# Oral argument — State of Arkansas v. EPA > Briefs, arguments, decisions, and more. URL: https://www.frixlaw.com/law-library/documents/argument%3A94684 ## Record - **Collection:** Oral argument - **Document type:** Oral argument - **Published:** October 22, 2024 ## Text Chief Judge Colleton, and may it please the Court. This is a case about whether EPA may tell a state to meet one benchmark and then hold it to another one. Here, EPA told Arkansas to use one set of modeling and a screening metric to draft its SIP. Arkansas did that. Then, years later, EPA disapproved that SIP because it didn't use different modeling and a different, more demanding screening metric. Well, it didn't quite say to use the modeling, did it? I mean, the language in those, I think there's three separate communications from EPA that you're talking about. It never said use the monitoring, right? I understand it said this may be of assistance or something along those lines, but I don't think it was anywhere close to a directive, was it? I mean, can you be a little more specific about what you think the meaning of, and I think we're talking about two memos and an updated model. Is that correct? So, we're talking about there's a modeling memo and there's a modeling issue, and then separate and apart from that, there's a threshold issue, which is a different model. So, I can start with the modeling issue. The modeling guidance that they issued in, I believe it's March of 2018, told states that they could use this modeling to assist them in drafting their SIP. If it was justified, right, on a state-by-state basis or by a technical analysis submitted by the states? That's the threshold issue, Your Honor, if that's their argument for the threshold. So, to deal very quickly with the modeling point, there are multiple memos in this case, so totally understandable. So, with respect to the modeling, what they said is here's the guidance that you can use to draft your SIP. That modeling answers critical thresholds about what is the state's linkage. States don't generally do their own modeling in this context. Instead, they rely on EPA. EPA tells them if you want to deviate from EPA's modeling, you have to justify doing so. Well, you say states don't generally do their modeling. I mean, some of the states here did, and I understand you're representing one who did not, right? But states are still responsible for submitting in their SIP adequate modeling, and you could use your own, right? That's correct, Your Honor. And, in fact, if we'd have been aware of what ultimately happened here, which is that they told us to use one set of modeling, that we could use it to draft our SIP, and then we relied on that to target the one thing that that modeling and their screening threshold told us to target, we would have been, actually, frankly, in a better situation. Because in that context, we would have been aware we had to draft our own modeling. We could have relied on that modeling. We could have worked with other states, perhaps, to develop that modeling. And then, if they wanted to deviate from the modeling that we prepared, they would have had to justify it, which is more than they did with respect to the modeling switcheroo that they pulled here. And that's because they basically didn't offer any explanation for why they were changing the modeling, other than to say, we think it's better to rely on more recent data, which is not consistent with the EPA's own past practice in this area of not pulling an agency switcheroo long after states have submitted their SIP. I mean, one thing worth highlighting here with respect to the modeling is we're not talking about a short delay after we've submitted our SIP. In fact, we submit our SIP, they wait well past the statutory deadline, which is a year, well past that, while they develop new modeling. And then they decide, after waiting years to develop it, to drop that. So you say that's not a usual practice, and there's a lot of cases here. And if I'm citing one or citing one that was raised by another party, just let me know. But wasn't it the Wisconsin v. EPA case that involved at least a somewhat similar, I think, different rule making, perhaps? But in that case, as I remember, Delaware took issue with the fact that EPA switched after the initial submission there, switched to a later model. And I think that was a D.C. Circuit case. And the D.C. Circuit basically said the SIP deadlines are procedural only the implication in the follow-up there was being the EPA was okay in relying on modeling that was developed post-submission. Is that an accurate understanding? And if so, how was your case distinguishable from that? So the difference is a reliance interest, Your Honor. So what Delaware was arguing was that post-SIP development, post-submission, that even though later modeling showed that Delaware did not have a nonattainment problem, Delaware was being deemed for nonattainment, which meant that they weren't meeting the NAAQS standard. And what they were saying when the later modeling showed that Delaware was, in fact, in attainment, the EPA should still hold upwind states to the previous nonattainment determination. And what — essentially what the EPA argued and then what the D.C. Circuit blessed was, well, that's not really consistent with the statute to say — when the statute says that you have to control for admissions, address admissions that go to neighboring states which are not in attainment or which will not be in attainment. So is your — is the reliance interest an essential element of your argument here? Are you suggesting that the EPA can never use post-submission data or modeling on a NAAQS, ozone NAAQS? Or is it — is the reliance an additional aspect of the argument or is a necessary argument — element of the argument here? I think the reliance is central to the argument in that they told us we could use a particular set of data. We should use a particular set of data. And again, they told us we could use it. We did that. We relied on it. We drafted our entire SIP based upon that modeling and the threshold that they gave us to target the one contribution, Allegan County, Michigan, that they told us to target. Now I would distinguish that from a situation where reliance doesn't factor in, where that later data shows that, in fact, there's not a nonattainment problem. We didn't contribute. So there's not a reliance problem, in other words, where that subsequent data confirms what the State argued in its SIP. In that situation, there is no reliance problem. So the fact that the data later showed, as we argued in our own SIP, that Allegan County — we didn't contribute to nonattainment in Allegan County. You can use it in that situation because there's not a reliance problem. We didn't rely to our detriment. The problem here, when they told us to use their data and use their threshold to target what it told us to target, is we drafted a SIP entirely around what they told us, and then they sit on it well past the statutory deadline to develop new data, and they spring it on us just 13 days before proposing to disapprove the SIP. I mean, the timeline here really matters. They sat on it for years past the statutory deadline, and then just 13 days before proposing to disapprove, they drop that new data on us that says, here are the new sites, and then they disapprove our SIP because we didn't address those new sites identified in that hot-off-the-press modeling. We didn't know about them, so of course we didn't address them. We did what EPA told us, and it's a basic principle that agencies can't use the rulemaking process to pull a surprise switcheroo. That is, they can't penalize a party for relying on what EPA told it to do. So that's the modeling. If I can briefly address, because Your Honor was asking about the threshold issue. In, I believe it was August of 2018, they issued a memorandum that said states could use a one-part-per-billion screening metric to determine significant contribution. Arkansas relied on that representation, using the modeling and that threshold to, again, identify the site that it needed to address in its SIP. Other states also relied on those representations as well. EPA then, in the final rule, the final disapproval, said, we're sorry, even though we told you the one-part-per-billion was the appropriate screening metric, it's actually .7, years after the SIP had been submitted. It's essentially, they said the standard was here, and we targeted that standard to meet it, and then all of a sudden, in the final disapproval, they told us, no, in fact, that wasn't the standard. You met the wrong standard. You should have met the standard up here. If you'll refresh my recollection, doesn't EPA also argue that your SIP was inadequate, even under the one-part-per-billion standard, under their modeling? Is that correct? And how do you respond to that? So one of, they make, I think, two versions of that argument, one of which is a harmless error analysis. But I think on one version of it, they say, if you, if they're allowed to use the new modeling, that in the final analysis, we still had one receptor over one part-per-billion. And they say that as a result of that, if they're allowed to use the new modeling, they would have inevitably disapproved the SIP. The problem is, that's not the analysis that the final rule actually does. The final rule does an analysis that says you're disapproved based on the six receptors that that final modeling, above .7, pinged you to. And you didn't adequately determine whether or not to impose controls to address those six receptors, not the one receptor. And that one receptor is only at 1.21, so it's just a fraction of a fraction above the one part-per-billion screening metric. If we did, if that is the standard, if they wanted to do that analysis and said, you know, that one receptor is enough, they could have done that analysis. They didn't do that analysis. Anything that the lawyer is telling you today is just a post hoc representation about how they would have approached that. And it's not automatically true that just because you are linked to one receptor, you're required to impose controls, and they would have disapproved our SIP. Instead, under EPA's own four-step framework, if you meet the screening threshold, they then proceed to the next step, step three of their four-part framework, where they look at a cost-efficiency metric. What does it cost to address that receptor? Here, that one receptor was 1.21, so just a fraction above the one part-per-billion. And it may very well have been that they would have agreed with us that controls were not required in order to address that fraction of a hair, given the cost of what it would cost in order to address it. They simply haven't done that analysis, and they can't represent today that that's the tact that they would have taken. The other alternative that they give is they suggest that they would have disapproved our SIP even using the old modeling and the old screening metric. That, again, is just post hoc representation. In fact, they say the exact opposite in the final rule. And they're not allowed to pull in yet another switcheroo now and change what the final rule says. The We're not relying on Allegan County. We're not disapproving you because of your linkages to Allegan County. I'll grant you they do say very disparaging things about the State's analysis of its linkages to Allegan County. But ultimately, what EPA says is we're not relying on those linkages. We're not doing a thorough analysis of Arkansas's analysis. And it disclaims any reliance on it. So they're not allowed to pull yet another switcheroo now and say, even though the final rule says the exact opposite, we're now going to say we would have disapproved you anyway. Frankly, we don't know that. And given what we do know, which is that their own updated modeling shows there is no nonattainment problem in Allegan County, they couldn't have disapproved us on that basis. They would have had to approve us. The guidance on the threshold, I thought it wasn't quite as clear as you made it sound. My recollection is that it was phrased more in terms of with justification or under certain circumstances you can use the 1.0. That's their characterization of it, Your Honor. I think what the guidance actually says is that EPA has done a thorough analysis of the one part per billion in the .7 metric. And that EPA, based upon its thorough analysis, has determined that the two metrics capture a similar magnitude of contribution. That's their language. A similar magnitude of contribution at most sites. Again, their language. And what that plainly communicated to the States is that it captured a similar number and that because on average it would capture a similar number, States could use that metric as a screening metric. And then it does include language that says this may not be applicable in every single circumstance, every single SIP. What that communicates is that because on average it's going to measure the same thing, you can use that one part per billion, but there may be reasons why you would depart from it. So the exact opposite of what they're claiming the memo says. And I think the plain text reading of that memo is you can use it unless there's a State-specific reason to depart. And frankly, underscoring that's what it meant is the fact that virtually every State read it that way. I mean, their position is every single State just misread it. It's also the way that EPA, frankly, read that memo back in 2020. As they argue in their briefing, they proposed to approve Iowa SIP in 2020, which did not offer State-specific justifications, instead relied on EPA's analysis in order to use the one part per billion. And they proposed to approve it using that metric. So what that shows is that at least in 2020, EPA didn't even read its own memo the way it's representing it reads it now. And that really, the fact that they don't even acknowledge that they've made a change here, that they insist that they're just reading the memo correct, is a separate administrative law problem, because it means they didn't acknowledge the change, which certainly means that they didn't give a logical explanation for making that change. In fact, it's doubtful that they actually could have given a logical explanation for the change, given that their own data confirmed that the difference in what was captured between one part per billion and .7 parts per billion was even smaller than they originally suggested back in August of 2018. Back then, they suggested that the difference captured between the two metrics was about 7 percent, so it's already a minuscule difference between the two. Their final rule actually says the difference between those two is just 5 percent. So if they'd tried to explain the change or justify the change, they would have had a real difficult time doing that. Why now all of a sudden they needed to use an even more demanding metric? So, again, I think that that really illustrates a critical problem with their analysis, even if they had tried to explain that away. What's the significance of the statutory timeline? I think what the statutory timeline reflects, Your Honor, is the idea that new information is always available. We're talking about forecasts for future emissions, that there's always new data, always new technology that's around the corner, and it suggests that Congress understood and told the agency, given that, you need to act within a confined period of What do the cases mean when they call it procedural as opposed to substantive? And is that correct? And what's the significance of the distinction? I think the reason why they call it procedural is because at least other courts have said that it's not a hard, enforceable deadline, that there is a . . . If they could, say, a week after the deadline, act on a SIP disapproval if there's no prejudice or reliance? I think, yes, that would be a very different case. We would not . . . You don't think it's a hard and fast rule? I don't think it's necessarily a hard and fast rule, but I . . . What's the significance of it? That it underscores, Congress understood and told the agency that this data is continuously changing and gave it sort of a confined, the amount of time it had to make these decisions. It suggests that Congress understood, as EPA understood, that this data is continuously changing. And if you take their approach where they're just allowed to sit on a SIP submission years afterwards and change the standard . . . All right. I'm just going to drill down, though. If it's not a hard and fast rule, what's the legal significance of the deadline? Again, I think it indicates that Congress was telling the agency that it needed to act within a confined period of time, given that data is continuously changing. Act within an amount of time or limit its consideration of data to that amount of time? Well, I would actually say that even if the new data in the EPA's previous position was even if the new data comes out on the eve of a SIP submission, they're not going to hold states to that new data. And that's because SIPs are not something that's prepared overnight. It's part of a long process. So when they originally gave us the data here, it was about 18 months between when they gave us the original modeling and when we finally submitted our SIPs. That's how intensive and technical the process is. So when they proposed to disapprove us, they released the new modeling, again, 13 days before proposing to disapprove us. They only gave us 60 days to respond and essentially drafted an entirely new SIP targeting this new stuff that they say you should have targeted in your SIP. You shouldn't have paid attention to what EPA said. One other brief point I did want to address is the venue issue. Yesterday, the Supreme Court petitions in Oklahoma and Utah's cases, the Tenth Circuit had transferred those cases to the D.C. Circuit. And that will obviously control the venue. The Court's decision in that case will obviously control the venue decision in this case. Now, we think it's very obvious that the EPA is wrong about venue here, given the plain text of the venue statute that assigns venue to the State if the Court were to wait to see how that comes out. No, Your Honor, as long as the stay remains in place. Have any of the other circuits responded this morning yet? I know you just got here, I just got here, and I'm not monitoring Westlaw. So I'm only aware that the Fourth Circuit canceled an argument that was scheduled for next week. I suppose it's too late for us to do that. There may have been discussion in the audience about whether that was going to happen this You're allowed and you're not required to use all of your time. Unless the Court has any questions. Thank you for your argument. Mr. Wilson, we'll hear from you next. Good day, Your Honor. May it please the Court? I'm Josh Wilson. I represent industry. We support the State of Arkansas and all its arguments, but I also wanted to raise a couple of technical issues with EPA's disapproval. You may, and you may also raise that by pushing the button to your right so that the microphone is closer to your voice. That'll help us hear better. Is that better, Your Honor? Thank you. So I'll start with two technical facts, and then I'd like to explain why they actually matter for the case. Fact one, EPA's CAM-X modeling that it used. EPA's, they call it their CAM-X or C-A-M-X, their photochemical model that they used. EPA did an analysis on how precise or accurate it was for these particular receptors and found out that it was approximately seven parts per billion standard error, which is, so if you think about it, plus minus seven parts per billion is the accuracy we're talking about on the model. The second technical fact is EPA separately, in another proceeding called their ozone sill or significant impact level, did some statistical analysis on actual observations of measured data at different sites and determined that anything below about one part per billion is just statistically indistinguishable from background inherent variation at a site. So why do those matter? I think it matters both to the substantive arbitrariness of EPA's submission, as well as to the harmless error point that they've raised in litigation. First on the arbitrariness front, because the states are given the primary role here under the statute, in order to disapprove a state SIP, EPA needed to come up with some basis for saying that Arkansas' SIP submission was arbitrary, and they did not do so. The only rationales that they gave in the record were, well, we have some technical nitpicks. But one of those was, I said, Arkansas' entire approach to looking at whether they were significantly contributing or not was Arkansas looked at, well, do we have a consistent and persistent impact on that downwind receptor? And EPA said, we don't even have to engage with that, because to be clear, our CAMEX model already determined you have a consistent, persistent contribution. But that's where that first technical fact comes in. In fact, their model did not, because the magnitude of error on that model was far greater than the tiny, minuscule impact that they were predicting from Arkansas. And this might be a different case if we're talking about a SIP, where EPA gets the primary authority there to set their SIP, and we can, maybe they get some sort of technical deference on those grounds. But here, they were supposed to be, they needed some actual reason to say that Arkansas' approach was arbitrary. And the reason they gave was, well, our model already says it, but that just simply wasn't the case. Second, it matters to harmless error, because EPA wants to say, well, none of, our whole modeling switcheroo was harmless. But that's simply not the case, because if Your Honors agreed with Arkansas' argument that it was actually a substantive administrative law problem to arbitrarily switch the modeling and not give the state an ability to address it, then we can't just assume that whatever Arkansas did in response is going to be deficient such that EPA will still be able to disapprove it. First, we can't even assume that Arkansas would fail to take any steps to reduce its contribution. The only reason they didn't before was that that receptor that EPA used to identify turned out to not actually be linked in any consistent, persistent way, and even EPA agrees now that it's not even a non-attainment selector. But secondly, they wouldn't be able to just give the same rationale and say, well, you're not even allowed to look at any other technical grounds for saying you're not consistently, persistently contributing, because they can't use the exact same rationale. That rationale is arbitrary, at least with all of the modeling they've submitted at this point. With that, I see I'm down to about 20 seconds, so I'll cede the rest of my time to Arkansas unless you have any further questions. Very well. Thank you for your argument. Thank you, Your Honors. Ms. Lineberry, we'll hear from you. Can I grab my water real quick? Yes, we want you to be well hydrated for the argument. Thank you, Your Honor. You may proceed. I'm not as tall as my co-counsel. Right. You can adjust them. Good morning, Your Honors, and may it please the Court. I'm Anna Lineberry on behalf of the United States. I'm joined at Council's table by my DOJ co-counsel, Ms. Miranda Jensen and Ms. Chin Lee, as well as EPA's Mosemary Caban of their Office of General Counsel. I will respond to the Arkansas petitioners, Ms. Jensen will respond to the Missouri petitioners, and Ms. Lee will respond to the Minnesota petitioners. I'd also like to address some overarching points on venue and cooperative federalism. Before turning briefly to venue, I'd like to step back for a moment and highlight two points that are critical to this case. The first is that these cases are about EPA's actions on state implementation plans that were submitted to the EPA. That means that under that provision, Section 7410A2D, states shall prohibit amounts of pollution that will contribute significantly to downwind pollution problems. Unlike other SIP requirements, which concern in-state obligations, the Good Neighbor provision requires states to look at how their pollution impacts other states' ability to maintain healthy air quality in compliance with federal law. The second point I'd like to briefly cover is that the pollutant at issue here is ozone. At ground level, ozone can cause serious health problems like asthma and heart attacks. Ozone pollution presents a collective contribution problem, which you'll hear us talk about a lot today. Unlike some localized pollutants, ozone can travel hundreds or thousands of miles, and small amounts from many upwind states can mix and combine to prevent a downwind state from achieving healthy air quality, even if that downwind state is doing everything that it can to keep its local emissions in check. So in the Good Neighbor provision, Congress created a mechanism to ensure that downwind states are not unfairly burdened by pollution from upwind states by requiring those states to prohibit those harmful interstate emissions. So with that background in mind, I'd like to address the venue briefly. So as my friends on the other side explained, the Supreme Court granted cert in Oklahoma versus EPA yesterday. So in Oklahoma, the 10th Circuit Merits Panel agreed with EPA that the D.C. Circuit is the proper venue to hear Utah and Oklahoma's challenges to the disapproval. The Supreme Court also granted cert on the United States petition for cert in another case involving the Clean Air Act's venue provision. This is obviously a new development. The 4th Circuit canceled its oral argument in the West Virginia case challenging the disapproval for next week. We're confident in our positions on venue and the merits, and we're of course prepared to present those arguments today. The Court has invested a lot of time and resources in preparing for this argument and gathering all this counsel here today, but we... We don't take a position affirmatively on whether this Court should or should not stay this case at this time, but we appreciate that the Court may want to see how the Supreme Court decides the threshold issue. So... Does that... Correct, Your Honor. This Court's one or two sentence preliminary orders are not the law of this case. We have these cases that seem to say the opposite, don't they? I think the Court's decisions in the... I think I'm pronouncing... It's Neffler case and in Iowa League of Cities and in Stanko all reconsidered administrative orders when finding that those orders were not... Did not address the issue with sufficient clarity and directness such that it had... Such that it created settled expectations. I thought there was like... I'm sorry, Chief. I was just going to say in this... I mean, they didn't write an opinion, obviously, but the issue was squarely presented and the motion was denied. It wasn't referred to the hearing panel, for example. Wouldn't that be clear and direct and... That's true, Your Honor. I think that as I understand this Court's cases, the sufficient directness and clarity would involve an explanation of exactly what is the basis for the Court's ruling. And I think that the cases that those decisions rely on include the... Its first Union National Bank case, which I think explained that in a little bit more detail. And here, I think it's important that we also do have the intervening case of Oklahoma versus EPA that found specifically that the action at issue here is appropriately considered in the D.C. Circuit. We still have a circuit split, though, right? I mean, the Fifth Circuit came out the other way and I don't believe it's changed, right? I mean, and the Sixth Circuit is with the Fifth. Is that correct? So there are three decisions. There's the Fourth, Fifth, and Sixth Circuit. The Fifth and Sixth Circuit cases were also preliminary motions panel orders that were re-argued at the merit stage. But those came with written opinions, both of those. I'm trying to remember. There's a lot of moving parts here, but I think those were opinion, per curiam opinions with published dissents in both cases, weren't they? The Fifth and Sixth Circuit decisions were not published. The Fourth Circuit's decision was, and EPA did file a petition to re-hear that decision on Bonk, which was denied as premature at that time. And the Court has obviously stayed the case, given the cert petition, before hearing argument on the merits. But the Fifth and Sixth Circuit orders are not published. And I think all three of those decisions were followed by persuasive dissents from other members of the Court. And so if this Court is interested in hearing argument on venue, I do think it should reconsider. We do think that it should reconsider the venue question. It should transfer the petitions to the D.C. Circuit because the disapproval is either nationally applicable or it's based on determinations of nationwide scope or effect. So the Clean Air Act's venue provision centralizes review of national issues. It directs judicial review to one court, the D.C. Circuit, if an action is either nationally applicable or based on those determinations of nationwide scope or effect, such that centralized review helps ensure consistent answers to nationally important questions under the Act. So on its face, the SIPP disapproval is nationally applicable because it applies to 21 states across multiple circuits and it applied a nationally consistent or a common analytical framework. So as the Tenth Circuit held in ATK and as the Seventh Circuit held in Southern Illinois Power Cooperative, venue is a facial test. And applying ATK, the Tenth Circuit merits panel in Oklahoma v. EPA found that the disapproval is nationally applicable and same. The disapproval is also based on those multiple determinations of nationwide scope or effect. So those determinations are technical or legal judgments that underpin the action. So for example, one such determination that's at issue here is that states are not excused from eliminating their significant contributions downwind because of either international emissions or the emissions of other states. So Arkansas and Missouri petitioners challenge that determination here in their merits briefing. West Virginia challenges it in the Fourth Circuit, and Utah and Oklahoma challenged that in the Tenth Circuit before it was transferred to the D.C. Circuit. So inconsistent rulings in these circuits on the validity of EPA's determination could lead to inconsistent treatment of similarly situated states, which is a result that the venue provision sought to avoid. So EPA made many other determinations of nationwide scope or effect. Those are detailed in the record on pages 9,380 to 81 of the disapproval, and they're further explained on page 392 of the response to comments document. Consistent with Congress's intent to centralized review of those nationally important issues, this Court should transfer the merits. So on the merits, the issue before this Court is whether EPA reasonably disapproved Arkansas's and the other states' good neighbor SIPs. So under the Clean Air Act, EPA is responsible for determining whether the SIP meets Congress's requirements in the Clean Air Act. The question is whether EPA provided a rational and lawful explanation for its determination that they did not. So under 7410K3 of the Act, EPA can only approve a SIP if it meets all applicable requirements. For good neighbor SIPs, the assessment requires EPA to make a bunch of smaller determinations about the SIP's compliance with the good neighbor provision, which is 7410A2D. Those requirements can be distilled basically into three parts. So the first is, will states contribute to nonattainment or decline in a downwind state? The second is, if so, what amounts of those emissions are significant? And third, are there adequate provisions in the state's plan to eliminate those significant contributions? EPA is not required to afford the states any deference in answering those three questions. So this Court has regularly deferred to EPA's technical judgments in SIP review cases like North Dakota and Nebraska. In those cases, this Court flatly rejected petitioner's characterization that they make here today that EPA's statutory role is ministerial and deferential to the states. In those decisions, this Court rejected those arguments. In the Nebraska case, for example, EPA disapproved Nebraska's BART determination, or its determination of the best available retrofit technology under the Regional Haze Provision of the Clean Air Act, because, quote, it was based on a flawed analysis and an unreasonable conclusion. So that's on page 668 of that decision. Nebraska argued that EPA substituted its analysis for the states, which is the same argument that petitioners make here today. What about the argument that the EPA changed the modeling? So EPA didn't change the modeling. Well, I misspoke, but you know what I mean. What about the argument that there was a change in the modeling that was considered? Of course, Chief Judge Colleton. So Arkansas tries to say that EPA departed from its own modeling memo, the 2011-based modeling, and the threshold memo. But the fact is EPA disapproved Arkansas SIP on its own merits. So EPA did consider the 2016 V3 modeling, but it was just confirmatory. Is that in the final rule, I suppose? I mean, I believe co-counsel for the state suggested that those arguments would be made by counsel, but were not made by the agency in the rulemaking process. That's not consistent with the record, Your Honor. The disapproval says on page 9354 that the full basis of its disapproval includes the proposals and includes the data that was available to the states at the time that they submitted their SIP. But how do we know they would have come to the same conclusion if they had not considered the 2016 modeling? So for Arkansas, Arkansas's significant contributions downwind were consistent across all versions of the modeling. In the 2011-based modeling, it was contributing to multiple receptors above EPA's historic 1 percent threshold and above Arkansas's preferred 1 ppb threshold. In the 2016-based modeling, either version 2 or version 3, the story is the same. Arkansas is causing significant problems downwind. So how did Arkansas avoid what you're suggesting the conclusion should have been? Was that through application of the maintenance memo, or how did Arkansas run afoul of what you're suggesting the models showed? The maintenance memo is at issue in the Missouri portion of this case. It's not something that Arkansas relied on for its arguments. So Arkansas's submission, even under the 2011-based modeling, was technically flawed for multiple reasons. This is not the type of case where there are two valid technical conclusions and this Court has to choose between them. And even in that case, this Court would defer to EPA's technical judgments. So is that undisputed, that we would defer? Judicial deference to agency is a bit in flux. Is that deference to technical determinations, is that still good law in EPA's view? I assume the answer is yes. Yes, petitioners don't contest that. I think their reply briefs even concede that EPA gets technical deference and that it's consistent with Baltimore Electric and Gas Company and this Court's precedents in North Dakota and Nebraska. I'd like to hide three technical errors with Arkansas's submission here today. So the first is Arkansas's reliance on back trajectories did not undermine its modeling. Arkansas filtered out those trajectories based on their starting height or their center line, and it gave those trajectories too much weight, given their simplicity compared to EPA's sophisticated modeling, which also considered meteorological data that would be captured by those back trajectories. The second is that Arkansas looked at emissions from other states contributing to the receptor that it focused its analysis on, Allegan County, Michigan. Using emissions data, it argued that downwind contributions to that receptor were insignificant compared to other states. But that ignores the point that I made earlier, that ozone is a collective contribution problem. Even Arkansas recognized that its contributions to that receptor were not de minimis because they were above its own chosen de minimis threshold, 1 ppb. So the good neighbor provision thus required Arkansas, just like every other state contributing to that receptor, to consider options for eliminating those significant emissions. And the third technical flaw that I'd like to highlight here today is that Arkansas only looked at emission control options for power plants. So power plants only represent 56% of Arkansas's nitrogen oxide emissions. Nitrogen oxide is a key ozone precursor emission. In assessing power plant controls, Arkansas applied a cost effectiveness of threshold from EPA's last good neighbor action, but EPA reasonably concluded, and this is explained thoroughly in the record, that it was inappropriate for Arkansas to rely on that older threshold to address the more stringent 2015 ozone max. So Arkansas is very focused on the decisions that EPA made about the appropriateness of a 1 ppb threshold and about its consideration of the 2016-based modeling. And Arkansas said basically that it was impossible for it to submit an approvable SIP because of those two things. But Arkansas is wrong for three reasons. First, EPA would have reasonably disapproved Arkansas's SIP on its own merits, regardless of where it was causing downwind ozone problems. Second, the threshold memo made clear, as Judge Grunder was pointing out earlier, that EPA may not approve a state's use of a 1 ppb threshold. In the alternative, Arkansas easily could have considered its many contributions above 0.7 ppb to Texas receptors, and those contributions were consistent across all versions of the modeling. Regardless of which modeling you consider, Arkansas is contributing to ozone problems downwind in Texas. In which you say they're consistent across the model, what's the first model where that's accurate, where your statement is correct? Is that the 2011 model? Yes, the 2011-based modeling. What was the place in the record you thought was best to show that the agency supposedly did rely on the 2011? So there are a few parts. The first is page 9354, Chief Judge Colleton. So it explains on that page that the full basis of EPA's disapproval includes the proposals, and includes the data that was available to the states at the time. And I think also on page 9355, where EPA is considering Arkansas's submission specifically, that Arkansas discussion, it includes all of the reasons that I just explained for why Arkansas's disapproval was appropriate, for reasons that are independent of the modeling that's considered, given the consistent linkages. And it incorporates the summary from the proposal. And the proposal at page 9810 critiques the SIPP under the 2011-based modeling, so under its own merits. So there are a few points in the record where that's clear. So it's important to recognize... Are you sort of abandoning reliance on the 2016 modeling? You seem to be saying that the way the Court should go is to say you don't have to consider the 2016. I don't think that this Court has... Sorry, did I mean to cut you off, Chief Judge? You seem to be saying we don't have to consider this controversy about whether EPA impermissibly relied on the 2016 modeling because you think the result would be the same under 2011, and the agency supposedly so stated. Is that what you're proposing the Court should do? Correct. For the disapprovals of Arkansas and Missouri. I don't think EPA should still disapprove Minnesota's submission based on the 2016 V3 modeling. But for Arkansas and Missouri, EPA's consideration of the 2016-based modeling is a red herring because its linkages are consistent across all versions of the modeling, and EPA had multiple valid bases for its disapproval of those submissions. So in all three versions of the modeling, Arkansas is contributing above its preferred threshold. Including its preferred 2011-based modeling. In all three versions of the modeling, Arkansas is contributing above EPA's threshold and even above 0.9 ppb to receptors in Texas. And those downwind citizens face real problems attaining and maintaining federal air quality standards, and they face real health consequences because of this pollution. The reality shown in the record is that emissions from Arkansas will contribute to downwind ozone problems. And Arkansas's conclusion that it need not prohibit any of those emissions was unreasonable, and EPA lawfully disapproved it as part of a consolidated action involving nationally important questions, and those petitions challenging that action should be transferred to the D.C. Circuit, or if the Court rules on the merits, it should deny the petitions. And I hope that EPA made it crystal clear that it was denying based solely on the 2011 data. No, I don't think the disapproval says it's denying based solely on the 2011-based data. I think it says it's denying based on all of the information in the record, including the 2011-based data that Arkansas wants this Court to only consider. But EPA also reasonably considered the 2016-based modeling. It was the best available data to EPA at the time, and so that data was confirmatory of the downwind ozone problems that Arkansas was causing. And so there isn't a spot in the record that says we're only disapproving Arkansas or Missouri based on the 2011-based modeling. Don't you run into a Chenery problem, then, with this harmless error argument? In other words, if it's not totally clear that the EPA was denying based on just the 2011, when you're now making this sort of harmless we would have done that, can we, under Chenery, consider that? Yes, Your Honor. I think the record makes clear that the 2011-based modeling is a valid basis for EPA's disapproval if the Court only wishes to consider that modeling. You keep phrasing it that way, it's a valid basis for it, but the question is, was it the basis for it? Not whether it would be a valid basis, but whether it is the basis that was cited by the agency. It's a basis. It's not the only basis. I believe so, Your Honor. I don't think that petitioners have argued that it's not. Well, I guess I should say, I think that this Court, under the standard of review, explained in Nebraska and in Alaska Department of Environmental Conservation, even if EPA explained that with less than ideal clarity, this Court should still uphold its action, which is well supported by the record. Is that same clarity there for the threshold question as well? Yes, Your Honor. So the threshold memo, I think, says very clearly that it, petitioners try to frame the threshold. I'm actually thinking about the harmless error argument. I mean, is it as clear that this would have resulted in the same thing, same denial, if they used the higher threshold? So the disapproval explains that while it wasn't, so while EPA thought it was inappropriate to rely on a 1 PBB threshold, it didn't disapprove any state SIP simply because it did that. It considered every state's justification for why it thought that reliance on the 1 PBB threshold would be appropriate for that particular state. Arkansas didn't provide state-specific reasons for why it could rely on that threshold, and EPA explained in the disapproval that the memo did not presumptively, did not say that the 1 PBB threshold would be presumptively reasonable for any state to use. The text of the memo itself is very clear about that. It says that states should consider whether the recommendations in the guidance are appropriate for any particular situation, and the guidance may not apply to all facts and circumstances under the SIP. So there's no change in EPA's position about the threshold memo, and it's appropriate for any particular state. So that's well supported by the text of the disapproval as well. EPA's disapproval of Arkansas SIP is well supported by the record. This Court should deny the petitioner's petitions. The Court has no further questions. Very well. Thank you for your argument. Thank you. We'll hear rebuttal if you could please stick to your time. Thank you, Your Honor. Just a couple of quick points. On the point about whether or not they gave this alternative analysis. Could we please turn on the timer? Oh, gosh. Sorry. Sorry. You may proceed. On the point about whether they conducted this alternative analysis, what I heard in response to Judge Render's question was that there's some generic statement in the disapproval about we considered everything that the agent or that the State submitted, that that's simply not enough. And I would also point the Court to they specifically disclaimed any reliance on Allegan County. This is on the at the appendix at page 2239, where they said that the receptor in Allegan County was not identified as being a nonattainment. That's footnote 52 there. They also say on 2239, because Allegan is attaining EPA, was, quote, not relying, end quote, on Arkansas's analysis of whether it contributed to nonattainment or maintenance problem in Allegan. They also say at 2240, they're not providing a detailed evaluation of Arkansas's analysis in this rulemaking, because Allegan County wasn't a violating receptor. And then at appendix 820 and 2235, they conclude that, in fact, Arkansas's contribution was half of what they originally estimated. So they've specifically disclaimed any reliance on Allegan County. Now they want to pull yet another switcheroo and say, well, even though we didn't say it in our disapproval, even though we explicitly disclaimed it, we'd like to rely on it now. And they're simply not allowed to do that. And it's also doubtful that they could do that, given that they now know there's no nonattainment problem in Allegan County. That simply wouldn't be consistent with a statute that says control admissions, which will contribute to nonattainment, when they know there's no nonattainment problem. And then third and finally, I heard repeatedly represented that the monitors or the receptors were consistent across all the modeling. That is not accurate. In fact, the only, the main receptor we were told to address in Allegan County drops off in all the later versions of the modeling. There's only one receptor that's identified in the various iterations of these modeling. That's one receptor in Missouri County that in the original modeling was identified as being below the threshold that we had to look at, the one part per billion. And in the final modeling, it's just a fraction of a fraction above that one part per billion metric. And they haven't done an analysis that says we would be required to control to address that one fraction, minuscule amount above a one part per billion. They simply haven't done that cost efficiency ratio. They can't represent it here today. They're simply not entitled to make that And then lastly, I heard multiple times that our analysis was technically flawed. Well, apparently it was so technically flawed that it was right. We argued that we didn't contribute to nonattainment in Allegan County. Their own data says that, unless the Court has any questions. Very well. Thank you for your argument. --- Source: Frix Law Library, https://www.frixlaw.com/law-library/documents/argument%3A94684. Public record. Not legal advice.