addresses issues related to level of control, subcategorization, and at the end of the rulemaking comes out with a regulation that has certain characteristics and consequences. And here this regulation, what N1A says is, in light of the study, address whether such regulation under this section is appropriate and necessary for power plants. Now, it may be necessary to regulate something like mercury if there's a public health risk, and that's the only hazardous air pollutant for which EPA has calculated a quantifiable public health risk. But that may not be the appropriate regulatory regime if, as EPA has explained here, their view of the statute is not to focus on whether such regulation is appropriate, but whether listing of power plants under Subsection C is appropriate, just like every other source can. Can I ask whether that listing is, and the minimum standards that that imposes, are subject to the categorization device that Justice Breyer was asking about? Could the agency say, well, you know, we're going to divide these into categories, and since it's too expensive for certain 80 percent of power plants that don't have waterfalls nearby, we're going to exempt them from these minimum standards. Can it do that? No, Your Honor, not at the listing stage. But once the source category is listed at the standard setting stage, they could consider subcategorization. They should reduce it below the minimums? Not below the minimums. Your Honor is perfectly right. The minimums depend on the categories and the subcategories. You can categorize in such a way that the minimums will be up here, or you can categorize in such a way that the minimums will be down there. During the rulemaking, Your Honor, arguments were made about subcategorization, and EPA ultimately subcategorized the power industry with respect to one limited set of sources, lignite sources, with respect to the mercury standard. But otherwise, EPA's position is once listed, it triggers an obligation to issue emission standards under the subcategory. So it triggers an obligation as to some standard, but again, the minimum standard can vary dramatically depending upon how the categories and subcategories are set up. And because the minimum standard can vary dramatically, so too will costs vary dramatically. So you're having the EPA consider costs before the EPA can know what the costs are. Your Honor, if I could respond to that before the other question at the other end of the bench. EPA does know what the costs are through the rulemaking process in which it undertakes notice and comment with respect to both the N1A determination and the emission standard. Section 7607D1C of the Clean Air Act lists subsection N as one of the provisions that requires notice and comment rulemaking under the special Clean Air Act procedures. And this is why the agency explained that there's no final N1A until the end of the process, until notice and comment, and we've taken and determined what the costs are. And this is then confirming what Justice Breyer said. The point that you had an opportunity and apparently took advantage of it to tell the EPA that it should subcategorize this source and decided to subcategorize just one piece of it. So what you're really saying to us is it's not the listing, it's the way they've set up their emission standards that I disagree with. Because they could have decided that there were subcategories that didn't require a standard at all. I'm presuming that they could have said anybody by the water doesn't have to do more because they're already part of the 12 percent. We're going to do costs by everybody else. That's not by the water. Conceptually, Your Honor, I imagine they could have subcategorized away the entire industry, but that's not what they did in this rulemaking. And with respect to certain of the regulations... No, they didn't do it, but you're asking us. This is a challenge to a regulation that's only piecemeal. Because you're arguing that they should have considered costs, but they obviously did before they issued the standards. You can't look at the standards and the emissions and a listing in a case like this in isolation. Justice Sotomayor, if I could try the answer to the question once again. The subsection M1A question is whether after considering the results of the study, and I note that the study also looks at alternative control strategies for any emissions that may warrant regulation, the agency determines such regulation is appropriate and necessary. So the focus of the determination is not on listing and whatever may flow from that, but the regulation that the agency decides to apply to address the remaining public health hazard that is identified from this source category. Mr. Brownell, did EPA say we are not going to take costs into account at the listing stage, because we will take costs into account through this subcategorization possibility that's being discussed? No, they did not, Your Honor, as I recall. The record of preceding this discussion of subcategorization has come up in the briefing. Why didn't they write? You see what's wrong? These questions are difficult because they're so hypothetical. It isn't true that 50 percent of the industry will use up all of those domestic products, et cetera. But they wrote this thing in a way that sounds as if even if that had been true, they wouldn't have taken it into account. So what they say is the EPA does not believe it is appropriate to consider costs when determining whether EGU is okay. You see the problem? So for me, and what the SG is telling us, don't worry. Maybe they should have written. Knowing what we know and what is undisputed so far, we don't think that the cost problem is big enough for us to warrant a cost benefit analysis or other consideration. Okay? Then they've taken it into account. So there's no problem. If it's within, it's at the reasonable. No, they didn't write that. So that's why I'm looking to see is the, it's really the SG, but I mean is there really a different way that they could eliminate this horrible scenario if it existed, which it didn't? Do you understand what I'm driving at? I'm trying to get your best answer on that. And I want to emphasize that this is not an argument about whether or not to regulate mercury where there's been an identified public health risk. It's whether the regulatory regime that has been defined here under Section 7412, which the government says is the listing that applies to all other source categories and the D3 minimum control technology standards that apply to all other source categories, is the right way to do it. Regardless of how you subcategorize, it's going to have tremendous impacts as a result of acid gas regulation and for a pollutant that presents no public health risk. We can't uphold an agency rule on a ground that they didn't adopt below, right? That's correct. Under Chenery? That's correct. So is your understanding correct that this is not an argument, a basis for decision that they adopted below? That's correct, Your Honor. My understanding of the basis for the decision below is that costs are irrelevant in our determination under M1A whether or not to regulate this source category under the typical subsection 7412 regime that applies to other source categories. But they're exactly right, Mr. Brownell. I mean, the agency at that point in time was only answering the very first question, the very threshold issue. And at that point, the agency said costs were irrelevant. But costs become relevant later in the analysis and in a variety of ways, through the 12.5 percent, through the 12 percent, through the categorization and subcategorization, through the determination whether to raise standards even higher. So costs, costs, costs later. But as to this particular thing, the agency said, yes, here we don't consider costs. We could, but we don't want to because there's all this potential for costs to come in afterwards. And, Your Honor, it's costs, costs, costs under the statutory criteria that Congress provided for setting D3 control technology standards and then having to find those standards. At the end of the process, the agency finalizes its M1A determination in light of the costs and impacts and other factors that are mentioned. Do you think that whenever the term appropriate is used in a regulatory context in the Clean Air Act that it demands a cost-benefit analysis? Your Honor, when you say in any context that's so broad, I don't think that I can say that it would require cost-benefit in any context. But in the specific context here where the focus is on whether such regulation is appropriate or necessary, that regulation has certain characteristics and consequences that we have talked about this morning, including the fact that it imposes on a pollutant that presents no public health risk $5 billion a year. Before you finish, because your time is up, can you clarify for me why this is, at this stage, something that we should be concerned about because there is this regulatory impact assessment that has said that the benefits vastly exceed the costs, and that's an impact analysis that has gone through the OIRA process and OIRA concluded that EPA appropriately calculated the cost. Yes, Your Honor, Mr. Chief Justice. The co-benefits, all of those benefits are co-benefits. Only $4 million to $6 million are associated with hazardous air pollutants. Those co-benefits that are in the regulatory impact analysis were not considered as part of the regulatory determination for good reason because they're important questions regarding their legal importance and relevance under the proper standard. And what I mean, Your Honor, is that PM2.5 is the pollutant, fine particles that is associated with these co-benefits. That is extensively regulated under the National Indian Air Quality Standards Program. In fact, those air quality standards were only recently revised to be tightened, and in the context of that proceeding, the agency found that the low levels of exposure for these co-benefits did not produce effects or risks that were of regulatory significance because they're too uncertain. So there are serious questions about legal relevance and importance. Thank you, counsel. Thank you, Mr. Chief Justice. Mr. Chief Justice, and may it please the Court, EPA's interpretation of Section 7412N1A should be affirmed for three basic reasons. First, it is the most natural and certainly a permissible reading of the statutory text, which directs EPA to focus on health concerns and doesn't mention costs. Second, it harmonizes the provision with Section 7412's structure and design because it applies the same regulatory logic to power plants that Congress directed EPA to apply to regulate hazardous air pollution from every other type of source, and that is decide whether to list for regulation based on health and environmental hazards alone and consider costs in setting emission standards. And third, as a matter of common sense and sound government practice, it was certainly appropriate for EPA to list power plants for regulation based solely on health and environmental hazards because that reflects the approach Congress chose not only in 7412 in this regulatory program, but in all of the major regulatory programs under the Clean Air Act. Could I go to your three arguments, your first point? You concede, don't you, that EPA could have interpreted this statutory language to allow them to consider costs? I think EPA read it as the best interpretation of the statute was it didn't provide for the consideration of costs at the listing stage. But under Chevron, if you adopted a regulation that said appropriate and necessary allows us to consider costs, you think that would be appropriate? I think the phrase appropriate and necessary doesn't, by its terms, preclude the EPA from considering costs, but under Chevron what the EPA has got to do is explain the justification for its reading of the statute. That's what it did. Since you're dealing with the term, I think it's said as capacious as appropriate, and since you could have issued a regulation allowing the consideration of costs as appropriate, you're saying that the agency deliberately tied its hands and said we're not going to consider something, we're going to issue a rule saying we can't consider something that we could consider otherwise. No, I think the best way to think about this, Mr. Chief Justice, is what the agency did was decide that it was appropriate to approach the question of whether to regulate power plants in the same manner that Congress found it was not only appropriate, but mandated to address the question. I understand your argument that they could have done that, but I think it's unusual for an agency to say when they want to do something that that's the only thing we could do. Agencies usually like to maintain for themselves as much discretion as they can, and it strikes me as unusual that maybe the agency could go ahead and not consider costs, but to say that we're prohibited from considering costs under the phrase appropriate, it strikes me as very unusual. Well, I don't think so, Mr. Chief Justice. It didn't say we're prohibited from considering costs. It said we are not going to consider costs at the listing stage in making the decision about whether power plants should be listed for regulation under Section 7412. I thought the rule was that the phrase appropriate and necessary did not allow them to consider costs. And the appropriate and necessary goes to the question of whether power plants should be listed for regulation under Section 7412, which would then kick in the regulatory mechanism by which standards were set and costs are considered at the standard. I had the same question as the Chief Justice. Let me just ask one more time. Could this agency reasonably have considered costs at Stage 1? I don't think the statutory text unambiguously forbids them from considering costs, but they determined that the best reading of the statutory text is that power plants, because after the study was conducted that Congress required, EPA determined that power plants are no differently situated than any other source of hazardous air pollutants regulated under Section 7412 and for every other source of hazardous air pollutants. What Congress mandated as appropriate was that you do not consider costs when you decide whether to regulate. You only consider health and environmental effects, and then you do consider costs under Section 7412B when you set the emissions standard. How is that consistent with this statutory scheme? If your argument is that EPA's only reason for doing this is that it wants to treat power plants the same as other sources, we know that that's what Congress didn't want, or it would not have enacted a separate provision for power plants. I agree, Justice Alito, that Congress proposed different treatment for power plants, but that doesn't answer the question. That just asks the question. The question is what different treatment did Congress prescribe? Now, what petitioners are arguing is that what Congress prescribed and mandated was a cost-benefit analysis that does not apply to any other source of hazardous pollutants, but that is not what the text of the statute says, and it is not what the history reflects. What the text of the statute says in N1A is that even though for every other source, regulation would proceed immediately, for power plants, there was uncertainty about whether power plants emitted hazardous pollutants at a level that would cause problems, whether the acid rain regulations would solve the problem, and whether there were alternative control strategies available. If I may, I'm sorry. Yes, sure. EPA, what Congress told EPA to do was to study those three things. Those go to health considerations, and then once EPA made a judgment about that, it was to decide whether to list power plants for regulation as appropriate, whether it was appropriate and necessary to list them for regulation. This is what I don't understand about your position, that Congress's decision to treat power plants differently, it seems to me, reflects the fact that Congress wanted at least to hold open the possibility that power plants would not be listed even if their emissions exceeded the levels that would result in listing for other sources. I don't see another reason why they would treat them differently. Well, I think— So if you—I can just continue. If that's—you may disagree with that, but it seems to me that that's a necessary inference from the statutory scheme. If that is the case, what factor might Congress have thought would justify allowing power plants to emit more than would be permitted if they were other sources? Now, the petitioners have an explanation, which is costs, and they say that power plants have to bear a lot of costs that other sources don't have to bear. In particular, the Title IV program, their emissions might exceed the otherwise permissible limit because they have participated in the cap-and-trade program, so they have contributed to the reduction in emissions in that way, in a way that wouldn't be reflected in their own emissions. So that is an explanation. Cost is what's missing. I don't know what your explanation is. So I know you're going to ask me to accept the premise, but I can't accept the premise because both the text of N1A and the legislative history tell you what considerations Congress left open, and they were all related to the possibility of health hazards. No one—the argument that Your Honor just posed is not in the legislative history and it's not in the text, and if Congress really thought that, then what they would have said to EPA is, take the—put the pause button, take the three years and study, don't subject them to the same schedule as everybody else, and study the cost problem. They would have expressly told EPA to study cost, and they did not do that. If all they were concerned about was health, why wouldn't they impose on power plants the same standard that's imposed on area sources? Just ask EPA, determine whether there is a—I forget the exact term, but is there an effect on public health, a significant effect on public health, whatever, if so, list. I think they came very close to that because what they said— after considering the results of the study, and they told EPA to study things that went to health hazards. And so that comes very close, it seems to me, and the reason I think it used the appropriate and necessary language rather than the language Your Honor suggests is because I think Congress, when it was legislating here in 1990, understood that there might well be uncertainty at the end of this—of the analysis that Congress directed EPA to undertake. There might be uncertainty about the projected effects of the acid rain regulations because I do—I do think it's important to clear up— I think that's what the legislative history said. I do want to clear up a misconception about that, I think, which is that the way these acid rain regulations unfolded, they were put in place in 1990 at the same time as 74-12, but they were to unfold over a 10-year period, five years until the first stage and then five more years until the second stage. So EPA was going to have to make a long-term projection here about which there could be some uncertainty. And so I think what Congress was saying to EPA was you may need to exercise your judgment here. And, in fact, EPA did exercise its judgment because it concluded— There are two parts to this argument. One is, well, what were they thinking of with the word appropriate if it wasn't cost? All right, I see your answer to that. Think about that. But the second, which I think your argument very much depends on, in my mind anyway, is, well, don't worry, because there is a way to take into account cost. So if, in fact—I mean, you know, it's a lot of money, $9 billion. And if you divide it by the population, you have $30 a person or a family of four of $120. That's a lot of money for people, for some people. And you should say, gee, you couldn't take it into account ever. It would have been $500. You know, it begins to look a little irrational to say, I'm not taking it into account at all. But you say, never fear, because they will take it into account when they set standards. And at that point, I read the thing about the 12%. Now, I've got the word similar source, which then can refer me back to the categorization of two things earlier. As I say, maybe. And then I have, aside from that, hey, here's what you do. When you're regulating, you look at the top 12 generators, and that's the minimum standard. So they might want to say, hey, that's not right. I mean, it's right. It says it. But if you go to the bottom 50 generators, you're going to see it's not going to cost $120 per family. It's going to cost $1,000 a family. And we have the APA saying, we won't even look at that. At that point, I begin to say, oh, my goodness, why? Why won't you even look at it? You could say it isn't true, but why won't you even look at it? And now the answer seemed to me to be in that word similar source and the classes and the subclasses, because were there such an argument? Maybe the EPA could say, don't worry. If there is such an argument, which there isn't, we have the power here under the statute to take it into account. Now, you know where that argument came from? From discussion and thought in my chambers. Now, maybe it came out of the briefs too. But is what I say right? Can the EPA take that into account? Or do they have to just blindly say, if it's the top 12, that's for everybody no matter what the cost? In which case, they can't take it into account ever, except for the word appropriate. Now, that's the argument. You got it, my argument. I got it. I'm going to make three points in response to it. The first one goes to the empirical situation in this case, but then I'm going to get to the theoretical question you asked. The first is this. $9 billion is a big number. This is an industry with $360 billion a year in annual revenues. So you're talking about 2 to 2.5% annual revenues, of annual revenues. And what Congress and what EPA concluded, Revenue doesn't talk about profit. Right. But this is a cost, and the cost is about 2.5% of revenues. And then what EPA concluded was that about 2% of electrical generating capacity could go offline as a result of it being uneconomic. So it's not a 50% or a 78%, 88%. Right, and as soon as you've said that, you've taken costs into account, which is what they say they wouldn't do. But then, now let me talk about the way EPA, under this regime, does take costs into account. The first point I would make is that the situation that you're on in describing the hypothetical is a quite unusual one. In the normal case, the 12% rule, it's a technology-forcing rule, the kind that Your Honor discussed in your concurring opinions in Whitman and in Entergy. It's that kind of a rule. And in the normal case, it's not going to have that effect. It means that this percentage of the industry has been able to meet this without an operator in an economic matter, and Congress is trying to force the rest of the industry to catch up. And as we know from multiple experiences, as Your Honor identified with respect to catalytic converters and motor vehicles and with respect to acid rain, it turned out that the cost was vastly lower on industry than EPA anticipated. It would be that there's a very great tendency to overestimate costs in that situation. But then the third point is Your Honor's point about subcategories. And Section 7412C1, which is the provision that governs the listing of categories, it mentions the availability of subcategories. And, in fact, the last sentence of it says nothing in this... Do you have that page, sir? I'm sorry. I'm sorry, Justice Kennedy. It's 35A of the appendix to our brief. And this is C1, and it's the last sentence. And it talks about the EPA's authority. It says nothing in the preceding sentence limits the administrator's authority to establish subcategories under this section as appropriate. And, in fact, that is what... Can those subcategories apply to the minimum standards? Yes, that's how it would work because you identify the category, and then you generate the standard based on things in the category. But I thought the standards are automatic. There are certain minimums once they find, on the basis of the study, that these things should be listed. I thought there is an automatic requirement imposed on them, which is the reason they're complaining. The requirement depends on how you categorize. And so if there were a situation in which one segment of the industry was so vastly different from another segment of the industry in terms of its economics, in terms of its technology, then EPA would have the authority to break those into two separate subcategories, and then you don't calculate the best-performing 12 percent, which is what the standard is, until you know it's the best-performing 12 percent of the people you've put in the category. The language that does that is the first sentence of 3, which says the maximum degree that is deemed achievable, not less stringent than the emission control, that is achieved in practice by the best-controlled similar source as determined by the administrator. That's what allows him to break it into categories and apply the minimums to the similar source. Am I right? That's correct. And EPA did that in this case. It broke out power plants that generate power burning natural gas. And it said that's a separate subcategory. Where in the record, where can you point me into the record of where this argument was made or considered by the agency, as opposed to Justice Breyer's chambers? Because it's a very important principle of administrative law that we will only uphold a rule based on the arguments that were considered and addressed by the agency. So I — as I said, it's not something I recall from — It's a course rule, Mr. Chief Justice, and you're exactly right in stating that principle. But our argument in this case is that this question here is under N1A. N1A says EPA shall regulate under this section if it determines that such regulation is appropriate and necessary, and therefore that when EPA makes a judgment to regulate under this section because it's appropriate and necessary, EPA is triggering all of — Well, you're just saying that the argument is right, not that the agency made it. Well, it's — I guess what I would say about — It's not enough that the argument be right. The agency must have rested its decision on the point. I think the agency in the order I'm being challenged here did use the approach. But beyond that, whether or not — It would be one thing if this were a case in which you had a situation in which EPA — where EPA faced a situation in which 50 percent or 75 percent were going to face vastly uneconomic consequences. But this case didn't present that situation. EPA made a judgment about — General Lippert, can I simplify your answer for you? Yes, I'd be delighted if you did that. All right. Basically, you have consistently in your brief, and so has the other respondents, basically said at the listing stage we don't consider costs, we consider it later. And everybody gave a few examples. Whether this example was given or not is irrelevant. The issue here was do you have to do it at listing? It's only some of my colleagues here who are concerned that when you issue standards, you never consider costs. And that's exactly right. The question here is whether EPA has got to conduct a cost-benefit analysis when it does the listing. And the logic of the statute of 7412 doesn't operate. At that point, the game is over. No, I don't think it is, Justice Kennedy, for several reasons. First, the standard under Section 7412D for setting emissions standards, once you've decided to list, that, as Your Honor's question to my friend General Lindstrom pointed out, that does take costs into account in the sense that a segment of the industry can operate economically. Excuse me. I didn't understand that. I thought that there were automatic requirements imposed once the plants are listed. Once EPA lists and defines the category for listing, then the automatic requirement that is applied is that everyone in the category has to match the performance of the best 12%. Where did these categories come from? I really don't like the fact that your friend on the other side was not permitted to mount an argument in opposition to this categorization theory that Justice Breyer's chambers devised. Usually we have arguments on both sides. This is an argument I never heard of, and I'm not sure it's right. But I certainly didn't know the agency to say, oh, we're just listing, but we're going to categorize the listing. I understand your point about the focus or non-focus on subcategories, but the point that we're just listing, we say that over and over again in our brief. And, in fact, the petitioners concede, and this is at page 5 and 6 of the UARG reply brief, that if this is just about listing, then, of course, costs are irrelevant. But it is just about listing. That is the way the statute works. I ask that you respond to the fairly dramatic disparity, your friend on the other side saying that 6 million benefits, 9.6 million costs. You respond with a different calculation that looks to, I call them collateral, the ancillary co-benefits. And then the argument is raised that that's not quite proper because you're using the HAP regulation to get at the criteria pollutants that you otherwise would have to go through a much more difficult process to regulate. In other words, you can't regulate the criteria pollutants through the HAP program, so you get, okay, here we've got this tiny bit of mercury, and so we're going to regulate, and that's how we're going to get additional regulation of the criteria pollutants. It's sort of an end-run around the restrictions that would otherwise give you less control over the regulation. What's your response? Let me address that. There's several points, and I actually need to make all of them, I think, to make this clear. The first point is that that's not an argument that any party has raised. One amicus brief raised it, and it was averted to a particular argument. Well, my team founded it. And here's the problem with the argument. The problem with the argument is that it has two problems. One is that once EPA concludes that a source emits a hazardous pollutant, and here EPA has concluded that these sources emit mercury at levels that are unsafe. I don't think petitioners dispute that. And by the unambiguous terms of Section 7412D, EPA is under an obligation to regulate all hazardous pollutants that the source emits. And that's the D.C. Circuit in a case called National Line 15 years ago. I understand how the end-run works. I'm just questioning the legitimacy of it, because what they would say, okay, you found a hap that you want to list, but you ought to consider only the benefits of regulating that. You shouldn't consider the bootstrapped benefits that should be addressed through the other side. I guess the next point I would make, Mr. Chief Justice, is that it's not an end-run and it's not a bootstrap. This is a well-regulating the surrogates. Regulating the surrogates is a well-recognized methodology that goes back decades, that EPA has used for decades, that the D.C. Circuit has upheld for decades, that that is a perfectly appropriate way to deal with getting at metals and other pollutants that would be hard to get at directly. And, in fact, the very argument that Your Honor is positing here as an end-run is one that was made in this same National Line case to the D.C. Circuit 15 years ago, and the D.C. Circuit rejected. So what EPA is doing here is basically what the D.C. Circuit said that Section 7412D required them to do with respect to regulating every hazardous pollutant that the sources emit, and what the EPA has said, and what the D.C. Circuit has said for decades, EPA is permitted to do, and that it isn't an end-run at all. This is just a normal way in which you regulate. But the issue that I think raises the red flag, at least, is that there's such a tiny proportion of benefit from the HAP program, and such a disproportionate amount of benefit that would normally be addressed under the criteria program. So it's not just that, yes, when you're regulating one, it's a good thing if it also has benefits with respect to other pollutants. But if your basis for regulating, what is the benefit from the co-pollutants that you get? Oh, it's many, many billions. Do you remember how much it is? It's sort of $30 to $90 billion. $30 to $90 billion. The benefit from mercury is, what, $4 million? Well... So you say, we get to regulate this. We get to regulate it because there's a $4 million impact on mercury. But when we do that, we get to regulate $35 billion, in a way that gives us $35 billion in costs on the other side. So I understand the idea that you can have, you know, it's a good thing if your regulation also benefits in other ways. But when it's such a disproportion, you begin to wonder whether it's an illegitimate way of avoiding the different, quite different limitations on EPA that apply in the criteria program. So I don't, I really don't think that. I understand that Petitioners have put the case that way, Mr. Chief Justice. I don't think that's a fair way to put it. What EPA did with respect to mercury was quantify one of the public health benefits. It did not quantify many of the other public health benefits. If you look in the Joint Appendix, and I think this is about page 910 to about page 940, there's page after page of charts in which EPA has listed the other benefits that come from regulating mercury and the hazardous substances that it didn't try to quantify. Part of the reason it didn't try to quantify is because quantifying those kinds of benefits can be very difficult and challenging. And frankly, that is one of the key reasons that Congress adopted, not only in 7412, but under the NAAQS program and the Motor Vehicle Program and the New Source Performance Standard Program, the approach it did of not taking costs into consideration at the listing stage, but only at the regulatory stage. Could you tell me about the natural gas? You were cut off earlier. Sure. EPA reached the conclusion that natural gas power plants ought to be a separate category and because they didn't generate public health problems at the level that would make them comparable to the coal and... So they were part of the listing, but not... Correct. They were part of the listing, but they were not regulated because you're discerning... Yes, Your Honor. Can I ask you another question about these subcategories? Your argument is that under the last sentence of 7412C1, the EPA can create subcategories based in whole or in part on cost. Is that right? I think it's more subtle than that. I think that if... It's not just that provision. There are numerous provisions within 7412 that allow for subcategorization. But I think that if there is such a vast difference in the technologies that the group of entities is using, that there would be that vast a difference in cost that there might well be a basis to treat them as a different subcategory. Why did the EPA say that? I thought the EPA said, we are not going to take into account costs with regard to the listing. Now, they could have said, we're going to take into account costs as to whether some categories should be listed and other categories shouldn't be listed. That's not what they said. They said, we will not take into account costs with regard to listing. We list all of these subcategories. Here's what they said, and here's why they said it, and I think this is critical. What they said was that we think it is appropriate with respect to power plants not to consider costs at listing and to consider costs at admission standard setting. And the reason we think it is appropriate is because that is the standard, that is the regulatory logic that Congress deems not only appropriate but mandatory for every other source category. And so one would have to conclude then that what Congress said was mandatory and therefore necessarily appropriate for every other category was inappropriate That's how I understood their argument entirely, and I did not understand it to say, you know, we can exempt some people from these minimum standards because we categorized them differently. It definitely was not allowed. That is their main argument, but I do think the point of the logic of EPA's position here is that you make the listing decision and then you regulate, as N1A says, under Section 74.12, and these are provisions in Section 74.12 that give EPA the authority. As I understood it, please. Are there regulations that set out the criteria for creating these subcategories? I'm not aware that there are. I just don't know the answer to that, John. So, well, without them, we really don't know to what extent if any costs are taken into account in creating subcategories, do we? Well, I think that, no, I do think that it's going to be based on differences in technology and operation, I think, from which you might be able to infer costs, but that's hypothetical in this case because this is not a case in which EPA needed to confront that question, except with respect to natural gas-fired power plants, which they did find to be so different that they wanted different treatment because it didn't have the kind of problem that Justice Breyer's hypothetical raised. You didn't have that kind of problem. You didn't need to face this issue in these cases, so EPA didn't. And if I could just make this point, because I think it's quite critical. Given that 7412's regulatory logic provides for listing based on health, emission standards setting based on cost, including consideration of cost, and given that that's exactly the same logic under the NAAQS program, it's exactly the same logic under the motor vehicle program, it's exactly the same logic under the new source performance standards program, that if Congress intended to mandate that EPA cut so deeply against the grain and make such a radically different approach with respect to this one category of sources, you would expect to see very clear legislative language to that effect. You would expect to see a direction to EPA in 7412 to study costs before making this judgment, General Verrilli, let me ask a question about costs. There are economic costs. There are other costs. Is it the agency's position that no cost can be taken into account? For example, it may find that a particular material has an effect on health, but it may find that eliminating it will have other effects that are even more deleterious to health. Could that cost be taken into account? If I may answer, Mr. Chief Justice, I think that cost would be taken into account in the OIRA regulatory impact analysis. Not for the list. But not for the list. Not for the list. I think that's right. Thank you. Thank you, General. Thank you. Mr. Smith? Mr. Chief Justice, and may it please the Court, we agree with the government that the EPA was not required to engage in a cost-benefit analysis before making the initial listing decision to regulate hazardous pollutants emitted by power plants applying the appropriate and necessary standard. And I certainly want to acknowledge at the beginning, clearly Congress did think that power plants needed to be treated differently. But what did they give them? They gave them a three-year pause in which the EPA was instructed to take account of the health effects of the particular pollutants emitted by power plants. They did this under an appropriate and necessary standard. And if I could address the issue of what those two words mean in the reading of the EPA, I would refer the Court to pages 226 and 227 of the National Mining Association Certification Appendix. I don't think it's necessary to read from it, but what the EPA said consistently throughout this record is we looked at two things. There was a claim made in the legislative history that these chemicals are simply not harmful enough to require any further regulation, that their effects are negligible. And they looked at that issue under the appropriateness rubric. And they said these are harmful chemicals, particularly mercury. In addition, under the appropriateness rubric, they looked at the question of whether or not there were technologies available to regulate them. The necessary rubric was used to look at the post-Clean Air Act, post-acid rain health effects that would persist. So they said, A, these are harmful chemicals, harmful pollutants, and B, under necessary, they will continue to be harmful after the acid rain program is kicked in. And that's how the EPA saw the two different words, and it's a perfectly logical way for them to perceive. Who would have guessed? I mean, that seems such an artificial division of necessary and appropriate. What? Describe it again. I really didn't understand it. The claim necessary means what? The claim... Everybody concedes the necessary means that there will still be health problems after the acid rain program kicks in. Appropriate means? The appropriate was intended to meet the claim made by the industry that these chemicals already are sufficiently harm-free that we don't need to regulate them, regardless of the effect... Why isn't that part of the first one? Well, it could have been done that way. They read them separately, Your Honor. I see. That's the way the government read them. It's a silly way to read them, I think. Perhaps so, Your Honor, but the key thing is the issues that they were directed to study, the issues that were then supposed to control the listing decision were the health effects of the pollutants that come out of these power plants. And they then deferred the issue of considering a cost to the second stage, just as occurs with every single other source of the same 189 hazardous pollutants that they were dealing with. If I could pause here to just clarify one thing about what happened at that listing stage. Natural gas-fired plants were not turned into a category. They looked at the health effects of natural gas-fired plants and said, we are exempting them entirely from regulation under this section under the appropriate and necessary standard because they simply don't emit these chemicals in more than trace amounts and there's simply nothing for us to regulate. So natural gas plants get taken out at the listing stage. We then have coal-fired and oil-fired plants as to which they begin to apply the subsection D standards, standards which were designed by Congress to limit the emission regulation to reasonable amounts, designed because the floors are, in fact, limited to what has already been achieved by comparable plants in the same category. Now, there was some question raised about whether or not this categorization was something that EPA recognized it could use to affect the emission standards and make them reasonable. In fact, as Mr. Brownell acknowledged, they did create a category, separate category, of coal-burning plants that burn lignite because it turns out that none of the lignite plants could meet the standards that would otherwise have applied with respect to mercury if they were in the category with the other coal-fired plants. They then, through this whole process, looked at the issue of categorization. They started out with two coal-fired categories. They ended up with, and they started out with one oil-fired category. In the final rule, there are four separate categories of oil-fired plants depending on what they burn and how they operate. So this whole process of separating out these categories to produce emission standards that make sense and are practicable was built into that process under subsection D. How can we tell? Could you just clarify for me that the categorization happens after the listing? Is that correct? Yes, Your Honor. What they listed was all coal-fired plants and all oil-fired plants but no natural gas plants. They then go to the question of what emission standards should occur. With respect to other sources, that's sometimes years after the listing decision is made. But they then have a process of saying, what are our categories and what is going to be... We have to get information at that point. What are the top 12% of the category would, in fact, are their emissions? They have to report in. They make a calculation about that. Aren't these just requirements above the minimums that automatically apply? No, Your Honor. The minimums are established by what the top 12% in the category that the EPA has to specify. I'm saying the categorization that allows you to reduce some people and not to reduce others, that applies to requirements above the minimums. Your Honor, the minimums are the things that are set by mathematical calculations from the categories. Everything else above the minimums is done taking cost into account under D2 expressly. Yes. So the only thing that is done using not express consideration of cost but consideration of cost indirectly but basing the regulations on what the top 12% are doing is the minimums. And the minimums are then altered depending on what categories you establish. That is the way that the regulation has worked for all of the sources that they regulated. That is the practical mechanism. But just to clarify that just for my own purpose, Mr. Smith, it's just you categorize one way, the minimums are down here. You categorize another way, the minimums are up there. That's correct, Your Honor. It can make a huge difference in terms of what the minimums are. Right. And there's a notice and comment process. They put out empirically what we found about what the top 10% or 12% are doing. They then get commented and as they happen here, they make different categories in the final rule adjusting. How can we tell the degree to which costs are taken into account if they are at all without knowing what the criteria are for creating these subcategories? Well, Your Honor, it's in the statute that they create the categories and applying this 12%. I don't... No, no. I thought the 12% came into play after you've created the categories. Right. It's part of the subcategory. So how do I know how they create the subcategories? Well, you can see in the notice of proposed rulemaking. They say, here are our proposed categories. Two coal categories, one oil category. Then they get comments about how we are not... What happens then is the people that comment say, we're so different from that category. We have these special problems. We need our own separate categories. And it's a rulemaking after the rulemaking that applies to the listing. Is that right? Totally after that, Your Honor. It's a second phase. It happened here that they announced it all at the same time, but only because there had been an 11-year delay. You confirm it was not made up in my chambers. Although... The fact is... They did a wonderful job figuring it out again, Your Honor. You're right. The brief said, the SG's brief, at the second stage of the regulatory process. That's what it said. A few pages later, they have the statute. It's a statute that states that EPA has to consider costs for generating generators. So not surprisingly, I read the statute. And reading the statute leads me to think it works along the lines you just said. That's correct. If you did have the most expensive set of generators in the world, you would ask EPA to create a separate category for them, in which case the top 12% would no longer be in your category, and you wouldn't have to do it. Now, what I'm asking is, if you think it's the system, that's what I read in the statute, I think the SG thinks that's the system, but is there a treatise? Is there an explanation that EPA has put out so that it is clear that it was not made up, that it's clear that this is the system that they follow? Would you refer me to a source? Your Honor, the only source that I can refer you to, and perhaps the government can supply something after, is the notice of proposed rulemaking and the final rule, in which all of this categorization process is laid out in exquisite detail, and you can see the comments that come in and say the categories don't work, we need different categories, and they then produce different categories. Is there something in the administrative record where EPA adopts that, when somebody says, you're not considering cost, that's a bad thing, we're going to go to the Supreme Court if you don't consider cost, and EPA says, oh no, we're going to consider cost, we're going to consider cost when we categorize the power plants. Is there a reference to the administrative record where there's something like that? Well, they certainly said, Your Honor, in the notice of proposed rulemaking, that we interpret the listing decision as being something that is based only on health and not on cost. Yeah, that the listing decision is not based on cost. Right. I want to know if there's anything there that says, but don't worry, because we're going to consider cost through the categorization process. I don't know whether they said that explicitly. That was so implicit in the whole system. This is a system that's been in operation for other sources since 1990. Well, implicit usually doesn't work when you're talking about an administrative record. Your Honor, they gave everybody the opportunity to attack the categories that they in fact proposed based on the argument that they were impractical for them. But I thought your position was that you didn't need to consider cost at the first step and that that would include your initial category. No, Your Honor, the categories are at the second step. The categories have the only thing that they did at the first step is say But you didn't take that second step. They did, Your Honor. They categorized oil-fired plants into four categories. They categorized coal-fired plants into various categories. And that was all done through a notice and comment process, which led them to different emission standards. Are you saying that that was done based on cost? Yes, Your Honor. It was done on what was feasible for these different technologies. How much money did that save? I mean, do we know how much of the $9.6 billion cost would be reduced by this categorization? Well, Your Honor, that's the problem here. I don't have that calculus, but I would point out that it's important to recognize that something like 90 percent of that $9.6 billion that $9.6 billion, has now already been spent. And the industry has not experienced the kinds of upheavals that we've had in the past. We've had upheavals that are being described. The rule takes effect in the middle of April, and so the idea that the result here was somehow ludicrous or outlandishly expensive is belied by the fact that the industry is bringing itself into full compliance. Instead of going to jail. Is that it? I mean, it was ludicrous, but it had to be done. Well, the other thing I would say is the idea that the $4.6 million benefit is that proper comparator is wrong on so many different levels. First of all, that was a single, one single health benefit was related to mercury. They didn't Is the $9 billion a year recurring annually or are you saying that most of this is capital investment? Most of it is amortization of the capital expenditures that have already been made. That's in all the briefs from the petitioners. There will be something like 40% will be operating costs and 60% is amortized capital costs. The industry has been able to do this and the situation now is we're ready to finally have national standards, which means the states that have been regulating in this area very strenuously won't continue to have mercury flowing across state lines as they have. And we have this national highly competitive electricity market where some companies have marginal costs that reflect that they're in compliance and some don't. And that is a very difficult situation. that's one of the reasons. I just want you to finish your thought. The 4 billion that they're referring to is only mercury. The agency did not quantify all of the other costs for the other hacks, correct? Actually, it didn't quantify many if not most of the costs for mercury because mercury causes developmental delays, attention deficit disorder, all these other things that are not quantified in the rule. It causes cardiovascular problems. A lot of things mercury does. It's an extremely poisonous neurotoxin. The other thing about the 30 billion to 90 billion, that is particulate reduction. And some of those things, the acid gases, which turn into particles because they become aerosolized, they go into your lungs as tiny droplets. All of those are in fact being taken care of in the controls of particulate. It is true that in controlling those hacks, you use the same technology and you end up controlling a lot of other kinds of particulate. Primarily sulfur dioxide, which causes premature deaths. When they did the calculation, they said we put these controls in to control hacks. It also saves a lot of lives because of sulfur dioxide that is not otherwise being controlled. Is this the basis for the EPA's decision? I thought the EPA's position was it doesn't matter how much the costs exceed the benefits, we just will not take costs into account at the listing stage. That is correct. They did not consider the cost benefit analysis at the listing stage. That is based on their reasonable interpretation of the cost benefit analysis. The EPA was not regulating effectively. Congress said we will force regulation of these chemicals. It gave one benefit to the power plant industry. You have three years and you can prove there are not serious health effects. It did not give them the benefit of having a cost benefit analysis done up front or create all the discretion in the world on the part of EPA to say we don't think we should regulate this category. Mr. Lindstrom, you have four minutes remaining. Any sub categorization that was going to happen has already occurred because we're talking about the rule that's been promulgated and despite any sub categorization that happened there's still $9.6 billion in costs being imposed on an annual basis. It wasn't the question presented. Is the question presented that not that you have to take that into account at listing but that somehow that ratio makes any emission standards wrong? Even if for some people it's really not backbreaking to do it? The question is whether costs have to be considered under N1 when you're regulating electric utilities. They do say it does when you're regulating at the emission standard. If you think about this, what happens under 7412C is you have a listing decision, a floor standard, and above the floor standard. You're taking out the categorization. They don't establish the floor until they categorize. You're making the necessary and appropriate determination when you're looking not at the list but whether such regulation is appropriate and necessary. That's the language in the statute. It says is such regulation appropriate and necessary? It's not just whether it's going to be listed. It's looking at what's going to happen. That's why they did both at the same time. They published the emission standards. They were looking at the costs ahead because they didn't know what the costs were going to be. You're saying they purported to make the categorization decision without taking into account costs? Yes. Any categorization they've done and they didn't consider costs? That's correct. Again, I'd like to return to one of the big picture principles. This turns or falls on what the agency did below. They've said costs are not relevant. They've ignored an important part of the regulatory problem. I'm sorry. They proposed those categories and everybody had the opportunity to say it's the wrong category, correct? And argue why it's the wrong category. Some people submitted complaints about costs relative to their technology and their kinds of plans, correct? Yes, Your Honor. It's not true that they proposed them, but everybody gets a chance to say this technology is different from the others or this kind of plant is different from the others and imposes a cost much greater than you're anticipating. I'm saying we're already past that phase. Any categorization we're going to do has already been done. You're saying it's already passed. It's passed because the final rule has been issued. Correct. I'm talking during the rule-making process. The rule-making process does permit the agency to consider the cost of technology in setting up categorizations. They've adopted the exact opposite position, which the costs do not matter. I thought we just heard the Lignite people, I can't remember that's the right name, they said look, we have special ways of producing our stuff and they're much more expensive so please don't put us in the same category as you put the other people in. And the agency said right, separate. Did that happen? And how would you do that without considering costs? Because the Lignite people were saying our costs are more expensive. I don't know how they did it, but they've said throughout that we're not considering costs. Thank you, Your Honor. Thank you, counsel. The case is submitted.