Oral argument — Billings Clinic v. Eric Hargan

Case number 17-5006 Billings Clinic et al. Appellants v. Eric D. Hargan Acting Secretary, U.S. Department of Health and Human Services Mr. Collins for the appellants, Ms. Foster for the appellee Ms. Foster for the appellant May it please the court, I'm Sven Collins, I'm representing the appellant hospitals. Today I'll refer to the appellee secretary as either the secretary or as CMS. And the court has raised a question of jurisdiction and I would just make two remarks on that and prepare to address any further questions the court may have. The first is that only a subset of the hospital's claims here at issue would be subject to the jurisdictional question the court's raised. And the remaining claims challenge all of the secretary's rulemakings at issue that are in the appeal. Is that true of 2011? I thought there was only one provider who raised a claim as to 2011. I'd have to verify this again, but my understanding is that for all four of the years there are hospitals who did not have the jurisdictional question that's been raised by the court. Billings raised it for 2011, but I'm just not sure if they were a certification or a dismissal. Your Honor, we looked at that issue and my understanding is that all the hospitals received an authority determination from the board enough to challenge all of the years. But of course, you know, we recognize that if that's not the case, that's an issue that relates to the jurisdiction on that particular year. So that eliminates the jurisdictional question, is that what you're saying? Correct. And the second point... Can you point us, maybe after argument, can you point us to where in the record it shows, at least as to 2011, there are hospitals that were not dismissed? Yes, I will, Your Honor. The second point is that all of the cases and claims as to which the jurisdictional question had been raised, they were presented to the agency, presented to the agency board. The board dismissed the case. They came up properly through appeal through the statute, 1395 OOF, and during the district court proceedings, the secretary waived any further exhaustion of those claims before the agency. And that is at the joint appendix at pages 93 to 94. But before you do that, when you say that as long as there's some hospital challenging every year, that eliminates the jurisdictional question. Is that right? Because each of these hospitals want their own money. And so for those that were dismissed and that are now here, don't they still want their own money? And so we do need to find jurisdiction as to each individual hospital that wants money or not? Absolutely, Your Honor. So it doesn't eliminate the jurisdictional question that there's somebody for each year who got a certification from the board. I'm very confused. And number one, my belief is that 100% of the threshold rulemaking is being challenged. There are plaintiffs who received, felons who received, and no authority determination from the board with respect to that. And obviously you raised the question about 2011. No, I think what she's asking is even if there are some hospitals as to which there isn't a jurisdictional problem, what about the hospitals as to which there is still one because there wasn't certification, wasn't a checkup, whatever? And that's the second point I'm addressing. Before you get to the second point, which is an argument about we have jurisdiction anyway, isn't the answer to that if we find jurisdiction, decide in those hospitals' favor, and those hospitals go back now to the board and the secretary has already waived exhaustion, they will, HHS will be bound by our decision in this case. So those hospitals will get whatever benefit everybody else gets. Yes, Your Honor. We believe they would get the benefit of the balance of the hospitals, this Court's rulings as to the thresholds themselves. And this raises, I believe, a novel issue, and it probably ties into this Court's line of health decision, which spends a bit of time talking about the board's no authority determinations. And we'd be happy to brief this further. The board's exhaustion requirement, we believe, can be waived, and here the secretary did do that. And what would happen as a practical... I'm sorry, is that waiver what creates jurisdiction in those cases, the ones where they were dismissed? Well, what creates jurisdiction is the statute where the hospitals filed appeal to the board, the board dismissed the case for lack of jurisdiction, and then in the district court the court ordered that there was jurisdiction, of course, but that was with the secretary's waiver of any further exhaustion. I get that the waiver happened after the complaints were filed. As to just those hospitals for which the board dismissed and certified their claims, then if they come to district court, the only thing they can bring up for review is that dismissal decision, right? They can't bring up a merits decision, because there is neither a certification nor a merits decision to be reviewed in those cases. Well, and that's where Judge Collier below determined that sending the case back to the board for entering a ministerial no authority determination was something that the court could find could be waived and have been waived. The problem is subject matter jurisdiction has to attach to the time the complaint is filed. And the waiver did not happen within that time frame. It didn't happen within the statutory time frame of, I think, 60 days for reversing a board decision. And the board's regulations, their own regulations, say that if the secretary is overturning a jurisdictional dismissal, the case will then go back to the board for the board to certify or to make a final decision on the merits. And the statute seems to require one of those two things before there's district court jurisdiction, certification or final decision on the merits. Neither of which they had. Well, the district court had absolutely original jurisdiction over the appeal because a dismissal from the board based on jurisdiction supplies the jurisdictional basis for the appeal. For any issue they want to raise or only for that decision to dismiss for lack of jurisdiction. That's what I was trying to ask. No, and you're right. You're very astute question that ordinarily, and we prevailed on the jurisdictional question. No question. Ordinarily, it would go back to the board to enter a ministerial expedited judicial review order. The no authority review decision that it entered for every one of the other hospitals, even some hospitals in the same expedited judicial review order. I get the sort of, it seems ridiculous to send it back so we know what the board is going to do, but I just want to ask you, you had a case where just one hospital and all that happened was the board dismissed it for lack of jurisdiction. And then they filed a complaint in district court. Could that complaint challenge, certainly could challenge the dismissal for lack of jurisdiction. Could it also raise merits issues? The complaint would also need to seek relief, which is what the hospital plaintiffs sought in our case, which is that it remain in the district court, notwithstanding the ordinary remand, and for the district court to make a finding that the jurisdictional, the ministerial staff. The jurisdiction had been waived. Yes. Can I ask you a question? I think this is more maybe one for the government counsel to think about while waiting. But the statute doesn't say anything about holding off while HHS decides its jurisdiction. The statute that gives us jurisdiction says the provider files a request for determination by the board of its authority to decide the question of law. The board shall render such determination writing within 30 days. The board fails to render such determination, the provider may bring a civil action. That's exactly what happened here, right? That's correct. There's no determination on your merits claim and on your claim about the regulations within 30 days. Is that right? That's correct, Your Honor. Okay. So as far as the statute is within 60 days. So as far as the statute is concerned, we have jurisdiction. Correct. The only thing that gives us any pause is that the agency itself thought it was without jurisdiction because of the lack of the certification of the amounts, right? Yeah, the board did. Correct. The board. And that comes not from the statute but from a regulation of the secretary. Yes. Correct? Correct. And whether or not, and although the secretary cannot waive our jurisdiction, the secretary can certainly waive the secretary's own regulations, at least as long as the secretary doesn't complain about it, which the secretary isn't doing. And you're not complaining about it, right? That's absolutely right, Your Honor. Okay. That's a question really for the counsel for the government. But that still doesn't answer, even if that establishes that we have jurisdiction, it doesn't answer jurisdiction as to what. It doesn't necessarily answer it, right? Because it doesn't necessarily tell us that we have jurisdiction as to the entire merits. Well, I think under the process Chief Garland laid out, the secretary's failure, the board's failure to issue a no-authority determination within the 30-day window would trigger the right to expedited judicial review under 1395-00F. So you think it bakes in the expedited judicial review, which then brings the entire kit and caboodle before us? Well, I think that's another avenue. I also think that the government's waiver in the district court, that there's no need for further exhaustion because once that happens, these claims purely come within the purview of the expedited judicial review provisions of the Medicare Act, which is the provisions through which these cases are appealed. Now, this case picks up where the Banner Health case left off, and it challenges the secretary's process for setting the outlier payment thresholds for 2008 through 2011. Banner Health recognized that a critical flaw in the agency's process to set the thresholds in 2004 through 2006 had been its failure to adjust for the uninterrupted decrease in the hospital charge ratios, and that failure had caused the secretary to set excessive thresholds that were not likely to pay out at the agency's 5.1 percent targeted amount. In other words, to bake in underpayments. Although the Banner Health court decision let stand the agency's first attempt in 2007 to account for the decrease in the hospital charge ratios, the subsequent record facts here demonstrate that CMS's method arbitrarily continued to bake in underpayments. CMS has argued for deference, and I'll get to the reasons why shortly, why deference is not due, but I'll just put to you four big-picture facts here. And I don't know if we have more time for our argument given the jurisdictional question, but I'm just going to focus now on the four things I want you to think about. Number one, the agency recognized that it had to adjust for the decrease in the charge ratios. It did that in 2007 because it had underpaid hundreds of millions of dollars when it had failed to do so in the prior three years. And that was central to this court's reasoning in the Banner Health decision with respect to those three years. Now, in 2008, after the secretary had applied its complicated method that didn't track the historical record trend. So they used a new adjustment methodology in 2007, is that right? Correct. So nothing that comes before 2008 is really relevant, right? There's several things that are relevant, Your Honor. One is how effective was the method? Because the agency adopted a method that didn't track the historical record trend. And, in fact, this court inferred that the agency might have thought that the record trend would have slightly trickled. I guess what I'm asking is if they were below for several years before, then they adopt a new method, which we blessed, and then they're below the next year. You can't look back and say, well, they've been below for many years without adjusting, because they did adjust. Now the question is whether this new adjustment is good enough. Exactly, Your Honor, precisely. So in 2008, we don't really have a history to rely on. That is, a history of bad adjustments in the past. Isn't that right? Except for what happened in 2007. Well, exactly. What happened in 2007, the first year they attempted to apply an adjustment, and it actually resulted in more underpayments than it had resulted in 2006. So pause on that fact. The next fact is in 2008, when they applied their method a second time, to account for the uninterrupted decrease in the charge ratios, the Secretary increased them. And in 2007, that irrational result was not addressed by the Secretary at all in the rulemakings. And the 2007 estimate the Secretary had made that there would be, essentially that the cost-to-charge ratio rate of decrease would slow to almost a trickle, did not come to fruition, because the 2008 record demonstrated that, in fact, it had continued its uninterrupted decrease at a much higher rate. And against this, in the 2008 record and after that, the Secretary did not provide any explanation of the record facts compared to its assertion, which unsupported assertion that its method was more accurate than adjusting using the record trend. And it's easy to understand why the Secretary's method went haywire in 2008. And that's specifically because the agency was modeling a change in cost-to-charge ratios. Cost-to-charge ratios depend on, and they're computed from, contemporaneous data. Costs divided by charges from the same fiscal period. But the Secretary's model used charge data from the most recent period and paired that up with cost data from four earlier fiscal periods. And the problem with that is costs and charges, charges were increasing faster than costs consistently, but they were both increasing at different rates. And so using the non-contemporaneous data made the Secretary's modeled adjustment factor ensured regularly and predictably that it would be divorced from the record facts, and as the record shows in each of the rulemakings here at issue. Now... With the actual payment, and... Yeah, and the Secretary defined that as 5.1% because that's the amount they offset from the hospital. Yes, projected rates. So now that the Secretary each year would look back and estimate how much, how close did it come to the target. So in 2007, when it did its look back in 2008, it saw that it had paid less with an adjustment factor than without. And then in 2009, when it looked back in 2007 and 2008, it saw again its re-estimate of 2007. Yes, indeed, we paid less. And then 2008, we barely moved the needle from where we had been before there was an adjustment factor. And in 2010, it looked back and found that it was actually above for 2009. Yeah, that's correct, Your Honor. So they missed two years, they undershot two years, and overshot one year. They undershot, and in fact, the record data which was presented to the Secretary in the 2011 rule showed that 2009 was in fact an underpayment. The commenters noted if you don't model past payments, if you actually look at the actual claims date. We have to decide each year at a time. You can't look into the future when you don't know what you know. But I take it your argument is in 2008, they knew that they had undershot in 2007. In 2009, they knew they had undershot in 2008 and 2007. But in 2010, they knew that they had overshot in 2009, right? Not only did they think that, but so did you. Correct. In 2000, you're absolutely right, Your Honor. So just as a big picture, why is it unreasonable for an agency to say, okay, we're only making estimates here, we undershot two years, and we overshot the third year? Well, first of all, Your Honor, there's not a word of argument or explanation like that in the record. That seems blindingly obvious. The whole point of your argument is that they're underpaying. And if the fact is they underpaid you in the first two, by estimation, and in the third year they overpaid, it seems, I don't know what the right race is, it seems obvious that that's their rationale for continuing along. And they did say in the 2010 record, they said, aha, the fact that we slightly overpaid means that our method is sound. But in the prior two rules, they said nothing about the fact that they had underpaid and concluded our method may not be sound. And, in fact, the record showed it wasn't sound. And, Your Honor, it would be a bit like if an archer aiming for a target fires five arrows, four go into the woods, one barely hits the target. Would this court defer to the archer's judgment if the archer can claim to be a marksman? No, but this is more like a statistical problem than an archer problem. And you have a mathematical formula which creates a set of plots on a line or on a curve. And you're trying to hit certain spots, but you can't be sure you're going to hit the exact spot that you want. And sometimes just by statistical variance, you end up not hitting it. And you don't have to really explain other than as long as, you know, you have a model that you think is going to get to the right spot, no model is expected to hit exactly on target every time. No, but at Santee, you are not going to be likely to hit that spot if you're using a method that the record has proved two times in a row to not work. And that is what we're focusing on here is not that they had to hit the ultimate 5.1% target. Of course, underpaying it or hitting it may reflect on the efficacy of their method. But the more focused... You said two times in a row. So in 2008, we don't know two times in a row, right? In 2009, we know two times in a row. In 2010, we know two times in a row and then one time over. So what do you do with that? So does that mean you win only in 2009? No, Your Honor. You just told me two times in a row is what requires it. No, I said that the archer fired four arrows. I may have said five. But we don't know those four arrows until you start looking at 2011. In 2008, the archers only fired one arrow. Your Honor, in 2008, the agency knew these facts. They knew that their prediction for 2007 for the adjustment factor was way off. Both of the two things they're modeling, one, the rate of cost inflation, that they're going to pair up with the rate of charge inflation, way off. They predicted an increase, not an increase. Two, the rate was going to decline to a trickle, which begs the question, why would you have an adjustment factor in the first place? If this was such a big problem that the agency was going to acknowledge it in 2007 that we needed to do it, why would you have one that goes to trickle? But the most glaring thing that happened in 2008 is the fact that its method predicted something that was counter to all the data. There is no data that would support it. So if the method were adjusted in response to this, if the methods were adjusted so that there would have been less of an underpayment, then there would have been more of an overpayment later. Could you repeat that? So if the reaction in 2008 is that something is deeply flawed, we're still getting underpayments, so we need to compensate for that, and the model is, say, you just make in a new factor that says we need to adjust it downwards by an additional .2 or something because we're predictably coming in under that, then it would have turned out there would have been more of an overpayment in a subsequent year in which there was an overpayment, right? Well, not at all because of that method. And, in fact, I mean, that's a great question. What the agency knew is that its model was not effectively addressing the issue. But in 2009, CMS both said, without any analysis, that our method must be sound because we have a slight overpayment, we need to raise the threshold. But also there were many other adjustments for assumptions in the basic payment rates, not the outlier analysis, that changed how much the threshold had been set in 2009. The threshold in 2009 was lowered, but it wasn't because the agency applied a proper adjustment factor that year. So the agency could conceivably have paid out more in 2009 with a proper adjustment factor, but that would not have meant the adjustment factor was improper. They were attempting to model a specific component of the variables, which was the steady decrease in the charge ratios. And this is all obviously very complex. Joint Appendix 4927 through 28 reflects the agency's discussion of those other factors I'm talking about, where they say the regular payment rates are going to change, but that doesn't have anything to do with predicting what costs are going to be for purposes of outlier payments. Further questions from the bench? Thank you. We'll hear from the government. May it please the Court, Sidney Foster for the government. To address the jurisdictional question that Your Honor asked, and first actually just to address the question about 2011, I think that there is a decision at J130 to 136, which granted expedited judicial review with respect to Billings Clinic. I think it might have been on reconsideration, and it was with respect to fiscal year 2011. So just to be clear. So you agree. Every year there's somebody who's... We agree that every year is covered by... And if we heard their cases only, with the same counsel obviously, exactly the same argument, and decided in their favor, you would be bound on remand with respect to the ones who originally found no jurisdiction to make the same payments? I would think that, I mean, as a procedural matter, I would think then that the district court would remand those cases to the board to make a decision on whether to grant expedited judicial review. And I don't know whether at that point the board would make a decision saying that it's voucher. I think it would probably go back to district court, and then the district court would apply. Would remand to the board, and the board and the secretary has waived the regulation with respect to jurisdiction. So with respect to the secretary's jurisdiction, right? So the board would then have to decide what to do. Either it would follow what we just said, or it would have to go back to the district court, because, I don't know, it seems to me that by that point there would be no regulation. Of course, to be a little even more clear, there is no requirement of payment even possible here, right? The most that can be done is a vacating and then a remand to reconsider payments. No one's asked for a specific kind of payment. That's right, Your Honor. So everything's going to have to go back. And everything eventually goes back to the board. I mean, if this court were to rule in favor of the plans that are here. I know it's contrary, but still hypothetically if we were. No, I think that's exactly right. There's no procedural or timing or any other kind of barriers that would come in for those folks that were dismissed? I'm sorry? There's no procedural or timing or any other kind of barriers that would come in for the folks who came here on their dismissal? We haven't disputed that there are any other barriers that would apply to them. That's correct. But to answer the question that you posed, Your Honor, I mean, I guess my first answer would be that we don't think that there's any need to decide whether or not that is a basis on which this court could find that there is jurisdiction, because we think that there's clearly jurisdiction under the salty waiver theory that was applied in Queen of Angels. So I think that would be my first point. My second point is that I didn't have time to look at all of the board decisions that dismissed for lack of jurisdiction, but at least with respect to some of them and perhaps all of them, it not only dismissed for lack of jurisdiction, but it stated that it was denying expedited judicial review for that very reason, and I think that's at least at J100, 103, and 105. I didn't get a chance to make sure that that was accurate for all of them, and it may not be, but to the extent that the board decisions also stated that they denied expedited judicial review, then it feels like that argument would be an uphill one. That is an argument against jurisdiction would be an uphill one. The argument in favor of jurisdiction on the theory that you were advancing seems like it would be more difficult, because your theory was that there would be jurisdiction. But if they actually denied expedited determination, then that goes right away, right? Well, that's correct, and that's what happened here. We had a final decision because it was a jurisdictional dismissal. I would say I don't know if a denial of expedited judicial review gets to be challenged, you know, if it's on the merits if it gets to be challenged. I can't take a decision on that, but I can say that here, because it was a jurisdictional dismissal, this was our final final decision. I like your first point best, which we don't have to decide any of that. What's that? Which we at issue don't have to decide any of that. I know. It's always good. So that's my response on that. To address the merits, all of the aspects of HHS's methodology for setting the fixed loss thresholds in fiscal years 2008 through 2011 that plaintiffs challenged in this case were also used by the Secretary in setting the fixed loss threshold for fiscal year 2007, and that's what was at issue in this Court's recent decision in Banner Health. This Court upheld the agency's methodology, concluding that it was not arbitrary and capricious. But that was in light of the evidence to date, right? Exactly. That's correct. And I think that's almost word for word what the case said, something like that. That's correct, although I think there are numerous arguments that plaintiffs make here that are materially the same as the arguments that were made there. I think there are a few arguments that are new, and I think one of them is this notion, this argument that over time evidence accumulated that the methodology was improper. We think those arguments are without merit for the reasons that I think Your Honors were discussing. There just was limited information available to the agency, not enough information available to the agency in each of the rulemakings here to conclude that there was something terribly wrong with the methodology. Part of what we're looking at is if somebody makes the argument that there's new evidence that has come to light and therefore you ought to rethink what you're doing, say we're in the middle of 2008 or 2009 and the providers come together and say there's new evidence that's come to light, this formula is just not working, you need to reconsider it, then part of the inquiry is what did the agency say in response? Right. And is what they said in response something that indicates meaningful engagement and not an arbitrary and capricious just kind of refusal to even engage? Right. And your understanding of what the agency said in response when these arguments were made is What is your understanding? What's your understanding of what the agency's response is? We do think it was, given especially the broader context of these rulemakings, which are very complicated, it's just one small part of a larger set of rulemakings. We have a ton of comments being submitted, so we can't go into excruciating detail with respect to each one. We do think that the amount that the agency said in each rulemaking was appropriate. If you look at fiscal years 2008 and 2009, what the agency said was it described the comments that said, Hey, the past estimates seem to be resulting in payments that are below the 5.1% target. We think this other methodology should be used as one where you just assume that the future rate of change is going to be based on whatever the past rate of change was in this national case-weighted average cost-to-charge ratios. Please change. And the agency acknowledges those comments. It describes them in detail, and it says, No, we think that our methodology is more appropriate. We like our methodology because it takes into account certain critical cost inflation data that the plaintiff's methodology doesn't take into account. As this court's decision in Banner Health noted, that data is actually contemporaneous with the kind of period in question that we're actually focusing on with respect to updating the cost-to-charge ratios. And so the agency said, No, there's no need for us to make a change at this point. I think it's true that the agency didn't explicitly say, And by the way, we're not worried about the fact that there's one or possibly two years of data showing that we have come in under the 5.1% target when you look at the actual data, but I think it was blindingly obvious that one or two years' worth of kind of data showing that they were potentially undershooting the target doesn't compel a need for change. And I think that's particularly clear when you look at the data that was available in fiscal years 2008 and 2009 to the agency. If you look at fiscal year 2008, the agency did have some estimates about what had happened in fiscal year 2007, but those estimates weren't based on actual claims and charges from fiscal year 2007 because that information wasn't yet available to the agency. Those estimates were based on data from fiscal year 2006, and the agency tried to estimate what would have happened in fiscal year 2007 based on that data. So even if we're looking at just fiscal year 2008 alone, the estimates about what happened in fiscal year 2007 were not even based on fiscal year 2007 data, so that kind of underscores the conclusion that the agency... By 2010, the hospitals had come forward with information that there were cost ratios that could be used that were substantially closer to the mark. And given that the misfiring was not getting closer year by year, I get that you could miss it and hopefully it would improve over time, but it actually, other than the dispute about what happened with 2009, keeps getting worse, not better. It keeps getting more off the mark each year and has continued that again. 2010 and 2011 is worse than 2010. So when they come forward with a proposal for a methodology that, for further demonstration at least, was more accurate, is the Secretary obliged at that point to at least explain why they're not adopting a better methodology when this one seems to keep getting further and further off the mark? If not then, when? Yes, so I would disagree with the premise that they came forward with evidence that their methodology was superior. I think actually that's not correct at all. I think that, in fact, there's only one comment that they cite in the record that purports to try to compare their methodology to the methodology that HHS used, and it happens to have been submitted in the fiscal year 2010 rulemaking, and it's at J538 and 557 we cited in our brief. If you look at what that comment said when it was comparing the two methodologies, and just even assuming the argument that the methodology that the commenter used is an appropriate one for gauging which approach is better at predicting cost-to-charge ratios, which it may not be, but assuming our U.N. know that it is, the comment showed that the plaintiff's preferred methodology worked better for one of the cost-to-charge ratios that HHS predicts each year, or that changes in cost-to-charge ratios, adjustment factors, I should say, that HHS predicts each year, but then HHS's method worked better for the other one. So this comment certainly did not compel the conclusion for HHS that it needed to all of a sudden change its methodology to the one that plaintiffs make. And I think that's the critical point. I was looking at J573. What's that? J573, which is the Federal Register Notice 44009 in the third column near the bottom. And the commenters, again, it may be that there's problems with their methodology. That's for you all to comment on. They found a variance of 0.6% for their methodology versus the 1.6% for the CMS methodology. And I guess I didn't see the explanation. I'm sorry, where are you, Your Honor? So on J573, so the rightmost column, very bottom, right before response. The commenters also put 5 or 6 lines up before response. See, the commenters compared its method to the CMS method. They've got a much smaller variance. The explanation to me just sounded like we're going to keep doing what we're doing. Yeah, I don't think, maybe I'm misunderstanding the question, but I don't think that what they're doing right there is saying, let's take a look at these two methodologies for projecting an adjustment factor for cost-to-charge ratios and let's see, like, let's look back at, you know, at what each methodology would have predicted back in fiscal year 2008 and let's see which one actually matched what later turned out to happen. Right? That's what the comment that I was referring to on 538 and 557, that it was purporting to try and do that kind of comparison. And that's kind of, you know, if the methodology of that commenter was appropriate, that's a reasonable approach to take and a question to ask. I don't think that's what's going on here. If I'm understanding correctly, all that's going on here is that the commenters are saying, hey, it's kind of the same argument that we were talking about a moment ago. The estimates of actual payments are coming in below 5.1%. You should, therefore, change your methodology. And then in response, HHS says two things. It says, number one, no, we actually prefer our methodology because it takes account of this important cost inflation data. And then it also notes that, you know, in addition, you know, we have you are right that we, well, they don't say you are right, but they contradict the kind of assumption that HHS has been predicting payments to come in below 5.1% by noting that the prediction for fiscal year 2010, or the estimate for fiscal year 2009, I'm sorry, was that HHS made payments above 5.1%. And even the commenter upon which plaintiffs, you know, continually rely for their arguments, noted in the fiscal year 2010 rulemaking that also estimated that the fiscal year 2009 payments at that point were 5.3%, and that was at JA 541 and 567. I think, too, just to step back for a moment and address your comment about why is HHS so far off, you know, in these years and shouldn't that be it? Not getting better. Right. I mean, I think one thing to just keep in mind is, like, the difficulty of the task that HHS was faced with. It's faced with an unenviable task of predicting what all different kinds of Medicare payments are going to be in an upcoming fiscal year. That task is riddled with an enormous amount of uncertainty, given the lack of information that is available, obviously, about what's going to happen in the future, an enormous number of variables in play, including ones that aren't at issue in this case but that could affect the calculation. And so I think given that massive amount of uncertainty, it's unreasonable to expect that HHS's predictions are going to map on to the 5.1% or, you know, that the actual payments are going to be exactly 5.1% or even very, very close to 5.1%. It's just this is a calculation that's riddled with uncertainty. It's something very sympathetic and it's math that's way beyond my pay grade. But there's a statutory obligation here. That's what makes this much harder. It's a floor. Congress has mandated that it won't be less than 5%. And it has pretty persistently been less than 5%. They would say 100% of the time because they dispute 2009. But persistently you're missing that. And it may well be that you all chose 5.1, which gives you very little margin of error here to comply with your statutory obligation. But you chose that for probably very good reasons. But you chose to give yourselves virtually no margin of error to comply with the statutory mandate. And then you miss it persistently. And since Congress has commanded that floor, can we keep going with going, well, you know, we're trying to get it as right as we can and it's really hard. And it seems extremely hard. But maybe you all need to shoot for 5.5 and then you'll get it right. I guess I'd say two things in response here. Number one, I actually think the statute points in exactly the opposite direction. If Congress really wanted HHS to hit 5.1% exactly or some other percentage exactly, it could have required that outlier payments be exactly 5.1% of total payments. It could have required HHS. It doesn't say no less than 5% and no more than 6%? It says that the projected outlier payments should be between 5% and 6% of projected total payments. I thought it was mandatory language. Am I just misreading it? It's mandatory language. And I think the point I'm making, though, is that Congress said that the calculations should be based on projections. If it was concerned that actual payments should also be 5.1% of actual outlier payments, should also be 5.1% of actual total payments, then it could have said either A, said so, or B, said first make these projections and then go back. But then at some point this becomes evidence that maybe the projections just aren't reasonable in this regard and need to be adjusted. And it may be complicated, but however complicated it is, if it keeps persistently producing the wrong number and you're not going, whoa, okay, 2008, that wasn't quite what we want. Let's tweak this and then we'll be a little less off. Maybe we're off, but we're getting closer to the target. But you were going 4.7, 4.8, and then we have the fight about 2009, but then again it's 4.7, 4.8 going in the other direction. Do you not have some obligation given the statutory mandate to? Right. Well, again, the statute just says that the agency must, essentially the statute contemplates, I think Congress must have recognized that there was an enormous uncertainty in making these predictions. Can I pause over this for two seconds here? What's that? The higher you go above 5.1, that money gets taken out of prospective payments. Exactly. That hurts the hospitals. Exactly. So that's the reason that that number was chosen, because they're going to get hurt on the other end if you go too high. That's exactly right. So this isn't a cost-free decision. Exactly. Second, have they raised the argument you should have gone higher than 5.1? No, they have not, and so they wouldn't have raised that target. Presumably for that reason, because it's not really to their advantage. That's exactly right. And if Congress wanted to make sure that the actual payments were between 5% and 6% and realized that there was kind of this uncertainty, maybe Congress should have directed HHS to have a target of 5.5%, but that's not what Congress did. Congress directed target HHS to make projections that were between 5% and 6%. Thus, Congress was allowing HHS to choose a target that was as low as 5.1%. Congress anticipated surely that the actual payments would not necessarily fall within the course of that target. Just to make sure I'm understanding this right, the reason it's effectively a zero-sum game and that if you go above, then it ends up coming back out of the hospitals doesn't have to do with outlier payments. At that point, you're talking about non-outlier payments that get reduced. That's exactly right, because whatever value HHS chooses for this target, if it's 5.1% or 5.5%, then they have to adjust kind of the standard payment for typical cases. The inlier payments. Right, exactly, and that's actually addressed in the fiscal year 2007 rule, and I think we cite the relevant part in our brief. I think the other thing that's important to keep in mind here is that we can't just look at HHS's method and ask, is it perfect, are there ways that it could be improved? The question here is, were their methodological choices reasonable? Before we get to that, why isn't your answer, what we actually end up paying doesn't matter, as long as our projections, so why do we have to have that conversation at all? No, that is our answer, and that's exactly right. It doesn't matter about the methodologies. The answer that we were getting here is we're trying to get it right, and this is our way of doing it, and if the answer is we've got it right consistent with the statute. It sounded to me like your point was that we realize that we want to have a closer relationship between projection and actual. Do you feel advised to do that as well? Right. I mean, the central question here is, is HHS's methodology reasonable, and of course HHS's goal is to try and get as close to 5.1% as possible, but of course there's a massive amount of uncertainty. It's going to be very difficult given all the variables involved and the limited amount of information available to HHS for HHS to get super-duper close to that target, and so the only question really is were the methodological choices reasonable, and here I think they plainly were, especially when you have to think about when comparing to the methodology that plaintiffs say is better. I mean, the question is not really was that another methodology that would have been reasonable. Perhaps it would have been. It's not even was that another methodology that was better, but did the record here compel the conclusion to HHS that their methodology was so much better than the methodology that HHS chose that HHS should have abandoned its methodology, and I think there's no showing in this record based on kind of the past data about actual payments and where they came in relative to 5.1%, and also with respect to specific showings comparing how that methodology is actually working out. Not by this case, but just as a matter of background, has anything changed since? Has the formula, has the same methodology remained in place? HHS did change the methodology in the rulemaking for fiscal year 2014 and have adopted a methodology that's more similar to the one that plaintiffs advocate using. Of course, that fact is not relevant to kind of evaluating whether or not it was reasonable based on the records before the court here to continue adhering to it, and I think just based on these records, there just wasn't enough information for HHS to conclude that its methodology was unreasonable, and of course, this is a very technical area involving a lot of complexity where this court owes HHS's expert judgments a great deal of deference. Further questions from the bench? Thank you. I'm afraid we're 10 minutes over, but we'll give you one more minute. If you could stick to one minute, that would be helpful. Thank you. The key word in the statutory language which this court interpreted in the county of L.A., Los Angeles, is likely, and the Secretary's process is now likely to set thresholds hitting the 5.1% target when there is a key component of it that is demonstrably not working, and that showed up in 2008 in the very first case here at issue when the agency knew, and the Secretary hasn't even addressed it in her briefs, its method produced an increase when it was trying to model a decrease. We've explained in great detail why that happened, but their method was not structured to ever come up with a rate of change in the decrease in charge ratios They were trying to model what happened from the march of the rulemaking to the march of the following rulemaking. Their method could never do it. And Judge Rinabesan, you asked what do the agencies say in the rulemakings, and that's really key here because the agency must connect the record facts to its conclusions. If the agency, if we, just to be clear about the remedy here, couldn't possibly order them to give payments based on your model, right? You don't expect us to be able to do that. No, Your Honor, it would be remanded. So if they remanded and the agency said the reason is we didn't have enough years yet under this model, and we had in the first year we only had one year, in the second year we only had two years, and the third year it actually went over. If that explanation had been given then, would that have been enough? No, it wouldn't have been enough because we're looking at the 2008 rulemaking. What did the agency say? How did it connect the facts and the record before the agency to the conclusion that its method was more accurate than the commenter's method, which, by the way, was the identical method adopted in 2014. It was not kind of like it. It was identical. So, and in 2009, how did they connect their assertion, their conclusion, that their method was more accurate when commenters critiqued it to the record facts? That's not the test, though. It doesn't have to be more accurate than one or the other. It just has to be reasonable, right? But here the agency is asserting its method is more accurate than using the record track. It is not asserting our method is reasonable. We're not going to compare it against. The agency is actually asserting a comparison. Well, on our view, our only question is whether it's reasonable. Our question isn't whether there's another one that, if we were the policymaker, we would adopt. That's the opposite of arbitrary. I think there's two points. One is that the Secretary's method is absolutely not reasonable. When it's demonstrably not performing what it's supposed to do, which is to measure what the record trend is going to be. And is there a better method? Absolutely. The Secretary claimed its method, the agency's method, is better, but it didn't examine on the record any of the facts and still hasn't even examined in its brief how that assertion could be true and how its method could be accurate when it's modeling a decrease by increasing. Thank you, Your Honor. Thank you. We'll take the case under review.