Oral argument — NRDC v. Michael Regan

Case number 20-1335, Natural Resources Defense Council Petitioner v. Michael S. Regan, Administrator, U.S. Environmental Protection Agency and Environmental Protection Agency. Ms. Forge for the petitioner, Ms. Buckley for the respondent, Ms. Snow for the interviewer. Thank you. We're ready to hear our first case this morning. Ms. Forge. Good morning, Your Honors. May it please the Court, Sarah Forge presenting on behalf of Petitioner and RDC. I'd like to reserve three minutes for rebuttal. Your Honors, drinking perchlorate can cause permanent and irreversible harm to the developing brains of fetuses, infants, and children. Nevertheless, EPA has decided not to set limits on the amount of perchlorate in drinking water and not to require public drinking water systems to test for it or inform people if it's found in their water. We challenge that decision on a number of grounds. I'm, of course, happy to answer any questions the Court has today, but I'd like to focus my arguments on the reasons that it's arbitrary and capricious because it's not supported by the record and is inconsistent with the Safe Drinking Water Act. Now, let me just ask you a preliminary question to that. If we determine that the agency does not have the power to delete the rule in the first place, do we need to answer the other questions? Your Honor, I don't believe the Court needs to answer the other questions in that. Fine, thank you. Why don't you start, why don't you just say a few words about your statutory argument first? About the statutory argument? Sure, Your Honor. EPA is a creature of statute. It has only the powers, in this case, that the Safe Drinking Water Act grants it, either explicitly or implicitly. And it's clear here from the text, structure, and purpose of the statute that EPA didn't have the authority to reverse a regulatory determination that it made 9 years prior. To begin, you see that in the text. The statute says not once, but twice, that once EPA makes a regulatory determination, the agency shall regulate. The second time, the statute sets out a clear deadline by which EPA has to do that. It has 24 months to propose a standard and then another 18 months to finalize the standard. But does it say that the EPA can't withdraw a determination or can't change a determination? Well, Your Honor, it certainly gives clear textual commands that the agency can't do so 9 years after the… Where's the textual support for that? Because a standard is due within, in this case, was due by 2014 at the latest. And once a standard is in place, there's an anti-backsliding provision that protects against it. But that's once there's a regulation. They actually didn't promulgate a regulation. You're right, Your Honor. Because the agency has failed to comply with the Safe Drinking Water Act, they've missed that deadline. Correct. But anti-backsliding doesn't apply unless there's already a regulation. No, Your Honor. But so a statute can… A statute can prohibit… Do you agree? Do you agree with that? That the anti-backsliding provision… I agree that there is no regulation in place right now. No, wait. Do you agree? Because the government, the EPA says this too in the brief that the anti-backsliding requirement doesn't… isn't triggered until there's a final regulation. Your Honor, I think the statute says that the anti-backsliding provision applies once there is a regulation. However, it's clear from the text of the statute that the Congress expected there would be a binding regulation in place within three and a half years. That's a different argument. The anti-backsliding provision says any revision of a national primary drinking water regulations shall be promulgated in accordance with the section, except that each revision shall maintain or provide for greater protection of the health of persons. It specifically deals with regulations. So anti-backsliding does not apply to a determination to regulate. Correct? Yes, Your Honor. I agree with that. However, there are a number of other provisions in the statute that indicate that Congress meant for the agency to take two steps in deciding whether to regulate. There is a ton of process that precedes a determination to regulate. And that is the stage at which EPA is tasked with deciding whether to regulate. And the agency has a ton of discretion in that. It gets to the pick by consulting with the scientific community and the SAB, which contaminants it's going to focus on. So I understand that, but it seems that the same amount of thought and process went into the decision to withdraw the determination. It seems that, yes, there is a framework put in place for a two-step process, which involves a lot of thought and care and research. But it seems to me that this statute doesn't say that you can't withdraw a determination or change your determination. And in this particular case, it seems that the decision to withdraw or change the determination also was accompanied by thought and care and regulation, although perhaps not as much thought and care as they should have put into it, which I'd like to hear more about. But in terms of just the agency's ability to withdraw a determination, I don't see where in the statute, in the text, it says they can't do that. Whether they did a good job of that is something I definitely want to hear about. But just can they do it? It seems that they can. Your Honor, we believe that Congress, in setting up this structure, was trying to demarcate two specific steps to the regulatory process. And there's a reason for that, right? Congress was trying to do multiple things in the 1996 amendments. It is true, as EPA and AWWA say, that Congress was trying to protect against any truly inefficient regulations. But Congress was also trying to force an agency that had been recalcitrant in issuing regulations to regulate. It was saying, go through a bunch of process, make a determination. But once you've made that determination, we're going to hold you to it. You need to issue regulation. Didn't they change the determination in this case because they previously determined not to regulate? That was a preliminary determination, Your Honor. They made a preliminary determination, as the statute calls for. They took public comment, as the statute calls for. And then they made a final determination, as the statute calls for. The statute is clear of what the agency is supposed to do after that, and it's not to flip-flop back and forth on that regulatory determination. In fact, the statute says, consult with the Science Advisory Board, but under no circumstances is that to delay your promulgation of a regulation. Congress knew that the agency would continue to consider the science and nonetheless wanted to hold the agency to the clear, unequivocal command, you shall issue a regulation. What's your answer to EPA's argument that if you're right about this, they lose, the agency loses all flexibility to adapt to newly discovered evidence or better best evidence and that they're locked in by the anti-tax law? Your Honor, that's simply not true. The statute provides for a number of offerings. Now, I think it's understandable to be concerned about what is the possibility? What if there is truly a scientific breakthrough and we realize this isn't a problem at all? That situation is vanishingly, vanishingly unlikely. Well, suppose it happens. I mean, here's an example. Suppose, for example, after 10 years of regulation for various economic reasons, the contaminant is simply no longer being produced. After 10 years of regulation, the anti-backsliding provision would be in place. So it's clear that science isn't, Congress knew that there might be a situation. But there's no more, the contaminants no longer being produced. In other words, the statute provides a number of programs. The anti-backsliding provision wouldn't, the EPA said they would, they would not be able to change the standards. They would not be able to weaken the public health protection. That's true. They could change monitoring requirements to reduce the burdens on water systems. States can issue variances for systems that can demonstrate that they're unlikely to have the drinking water, to have the contaminant in their drinking water system. There's a number of other offerings that the statute provided to deal with that situation. But even if the court is not convinced by our statutory argument, the way that agents, the agency undertook this decision here is not consistent with the Safe Drinking Water Act. And I'd like to start by explaining the problems with the levels of public health per chlorate contamination that EPA selected as its levels of public health concern. Now, let me just pause for a minute and explain why the maximum contaminant level goal standard is relevant to this case. We're not arguing that the Safe Drinking Water Act requires EPA to set a level of public health concern that meets the maximum contaminant level goal. But this court reviews an agency's action on the basis and the rationale that the agency provided in the record. And here, if you look at the final decision document, specifically, if you look at 85 FEDRAG 43995, EPA equates the levels of public health concern with the maximum contaminant level goal standard. The agency says we are selecting, as our levels of public health concern, the three levels that we propose as maximum contaminant level goals. And if you want to find our rationale for how we selected those numbers, look at a technical supporting document that we issued at the time of the 2019 regulations. And if you look at that technical supporting document, and I'm looking particularly at JA 373 and JA 415, it is very clear that the agency selected these levels because it believed they would avoid any adverse health effect. So either there is no basis in the record that explains how EPA selected these levels, or the basis that the court has with which to evaluate these levels is the level that EPA explained, which is the level that's set for maximum contaminant level goals. Here, the thresholds that EPA set do not meet that level. EPA selected three potential levels of public health concern, all of which correspond with measurable decreases in the average level of IQ across the sensitive population. IQ loss, whether it's one point or two points or three points, is an adverse health effect. That's clear from common sense. You can imagine a doctor prescribing a medication to a patient, and they would never tell the patient that this will have no adverse health effects, knowing that there is a population of patients who would suffer, on average, one to three IQ points loss. And those losses are very significant when you look particularly at the ends of the bell curve of IQ, particularly low IQs and high IQs. It has significant impacts for those populations. Is it your view that the MCLG has to be zero? Is that the level of the chlorate has to equate to a zero percent drop in IQ, or can it be five? It has to equate to no adverse health effects. So I presume there is some level of IQ loss that will not be measurable on a population level. And I certainly think EPA could set the level if there's no measurable IQ loss. Do you know what that level is in terms of parts per billion? I don't, Your Honor, because there's nothing in the record that establishes that record, that level. What the record establishes is that it's not 1856 or 90. Your Honor, I see that I'm into my rebuttal time. So if the court doesn't have any further questions at this point, I'll reserve my remaining time. All right, that's fine. Thank you. We'll hear from Ms. Buckley. Good morning, Your Honors. May I please report, Sarah Buckley from the Department of Justice on behalf of EPA. With me at council table is Pooja Parikh from the Office of General Counsel. I want to first address briefly EPA's authority to withdraw the regulatory determination and then address why EPA's technical analysis was a rational exercise of a technical scientific judgment. This court has long recognized that agencies have the power to reverse course change position. Wait, let's just, before you get into that, to me, you know, as I read your brief, the brief argues that the agency has inherent authority, right? That's what, in fact, your brief mentions the word inherent authority, I think, 13 times. But this court has made quite clear that agencies don't have any inherent authority. That authority has to come from a statute, either expressly or implied. So the question before us is where in the statute does the agency have the authority to withdraw? Do you agree with me that that's the question? Yes, Your Honor. I would say that this court has characterized the power to reconsider a decision as inherent authority, as inherent in the power to decide. No, I don't think so. I think agencies have implied authority to change or withdraw a regulation when the statute permits it or when the statute's silent and there's an ambiguity there. But I don't know of any case that says that, any recent case, which says that agencies have inherent authority to change a regulation. They only have the authority that comes from the statute. And all the cases you cite here, every one of them, every one of them, including Fox Television, they are all cases where it's a generic statute. Agency issues a right, Congress directs the agency to regulate, and there's no precise provisions about when and under what circumstances that are listed in this statute. They're totally different statutes. So you have to, you can't, you're not going to get anywhere by citing cases from, that rely on different statutes. It seems to me the agency to prevail here, and, you know, you might, but you have to point to implied, you agree it's not express, but you, so you have to appoint to implied statutory authority. And the question I have is, where does that come from? Well, Your Honor, I think that it is implied by the statutory structure and the text that Congress provided for when EPA can make the determination and what factors it has to rely on. Congress made clear that the determination of whether drinking water regulation is worthwhile when EPA has the authority to regulate. Depends on several factors, including whether in the administrator's sole judgment, there's a meaningful opportunity for health risk reduction. Thank you. I'm sorry, let me just finish. You're absolutely right. You know, and if you look at the statute carefully, that's basically the only place where Congress has given the agency any discretion. It says, it says, if the administrator makes that determination and sole judgment, you're right about that. But there's no other discretion here. It says, it says that once the determination is made, the agency, it says shall regulate, and then it says shall issue the regulation within a specified number of years, and then shall issue the final regulation. So it says shall three times there. Elsewhere, it uses may, so Congress clearly knows the difference. So I don't, I guess I don't see where the implied discretion to withdraw regulation comes from. Well, in two other places of note in the statute itself. First, the fact that Congress expressly makes a negative regulatory determination, not to regulate, judicially reviewable, and does not do the same for a positive regulation. I'm not sure that I'm not sure that helps your case. But I actually think it hurts your case. Well, yeah, I think Congress is basically saying no reason to review the determination because it will end up in a regulation, which will be reviewed then. Right. I mean, I think that hurts your case. Well, your honor to try to change your mind on it. We would be left in the anomalous situation that the only way to address an error or a change in circumstances that is relevant to the factors that Congress gave EPA would be on judicial review of a regulation that EPA has. Well, is that is that really right? Let's you mentioned two things. You mentioned an error and a change in circumstances. Okay, so let's take the air. First of all, there's there's two sets of notice and comment rulemaking and judicial right over the initial determination. So it's pretty unlikely that an error is going to creep through there. I mean, it seems to me quite unlikely that that would ever happen. Your stronger argument is future development, right? Change circumstances, better research, correct? Yes, your honor. But I don't see why the agency doesn't have the authority. You say in your brief, you can't have less stringent regulations, but that's not what the anti-vaccinating provision says. It says that it says that you have to maintain the level of protection. And it seems to me that if there's new research, which shows that less a less stringent standard would maintain existing health protections, that the agency can change it. There's nothing in the statute that would prohibit. Well, your honor, the statute would continue to apply no matter what monitoring and record keeping requirements, which do impose a cost on public water systems. And in fact, those kinds of monitoring requirements for contaminants that are very unlikely to be there is the concern. One of the concerns about costs that motivated the 1996 amendments discussed in the Senate report. So if EPA is forced to go to regulation, even when new information or a more sophisticated analysis, as was done here, demonstrates that the contaminant is not actually present at levels of public health concern with a sufficient frequency, then EPA would be forced in the position that Congress expressly said it was trying to avoid. But it wouldn't. You're talking about the anti-vaccinating provision, right? Not only the anti-vaccinating provision, your honor, but the regulation would necessarily include monitoring. So the anti-vaccinating. Well, does EPA have authority under the statute to provide exemptions for monitoring? Yes, your honor, but only after several years of initial monitoring for the public water system to then demonstrate. And also, what do you think about the provision that says, okay, so it says, it says shall issue regulation, shall issue within specified number of months. And then it says, it also says the agency, excuse me, shall review and revise the regulation every X number of years, right? That's correct. It doesn't say, by the way, it says shall review and revise. It doesn't say, does it, revoke? It does not say that the agency can revoke a regulation, but this regulatory determination is a preliminary step on the road to regulation. And so the fact that the statute speaks expressly about the national primary drinking water standards and what EPA cannot do with respect to those once they're promulgated implies that there is more discretion and the power implicit in this particular statute that we're concerned with here. Congress says if that first determination is made, then EPA shall, right? EPA has to regulate it that way, right? Yes, your honor. Now, once that's done, there is nothing in the statute that says, then they can do away with the regulation, right? That's correct. And Congress's apparent purpose, you can consider the purpose of the statute is to make sure that contaminants that have that, that are past that first step are regulated, are regulated. That would be fair to say, wouldn't it? Your honor, I would amend that to say it's fair to say that Congress was very concerned with regulating where the administrator determines that there's a meaningful opportunity for health risk reduction and that the change in statute from a system prior to 1996, where EPA was obligated to regulate 25 new contaminants every three years, to one that moves that into a discretionary determination, where the agency... I'm having a hard time getting that moved into a discretionary determination. I do not understand why statute doesn't seem to say, if that threshold is passed, then you shall regulate. In this case, for example, you say, well, in two states, they were able to reduce it. Therefore, we take it away. Congressman doesn't seem to have wanted that. They did not give you the authority to do that. I'm like Judge Taylor. I'm not sure there is an inherent authority. I'm not sure there's even an implied authority here to do away with the regulation that Congress told you you had to put there. Well, your honor, in similar circumstances, such as New Jersey versus EPA, where the court has assumed that EPA would have the authority to undo a decision, except if there is another mechanism capable of rectifying the error, and in that case and in others, it is only where EPA did not follow the process that Congress has set out. But Congress didn't set one out here, right? That's right, your honor. So that could be an indication that Congress didn't want one. Congress did not want a way out. Uh, I hate to suggest this, but the lawmaking branch of government is the legislature, and if they chose to change that, they could, but at this time, they didn't put it in there that the agency could do it for them. I'm not sure where it's coming from that they have an implied authority to undo what Congress told them they had to do. So if you look at the statutory language, what triggers the obligation to regulate, shall regulate, is the determination. That's correct. You make a determination, then you shall regulate. But if before you actually promulgate, you withdraw the determination, or you change the determination, is there still a duty to regulate based on the withdrawn determination? It's EPA's position that there is no longer a duty, and in fact, no longer authority to regulate. So I think that the statutory language does seem mandatory. Once there's a determination, you shall regulate. But there is nothing in the statute that says that before you've actually promulgated, if you withdraw the determination, or you change the determination, that you still have to regulate according to a determination that has now been withdrawn. We agree. And this court has never found that the existence of a deadline in a statute is a reason to indicate that the authority to reconsider a determination is abrogated. And this court has also not seen that language such as shall abrogates that. In the Idaho Conservation League decision, the EPA's prioritization, the deadline was 26 years prior, and the court still affirmed that EPA had the authority to determine not to issue regulations if that's what the record before the agency demanded. And here, I think it's important that we're not in a situation, an extreme situation that could arise, and that NMDC's position would hold EPA's incapable of rectifying a clear error, like a study that the determination was relying on is retracted, or there's a computational error. I understand that. And I'm just wondering, if the statute does not speak to whether a determination can be withdrawn or changed prior to promulgation, is the agency entitled to any deference about that? Deference in determining whether it has the power to withdraw or change a determination before promulgation. I think the agency would be entitled to some deference rather, but I think it's the clear best reading of the statute and fairly in line with this court's precedent about agency's reconsideration. Let me make sure I understand your question to Judge Vance, your answer to Judge Vance's You're saying that it requires no deference on our part to accept your argument that the agency can withdraw a determination anytime prior to issuing a regulation, that that's clear in the statute? Is that your point? I'm not challenging, I'm just asking you. In other words, in the old days when we had Chevron, we used to talk about Chevron 1 and Chevron 2. Is this clear as far as you're concerned, or is this something we would have to defer? We would have to defer to the agency's judgment. I think it is clear, Your Honor. And where does the clarity come from? It clearly comes from the implied elements of the statute that we've discussed about the factors EPA has to apply, the judicial review provision, anti-backsliding only attaching after a regulation is promulgated. But it is also clear from this court's precedent about the power to reconsider, the power to reverse course of being implicit in, if not inherent in, the power to decide in the first place. So if I were to think, you know, maybe EPA's right about this. Maybe it does have the power to withdraw before it regulates. I'm not saying I do, but if I did, you know, but if I don't think it's clear, that is, if I think it's an ambiguity in the statute, what would we do there? You don't claim any Chevron deference in your brief. That's correct, Your Honor. We didn't raise that. So what would we do? I think you should deny the petition for review. Well, we can't. You haven't claimed Chevron deference, and I think the statute's ambiguous. I'm not saying I think it's ambiguous. This is my hypothetical. I cannot think, I'm canvassing the cases that I can remember in my head, Your Honor, and I cannot think of another case that involved this implicit or implied agency authority to reconsider, resting on the agency's interpretation of its statute. Is there some reason why you don't want to rely on Chevron or ask us to defer to the agency? I'm just wondering. Well, Your Honor, I'm not sure that in the agency decision itself, Chevron or the agency's interpretation was invoked. If the agency discussed the reasons that I laid out for you today as its interpretation of why it had the authority. So I suppose, yes, without invoking Chevron by name, the agency certainly offered an interpretation of the statute and to the extent that you think that the statute is. Maybe you just don't want to sign in, decide CNRDC versus Chevron because NRDC lost that case. Well, Your Honor, I'm trying to stick to the arguments that we advanced in our brief. And grounds that the agency set out in their regulation. I apologize, I'm talking around it, but I guess. But is Chevron waivable? Chevron deference is not waivable? No, Your Honor, it's an interpretation canon, interpretive canon. In any event, just to briefly touch on the merits of the decision here and to explain what EPA did from a technical standpoint, I think there's two parts to this analysis. First is the legal question of what target was EPA shooting in this analysis? And I think it's clear from EPA's preamble here that it was making a determination about the levels of public health concern. Yes, it relied on a health effects analysis that was developed for the purpose of setting an MCLG, but it never finalized a regulatory determination that those are the levels that met the MCLG standard. It's final action here was applying the health effects analysis, knowing what we know about how perchlorate actually leads to adverse effects that is not present in public water systems with a sufficient frequency at levels of public concern. And then on the underlying merits of that technical analysis, EPA rationally and at the behest of the Science Advisory Board, developed a state of the science cutting edge model to, for the first time, connect the literature on iodide uptake inhibition. So I find it really surprising and disturbing that the EPA proposed a MCLG that is associated with a 2% drop in IQ in the sensitive subpopulation. And the definition of an MCLG is a level at which no known or anticipated adverse effects on the health of persons occur and which allows an adequate margin of safety. How could the agency possibly find that a 2% drop in IQ has no adverse effect on the health of persons? The question boils down, Your Honor, to a data interpretation question. For this type of modeling, biologically based dose response modeling, the way that you interpret whether you are observing an effect from the contaminant is to look at a relative change from the mean. As EPA's benchmark technical dose guidance said, a range of those relative changes from the mean have been used. One standard deviation, 10%, or for data such as here, data from epidemiological studies, 1%. EPA considered the range, proposed the range of 1% to 3%. And in the end, finalized the determination that it does not appear with sufficient frequency at levels of... I still don't understand because it still boils down to a 1% to 3% drop in IQ and a finding by the agency that this is not an anticipated adverse effect on the health of persons. And if we think that that's arbitrary and capricious, you would lose on the merits of your exercising authority if we find that you have the authority to change this. Yes, Your Honor, I think that the important point is that this is a data interpretation exercise. That when you have a continuous outcome like IQ and you're using a model such as this, you have to be sure or have some degree of certainty that what you're observing is not natural variation, but you're observing the effect of the contaminant. And it's the relative change from the mean that scientists use to determine whether they are actually seeing that adverse effect. That's consistent with the technical dose guidance. And as you heard earlier, NRDC hasn't been able to identify in the record anything that would have supported EPA using a lower level. So these levels of perchlorate are associated with drops in IQ. 1%, 2%, or 3% drops in IQ. And you do not look, and there's nothing in the record, as we just discussed with opposing counsel, about a lower level, which would be associated with no drop in IQ, which would seem to fit the statutory definition of MCLG. So if you didn't consider that, how is your action not arbitrary and capricious? Well, Your Honor, because the approach to defining the adverse effect, there are multiple approaches that one can take. One could take a no adverse effects level approach. That is, as I understand it, a technical way of interpreting data using different tools. Is that statutorily required by the definition of an MCLG? No, Your Honor. Our position is that the statute does not dictate what particular technical approach EPA has to take. No, no, no. But it has to have no anticipated adverse effects on the health of persons. I understand, Your Honor. Statutorily. Statutorily, correct. And EPA attempted to design this model in order to set that level. Its multiple peer review panels determined that its model was fit for purpose for doing that. And when you have that kind of model— I'm not reviewing fit for purpose. I'm reviewing whether you've met the statutory definition of an MCLG, which based on the record, it seems that you have not. And then you equated that with the standard to regulate. You equated it with the standard for levels of public health concern. Your Honor, the statute contains two legal standards there. I understand, but you equated them. What EPA did and what EPA said was that it used the public health effects analysis that it did admittedly create for purposes of setting the MCLG and said that, well, at those levels, even at our most stringent proposed MCLG level, not defending that as the MCLG— At your proposed level, there could have been a lower one. Your Honor, the agency put those levels out for public comment. There is no public comment supporting using a smaller than a 1% relative change from the mean for this purpose, for this technical purpose. And part of the reason why EPA had to use this particular model is that there is not, in fact, literature that directly says that perchlorate causes adverse neurological developmental effects. What we have is a literature that connects perchlorate to iodine uptake inhibition, which both the National Research Council and the Science Advisory Board told EPA is a precursor to adverse effects. So when EPA brought this problem of how should we set an MCLG to the SAD in 2012, the SAD recommended that EPA develop this kind of model in order to actually connect that literature to the other existing literature, which was how does thyroid disorders and iodine insufficiency relate to neurodevelopmental effects. So EPA developed the model to connect the two literatures for the first time to actually connect perchlorate to adverse effects. And it did so because it needed to meet that statutory standard. It didn't believe that relying on the health reference levels that it had used for the 2011 determination and before, which, again, focused on iodine uptake inhibition, would support a determination under the statutory standard for an MCLG. In doing that, EPA then found for reasons that include the body's ability to generate other thyroid hormones to make up for an iodine deficiency, that the actual level that would be associated with adverse neurodevelopmental effects in the sensitive subpopulation was higher than what they had considered based on the health reference levels from iodine uptake inhibition. And so given that conclusion and using the interpreted questions there, EPA then said, well, our levels of public health concern, that is a different, that's a different text than the standard that applies to the MCLG. It substantially means something different. We think it means it's more flexible. We think we have used health reference levels before. And we think that given the data before us, that this contaminant is not present with a sufficient frequency at levels of public health concern, and that we lack the authority to regulate it. So I'll just say that you seem to be trying to rely on this model, which is subject to criticism from the scientists. Let me just breathe. But the bottom line is what you came up with is a level of perchlorate that is associated with significant drops in IQ level. And there's a statutory standard that talks about adverse effects on people. And I don't see how that fits. And then in the regulation, it says EPA use these potential MCLGs as the levels of public health concern. And so if you're going to equate these MCLGs that do not meet the statutory standard with the levels of health concern, it seems that you have not properly performed your function. Because there's, in my view, clearly an error in terms of setting perchlorate levels that allows what appears to me a significant adverse effect on persons, and therefore does not meet the statutory definition of an MCLG. And so I feel like the agency is trying to hide behind a screen of, this is highly technical, we used a model, and saying that that should be enough. And I don't see how that fits the statutory regulatory standard. I understand your argument. I won't belabor it much further. I will say that, again, there are two questions here. What is the standard that applies to EPA's action? EPA did not finalize this as an MCLG. I'm not here to defend the agency's MCLG decision because it did not make it. Instead, it determined whether or not this contaminant was present at levels of public health concern, and whether it presents a meaningful opportunity for health risk reduction. But you define the levels of health concern based on these MCLGs, and you don't know anything below the 1%. We don't have information except about these MCLGs because that's how you define levels of public health concern. And then again, Your Honor, I think the definition of the levels of public health concern were at the levels that were also being used for the MCLG. But that is an underlying scientific analysis, not unlike the health reference levels that were used before that EPA could then apply to the statutory mandate it was given. So there is an underlying body of scientific evidence, and EPA applied the statutory standard to making its regulatory determination. In addition, on the technical point underlying, I think it's important to note that not only is this consistent is using a relative change from the mean consistent with how you interpret data in this kind of model, but EPA built additional conservative assumptions into that model. First, in defining the sensitive cell population as the thyroid deficient mothers, first trimester mothers, the offspring of those thyroid deficient mothers, and built in a number of extremely conservative assumptions about those mothers, thyroid function, iodide intake, and TSH. Once after applying the literature to that, it applied additional conservative assumptions about how much those individuals take in chlorine in their drinking water versus their food, their body weight, etc. All of those conservative assumptions were put in there to be health protective and help support EPA's conclusion that this is a level that would not be associated with public health effects of concern. One other point, just a technical one, the level 18 micrograms per liter is a level that would avoid a 1%. I'd like you to explain that, because you say that several times in your brief, avoid a 1%. I thought these are levels that are associated with a 1% drop, which implies that it causes it, not that it avoids it. The idea is that if we were to regulate at that level, it would avoid that 1% effect. It would avoid that 1%. Why? Because they're allowed to be at that level. That's the maximum level. That's the maximum level goal is let's have that or less. And so if you had that or less, it would avoid a 1% IQ effect. They can go right up to that, which would be a 1% drop. They can go right up to 18, if that's the level, right? So that is associated with a 1% drop. If you're setting a maximum limit, you can't assume that everybody following your regulation is going to be way below it. They're going to go right up to what you're allowing. Your Honor, one quick point just about MCLGs versus levels that they would have to comply with. MCLG is the goal, and then a regulatory level can be set higher than the DLG. But even here, the question was, if there's per chlorate out there in what we're making water systems at 18 or higher, are we concerned about that? Is that a level of public health? I understand. And at 18, it's associated with a 1% drop in IQ. So I don't see how that avoids a 1% drop. I think, again, it was because if you regulate that level, it would avoid it. It was the terminology of the agency. The Court has no other questions. I see my questions. All right. Thank you. Ms. Snow. Good morning, Your Honors. May it please the Court. I'm Brian Snow on behalf of Intervenors American Waterworks Association. I'm sorry. Could you speak up a little, please? Yes, Your Honor. I'll start by discussing the authority to withdraw determination. I think an important distinction here, Your Honors, is this is not a regulation. It's a determination. That is a preliminary step towards regulating. And Judge Tatel, I just wanted to clarify that because in talking about some of the case law earlier, you mentioned inherent authority to withdraw regulation. And here we have what is clear in the statute to be a preliminary step towards that. Judge Pan, I think you have the absolute right reading of the statute here for that reason, that the provisions that would require a regulation are where the provisions say when you have a positive determination, then there is a requirement to issue a regulation and to do so on a specific timeline. But when you no longer have— The statute says, hold on a second, okay? So this isn't a statutory— It says it. It says that once the determination is made, the agency shall regulate, right? That's what it says. It says once the determination is made, I'm not quoting the statute, but it says shall regulate. It doesn't say the next sentence says shall issue a regulation within a certain number of years, but it does say determination is made, it shall regulate. Isn't that what the statute says? It says for each contaminant that the administrator determines to regulate. And here we no longer have that positive— It's a determination. It shall regulate. But here we no longer have that positive determination. And I think, Your Honor, the more important piece here to fit into that is the three statutory criteria for regulating have to be present at the time of regulation. If EPA determines that those statutory criteria are not present when it goes to issue that regulation, it cannot issue a regulation. And so it cannot be that this determination is the final question, especially because we know that Congress specifically added these three statutory criteria out of a concern with over-regulation of substances that just weren't occurring at frequencies and levels of public health concern. It would make no sense for Congress to have added those three statutory criteria and then not give the agency the ability when it went to issue that regulation to withdraw that determination. So I do think Judge Pan has the best reading of this statute that there is a requirement as long as there is a positive determination. But if there's no longer a positive determination, then there's no longer a requirement to issue a regulation. And do you have an opinion about deference to the agency on this? Your Honor, again, I don't think we believe that you need to have deference to the agency on its interpretation here because I think it's clear from the statute that you have to have those three statutory criteria, but you certainly could defer because this is how EPA also explained why it was withdrawing its determination. And again, I think, Congress would not have thought that they needed to say anything different about this given that this was just meant to be a preliminary step towards regulation and regulation would then occur on the basis of those three statutory criteria. Judge Pan, I also just wanted to address your concern about the proposed levels that EPA is talking about here. So again, unfortunately, the science is quite complicated in a situation where you don't have a binary, good IQ, bad IQ. It's not the same as, are you sick with this? Are you not? And so the agency had to take a number of steps to try and correlate what we do know about perchlorate and iodide inhibition with any possible neurological impacts. So unfortunately there's not a smoking gun study here that shows that perchlorate and drinking water at these low levels corresponds to any specific neurological impact. What we have instead is knowing that perchlorate given at very high levels in medical doses impacts iodine intake. And we know that that can lead to impacts of iodine deficiency. And we know that those can lead to impacts on maternal hormones. We know that that can lead to neurological development issues. So the agency was trying to connect all of these dots and pick numbers that they could then use to look at the occurrence data we have for the public water systems and determine if there was going to be a meaningful impact or opportunity for national regulations to have a meaningful impact. Now what we know and what's in the record is not just that EPA looked at these three levels that they picked as ways to correspond the best version they could of figuring out whether or not there'd be a health impact at levels of perchlorate. But we also have all of the data from the national survey that EPA did under UMCR 1. But we know that perchlorate even at the lowest levels of detection was still showing up in less than 2% of those surveyed results. So the agency had before in the record evidence not only at these three levels but evidence of the overall level of happening in our water systems. And I think that's important because these statutory criteria were meant to address not just levels but also levels of occurrence. So here we know that only based on EPA's record only 15 water systems out of more than 62,000 around the country. Less, this is about 0.02% of the water systems would have perchlorate at these levels that would need to be regulated. We didn't get a chance to address this with the EPA but in determining what the prevalence was they apparently updated data from 2001 to 2005 with more recent data from California and Massachusetts but only replaced the samples that were positive for perchlorate. Why didn't they replace all of the samples from California? Why did they focus only on the ones that were positive for perchlorate? Because the claim is that that biases the data. So Your Honor, first I just want to point out that you can look at the also in the record the 2008 determination where EPA didn't remove any of these samples and they came to the exact same conclusion. Removing these samples really didn't move the needle. Even there they said less than 1% of water systems in this country had perchlorate at levels there. They looked at 15 parts per billion which is even lower than the levels they looked at here. So again, we're talking about a very, very small percentage. So what happened in 2008? That was when they made the preliminary determination not to regulate perchlorate. So EPA has reached the same. But did they use the UCMR? They used the UCMR. So exact same set of data without removing any of these samples. Came to the same conclusion. Again, found that less than 1% of our water systems had perchlorate even at their 15 parts per billion. So lower than the 18 parts per billion that they used here and determined that regulation, they made a negative determination regulation was not warranted because of the very low levels of occurrence happening around the country. As to the specific pieces of data that they removed here. So EPA actually used neutral criteria in what they were doing. They did a couple of one where there was updated information from the states that now regulated indicating that there were no violations of those state regulations, which I believe California is a six parts per billion. Then they knew that those water systems no longer had those elevated levels of perchlorate. And so they updated based on that. Yeah, they selectively updated, correct? They only updated the samples in the UCMR that were positive for perchlorate. Why did they just replace all of the California data with the new California data? Your Honor, I think because state regulation would have required all of those water systems to be below that level, right? So we know that those water systems no longer have the elevated levels. But again, I don't think that's something that this court needs to get into because within the record that was before the agency, they again had that all the data showing that even if you used UMCR1, even if you didn't remove any of those data samples, you'd get to the same place. This perchlorate is, and this is good news, Your Honors. Perchlorate is not occurring in our water systems at levels of public health concern. That should be something that everybody feels very good about because there's certainly a number of other issues. In this case, that's what your position is, but they don't seem to think so. Understood, Your Honor. But we also have a statutory scheme where the third factor is that this has to, in the sole discretion of the EPA administrator, provide for a meaningful opportunity to impact the public health concern. I completely understand why it makes everybody uncomfortable to talk about drops in IQ. And we recognize that iodine deficiency is a real public health concern, but based on the record before the agency, Safe Water Drinking Act regulation doesn't provide a meaningful way to address that public health concern. I'd say a good example or analogy that the EPA has dealt with in the past is sodium. Sodium is ubiquitous in our drinking water, and we know that sodium can cause heart disease or stroke. The EPA came in and did an analysis and found that the contribution of sodium to the health concern was just not a justification for regulating of our public water systems. Again, Congress introduced these three statutory factors out of a concern that when we add substances to Safe Drinking Water Act regulations that just aren't at national frequencies, then frankly, what we're really doing is overburdening our public water systems and distracting and distracting from their ability to monitor and come into compliance with more pressing health concerns. That's exactly why the statutory scheme was put in place. And that's really the issue you would have if you were to force the agency to regulate perchlorate here, because the record shows that it is just not occurring in public water systems at levels and frequencies on a national basis that would allow for there to be a meaningful opportunity. And our excuse me, NRDC says in its brief that that since the passage of these amendments, the 20 some years since these amendments were passed, EPA hasn't issued final regulations for any contaminant. Is that true? That's correct, Your Honor. Only there are about 90 contaminants that are listed right now. EPA goes in by statute every five years and has to consider five additional contaminants and whether or not it's appropriate to regulate those. EPS has to come in every six years and look at the contaminants that are already being regulated and determine if they should make the regulations more stringent. So there's certainly plenty that's happening under this statute. But thus far, EPA has not determined that that there was a reason to regulate additional substances outside of those original 90. Any other questions? Thank you very much. Ms. Fort, you have two minutes. All right, Your Honor, since a lot has been settled, I'll try to get to as much of it as I can. On the occurrence analysis, council's numbers assume that EPA undertook a reasonable update. And the reality is that the agency did it. The agency only went back to two states, the two states in the nation that are the only two places that have since regulated perchlorate. And within those two states only went back to places where perchlorate had previously been detected. Of course, those sites have gotten better. We know there's a problem and the states have since regulated it. That says nothing about what is happening on a national basis. And it's absolutely not a reasonable basis for the agency to decide that perchlorate is getting better. I will also say that it is not safe to assume that because the other places are subject to state regulation that they are necessarily in compliance. EPA's own brief at page 51 shows that California has detected perchlorate exceedances in 11.4% of large public drinking water systems between 2007 and 2018. It's simply not true that this is a problem that has been solved. I also want to go back to the EPA's point that this is about a model. This is a scientific analysis. We're not challenging the model. We're not challenging the agency's scientific analysis. The agency itself described the levels that it selected as a policy decision. EPA says there's nothing in the record to suggest that a lower level should have been selected. I'll refer the court to JA 743, which is EPA's own response to a comment from AWWA, in which it says, small changes in IQ become important when considered on the population level. For example, one study argues that although a one point change in IQ is within the standard error of an individual measurement and would not be regarded as clinical disease, it is still highly significant on a population level. That is completely inconsistent with the statutory command that the Safe Drinking Water Act has given EPA. I'll refer the court on the question of whether it avoids IQ loss to 84, FEDRAE 3557.01. And finally, Your Honor, EPA, on the reconsideration point, EPA is trying to preserve its authority to resolve a situation that is entirely hypothetical. The possibility that there might someday be a contaminant on which there is a dramatic change in science. It is failing to respond to the actual problem that Congress is trying to address, which is that there are a number of contaminants out there that we're not testing for, that people don't know if they're in their public water and that are causing real harm to public health. And that is the problem that Congress was trying to solve. And that is why Congress said, once you make a determination that it meets that criteria, we're going to hold you to it and we're going to make you actually protect public health, which is the purpose of this statute. Thank you. Thank you, Your Honor. The case is submitted.