Oral argument — American Hospital Association v. Thomas Price

and his official capacity as Secretary of Health and Human Services Appellant. Mrs. Salzman for the Appellant, Ms. Stetson for the Appellees. Thank you, Your Honor. May it please the Court, Josh Salzman on behalf of the Secretary. The record compiled following this Court's remand establishes three significant undisputed facts. First, there continues to be a significant mismatch between the resources appropriated for Medicare ALJ hearings and the demand for such hearings. Second, this mismatch is not a mere product of discretionary decisions with regards to the recovery product contractor program. If that program were eliminated in its entirety, the number of appeals filed each year would still be significantly greater than the agency's adjudicative capacity. And third, the agency has devoted significant efforts to backlog reduction, and as a result, the backlog today is smaller than it was the last time this case was before the Court. Now, all that said, it's also undisputed that providers are waiting far longer than 90 days for a hearing. That's unacceptable, and the agency still considers it to be a high priority to reduce and eliminate the backlog. But it does not follow that mandamus, and particularly the remedy ordered by the District Court, is an appropriate or indeed even productive response to this state of affairs. The District Court has ordered the agency to resolve some 350,000 appeals by the end of this calendar year alone. The agency has the resources to adjudicate perhaps a quarter of that amount. How is the agency expected to comply then? The only suggestion that has ever been made by anybody is through mass settlements. But perversely, the District Court's very order deters the kinds of merits-based settlements that the agency can enter into consistent with its statutory and fiduciary duties to the Medicare Trust Fund. This is true for several reasons. First, the very specter of a mandamus hangs over the process. A settlement requires a willing counterparty. And when there is an expectation on the part of the provider community that a judicially prescribed remedy will provide higher compensation than anything that the agency can in good faith agree to, interest in settlement actually declines. In fact, the agency reported in a recent District Court status report that precisely that has happened. I take that point. But could the agency basically do some triage and do the following? It appears that the agency has already taken steps to prioritize ALJ resources towards appeals by beneficiaries and other people who are financially vulnerable. Could the secretary also prioritize certain settlements and say, look, we're going to bargain with you in good faith. If we can't reach an agreement and we feel that we've acted in good faith, then you go to the end of the line and we move on to the next tranche of settlements. The agency certainly can work with different groups of providers to try to – I'm having a little trouble envisioning exactly what a tranche is for these purposes. Because once you start talking about settlements on the scale that at least the plaintiffs here have proposed, something like the entire hospital industry, what you have within a given tranche, as you put it, is a fairly diverse set of claims from providers, some of whom are much more scrupulous and have much higher success rates than others. So when you start trying to negotiate – I understand that. You can put them together in some whatever way that you feel like they are alike, based on whether they're all of the same party or they're all of the same classification, whatever. But basically you say to them, look, we're going to do our best to try to resolve these. And if we can't resolve, reach a resolution, then you go to the end of the line. Your appeal will be heard by an ALJ after everybody else's because we're prioritizing it now for this purpose, but we're not going to prioritize it for other purposes. Is there anything under the law or under your regulations that would prevent you from doing that? So the agency is engaging and has made available what is called settlement conference facilitation. So with individual specific providers, the agency is very interested in resolving all of the claims of a provider on a fair basis and not just dealing on a claim-by-claim basis. That said, in terms of sending somebody to the back of the line, the agency generally, with the exception of putting beneficiaries to the front, the agency has sort of felt that maintaining a first-in, first-out approach is what's fairest. But the agency does have the ability to prioritize. And while nobody has made a suggestion quite along the lines you're suggesting, I do think the regulations would allow the agency to... So the agency could do what I just suggested and could say, look, either settle or go to the end of the line, or you can escalate your claim and forfeit your right to a hearing. Or if you think that we're acting in bad faith, you can go back to the district court for some other relief, but those are your options. Well, I hesitate to say that I think there would be significant consequences if the agency certainly hasn't been given the opportunity, and against the backdrop of this mandating disorder, they certainly can't do anything you're describing here. If the agency was... Why can't they? Why can't they? Why can't they tomorrow say, this is how we're going to resolve this backlog? We've got to figure out some way under this order to reduce this backlog by 350,000 cases by the end of 2017, and this is what we're going to do. So I just want to be sure I understand. Basically, the agency... First of all, the scale of this is quite difficult because you are dealing with thousands of different providers here. So this isn't something where you can engage in unilateral negotiations with individual parties. It's time-consuming, and there's a limited amount of resources for this. But if the agency were to come to a provider and say, we will pay you $0.35 on the dollar, take it or leave it, and if you don't take it, you go to the back of the line, I would imagine that there would be significant resistance to that. And in fact, what would inevitably happen is if there isn't, they would say, you have to resolve this. We have all the leverage here. The notion that the agency has the leverage to make a take-it-or-leave-it offer under the circumstances, particularly against the backlog... I'm not saying take it or leave it. I'm saying that you're arguing that we have abuse of discretion review, right? That's correct, Your Honor. And what we've told the district court was to balance the equities. It seems to me that your argument is that the equities weren't appropriately balanced here because, A, we don't have the resources to eliminate the backlog, given what has been appropriated to us, but also, B, the only way to do so is through mass settlements, and we have basically been put in a position where we're going to have to settle these claims in a disadvantaged position. And I'm trying to test that argument, is how disadvantaged the position really is. And it doesn't necessarily mean that you have to take an unreasonable position and say take it or leave it, but it could be that maybe you have some leverage that you're not recognizing. I wish that were the case, Your Honor, but against the backdrop of mandate and disorder, the leverage really is exclusively with the providers. I also want to point out that it's not just about leverage. There are at least two other significant perverse results of trying to negotiate against the backdrop of this mandate and disorder. One is it perversely incentivizes more appeals. Right now, about half of the claims that are rejected at the second level of administrative review are abandoned. So you can see that at JA-149. So that's 150,000 claims that never enter the backlog. But if you create a system where the agency is required to settle any claim that reaches the third level, all it does is incentivize providers to file lots more appeals, which just compounds the problem. The other thing I would point out is this adverse selection issue, because I think it's quite significant. In order to be able to address this situation, if you want to come up with one-size-fits-all settlement terms for an entire group of providers, the inevitable result is those terms will be most attractive to the claimants with the weakest claims, but the claimants with stronger claims are going to view the offer as inadequate. You're not even going to be able to settle those claims, and all that's going to happen is the Medicare trust funds are going to be depleted by payouts for the weaker claims. So this isn't a very effective, ultimately, backlog reduction process, and it jeopardizes the Medicare trust funds by vastly increasing the potential liability that the trust funds are facing. Let me ask you this. You haven't proposed any alternative deadlines. You haven't said we can't clear the 350-some-thousand by December of this year, but we can clear so many. Have you in these status reports, or have you? I mean, I haven't seen anything that indicates that you're doing anything but saying we can't do this. So we've proposed a number. We've said mandamus shouldn't issue, but if mandamus should issue, we proposed at JA-146 and JA-152 to 153 a variety of alternative remedial measures that might be appropriate. We haven't committed to a particular deadline because we don't think we could do that in good faith. Given that there's a continuing mismatch between the resources for adjudication and the workload, about 80,000 appeals or 90,000 appeals can be decided each year, north of 150,000 are being filed every year going forward. So given this sort of structural mismatch, the agents can't in good faith say we think we can resolve this by X date. We know this is going to involve some work from Congress, or it's going to involve a great deal of accommodation from the provider community that we can't guarantee. We can't hold ourselves to that schedule. Now, if the entire provider community came in and said we're willing to work with you and settle at 30 cents on the dollar, then potentially, eventually we could work this. But in a representation to a court, we certainly couldn't commit to that. Likewise, while the agency is committed to working with Congress to come up with a solution for this problem, we certainly couldn't tie ourselves or commit to a specific date because we don't know the case of legislation. Let me ask you this, though. You just said if we could get all the providers to come in. You've got one provider of durable medical equipment that's responsible for 24% of all pending appeals. That's a quarter of the appeals. If you took the top ten providers and concentrated on them, and this particular one who has the 25% also has an error rate of only 25%, which means he has a success rate, I would think, just hear me out, of 75%. Offer him 75 cents on the dollar. Why can't you do that? That error rate actually is, and this isn't as artfully phrased in the declaration as it should be, what that error rate means is not error rate at the third level of review. It's not that they prevail 75% of the time in hearings. 25% of the claims they file as determined by a validation contractor. So just at the outside, a quarter of what they're putting into the system are claims that are rejected. So that projects out to an ALJ success rate that's far, far lower because many of those claims. We don't know that. This is what the agency informs me. But I would also stress that any settlement is going to require a willing counterparty. And the other thing that needs to be said, and I want to be clear, I'm not talking specifically about this one 24% provider or any other specific provider, but the other thing the agency needs to take into account are program integrity concerns and law enforcement concerns. Medicare fraud is a real phenomenon, and there are claimants out there where there are potential false claims act liability. There's potential issues related to people putting bad claims into the system, and the agency can't be in a position where it's paying out claimants where there are these major program integrity concerns. I've got one more question, and I think you alluded to it in response to Judge Wilkins' question. You said you put the beneficiaries at the front. Is this the system that's available not just to the providers but to also Medicare patients who are disputing whether Medicare has paid the 80% or whatever? It is, Your Honor. That's a smaller fraction of the workload. I think it's somewhere between 5% and 10%, but it's on that scale. The agency from the get-go recognized that these were the most vulnerable population, so they prioritized them. Actually, at this point, there is also now a consent decree out of a district court in Connecticut, so that is no longer even a discretionary decision by the agency. The agency is now required under the terms of that. To put the beneficiaries first? To put the beneficiaries first. Okay. All right. No further questions? Thank you. Good morning, Your Honors. May it please the Court. I'm Kate Stetson, representing the appellees. Judge Wilkins, I want to start with the colloquies that you had with government counsel because I think they frame the two pairs of questions that are at issue here. The first pair is the sort of broader questions. Can mandamus issue, which is what the first D.C. Circuit panel decided? And then the second question is should mandamus issue, which is what the district court decided, and that's what gives rise to the abuse of discretion standard of review. That's the first pair. The second pair is the colloquy that you were having at first, which is the difference between can't and won't. What the government is saying at this stage in May of 2017 is that it can't meet the four-year plan that the district court set forth in its order. I would suggest that one place to start with this is to go back, in fact, to the first D.C. Circuit decision. A lot of this discussion also took place in the context of whether mandamus could issue. But the problem I'm having is I don't understand where the four-year deadlines came from. I mean, we can't just pick numbers out of thin air. I mean, I didn't see anything in the briefing or in the appendix that explained how those numbers were arrived at. Right. One of the issues, Judge Wilkins, is the question that Judge Henderson posed to government counsel, which is that government counsel didn't essentially offer any other alternative timeline. So that's the first issue. The second response is the district court concluded upon reviewing the suggested timeframe. And remember, we essentially gave Judge Boasberg two different paths that he might follow. Path number one is order the secretary to take certain actions. But what the district court said in his decision was, I am reluctant to intrude on the discretionary processes of the secretary. That leads me to the second path, which is these phased reductions in the backlog. And what the district court found was that those phased reductions were thoughtful and reasonable. Did the district court ever find that the government can meet any of those deadlines? Because I can't find the district court having found that. I think the short answer is yes. If you look at the last page of the district court's decision and again at the final page. The last page of which decision? Of his December 6th opinion. One of the comments that he makes. Tell me the JA cite that well. JA 165. I'm sorry. So the third to last page. Defendant nonetheless argues that imposing such a timetable would require her to make payment on Medicare claims regardless of the merit of those claims. And that is the crux of what the government is asking here. I'm looking for the finding by the court that the agency can actually reach these timetables. That the agency can reach those timetables? Yes. That's not impossible. I think it's in the very next sentence. The timetable doesn't require the agency to do something that it is statutorily prohibited from doing. All it demands. Yes. All it demands is that the secretary figure out how to substantiate the claims. Right. And the district court and the secretary says it can't. And the court didn't say I disagree with the secretary. The court didn't make a fine. Let's put this as a hypothetical. All right. Imagine there are 10 billion claims. And Congress provides $1. But Congress also says you must resolve every single claim in a week. Now, you would agree under those circumstances that whether or not mandamus is jurisdictionally available, that it would not be granted by a court. Do you agree with that or not? In those circumstances, I would agree. Because I think under Alabama Power v. Kossel, under Sierra Club v. EPA, under NRDC v. Train, the agency would be in a position to bear what those cases call the heavy burden of demonstrating impossibility. So, Judge Garland, the fact that Judge Boasberg didn't specifically say I find that you can do this, he found that the agency had not established that it couldn't. No, I don't. Which for these purposes is sufficient. I have to say he didn't find that either. This strikes me as I think the district court is doing the best that the court can, given that we gave an instruction that says on the one hand and on the other hand, which is all we provided. But I don't see a finding that it's not, whatever the burden is, whatever the burden on the government may be, Train requires that a district court not order an agency to do something that's not possible. And that requires a finding, it seems to me, by the district court that what the government says is impossible is actually possible. And I don't see that. I think what the district court did is laudable, given that the district court did not want to compel particular things. But when you set a timetable, I mean, the timetable we're working off now is 30% reduction by the end of this year. Government's position is that it's completely impossible. And the district court didn't say why it's not impossible. The district court just says figure out a way to do it. Now, that may be the pressure that we put on the district court by the kind of opinion that was issued, but we certainly didn't direct that. And if you agree that the court can't compel something impossible, why isn't there a requirement that there be a hearing at which a determination is made of whether each stage of the timetable is possible? Because that hearing would have been premature in December of 2016. What was in front of the judge in December of 2016 was the question whether the agency should be mandamus. He answered that question, of course, in the affirmative. If the agency in the beginning of 2018 comes into the district court, presumably on our motion to enforce the mandamus, comes into the district court with proof, not just supposition, not just prediction or speculation, but actual proof that it could not satisfy that benchmark, then we would be having a very different conversation. Well, what conversation would we be having? See, they're under threat of contempt here. It's not an automatic contempt. You asked for that, and the district court refused that. But they're still under threat of contempt. That's a pretty big threat at this stage, unless there is some possibility even that they can achieve it. And I understand all the different points you make about different ways that might be possible, but the district court never said that any of those was actually possible. And to put the agency under the threat of contempt under those circumstances both seems problematic but also raises the unintended consequences point that counsel raises, that putting the agency under the hammer here is going to increase the number of appeals rather than decrease them. But why wouldn't the imposition of the status reports answer that in some respect? Because I see Judge Boasberg as saying, look, this is what I'm proposing. I want these status reports. And in those status reports, I don't think HHS can just wait until the November status report and say, guess what, we're not even close to making it. That's right. What he had in mind was that as that deadline gets closer and closer, if the status reports not only say we're not going to make it but have some alternative, we can do this, but we can't do that. We can get rid of this many, not that many. I mean, I think that's right, Judge Henderson. And I think that points to one of the issues that both the panel and the district court have, which is we are talking about predictions. And we heard in government counsel's response, I think, to each of your questions the accommodation that there may be ways. You mentioned the top ten. You mentioned triage, back of the line, and so forth. Government counsel said at one point that that hadn't been proposed before. This is exactly the point of an injunction. What the NRDC versus train court said is an injunction is supposed to be an adrenaline shot. It is supposed to cause the agency to move more quickly, to think of the things it didn't before, and to solve the problems in front of it. So for the district court to impose on the district court in 2016 the obligation to resolve at that time whether the agency could or could not by 2020 comply with a mandate that he declined to put any specifics on, correctly, we think, that to me is imposing too much of a burden on the district judge. What would be our standard of review? The district court at the end of this year hasn't met the 30% deadline. The district court says, okay, you're in contempt. Is our standard of review the same as it is today? And this is where we just decide abuse of discretion. I think the overall question presumably would be abuse of discretion review. But in order for the judge, if I'm sort of extrapolating correctly from NRDC versus train and CASEL, what the judge would have to find is actual factual findings presumably backed by not just speculation from the government about an inability to comply with a settlement, but data showing, for example, that hundreds of thousands of these claims are potentially fraudulent, data showing that they are not susceptible to mass settlement. There is no such data in this record. What you find in the government's brief are these kind of broad, somewhat hedgy statements. But if that's true, the district court could have said that. The district court didn't say that. The district court just said, I'm not going to tell you how to do this, but do it. But the district court didn't say, disagree with respect to the data or anything else. Yes, Chief Judge Garland, that is what the district court said. But I am not familiar with a single D.C. Circuit case that suggests that Mingdamas is inappropriate because the agency predicts that four years hence it won't be able to comply with something. In fact, what you see over and over again in that trio of cases that I've mentioned is that the court says, we understand that you are speculating about impossibility. But the burden is especially heavy on you in that circumstance to show that something is actually, factually impossible, not just to posit that this is going to be very difficult. That's why each of those courts said, come back to us or go back to the district court in the first instance. But here's my problem, is that I think it's undisputed that the ALJs just have a physical capacity of addressing what is about 90,000 appeals a year. That's right. We've got a backlog of close to 700,000. Even if nobody ever filed another appeal, then that backlog would take, under what is physically possible for ALJs to do, about seven years or eight years to clear. So the only way that this is physically possible is through some measure of mass settlements. In the specter of contempt, and when we're balancing the equities here, an agency now has a gun to its head, but no one has a gun to your client's head. And you make arguments in your brief about how there's an incentive for the RACs to want to claw back money because they get paid on a contingent basis and they're not penalized when they lose appeals. Right. Well, your clients have the same incentives. They're appealing everything, and they're not settling anything. And the declaration from the chief ALJ says that. And the most recent March 2017 status report shows that the recent efforts at settlements aren't working. So it seems like basically the equities aren't in favor of the agency, and it's just being set up to fail here. Judge Wilkins, I'm going to work backwards through your question. I first need to take issue with the statement in the status report that settlements aren't working. That was a speculation on the part of the declarant. In fact, there hasn't been a credible settlement offer since the mandamus issued. Working backwards still, of course there is an incentive to settle, and that's in two respects. The first is we are talking about hospitals, including one of these plaintiffs, Baxter, who didn't have enough money to fix its roof. If Baxter is confronting a situation where it is considering taking, for example, 68 cents on the dollar, which was the last settlement for, by the way, 380,000 claims, is it going to take 68 cents on the dollar now, or is it going to play out the string for the next four years on the theory that maybe at the end of those four years the district court might issue a default judgment entitling it to a little bit more? That's the second issue. The third is remember that with these phased targets, the agency doesn't need to peel off everybody at once. It needs to peel off 30% at a time, which puts the hospitals that are contemplating whether to settle in an interesting position because if a hospital decides that it's going to hold out and wait for a better offer, but 31% of hospitals get in front of it, it has forfeited that opportunity. So I need to push back strongly on the issue that there is some kind of a perverse incentive for all of these hospitals simply to wait until the end of this laborious process, and somehow they will magically then get 100 cents on the dollar. I don't think that's the reality. I also need, if I could just make two more quick points. Mr. Saltzman mentioned in the beginning of his argument that the backlog today is smaller than it was. That is also something that he said to the panel last week in the Casa Colina case, but he then went on to say, and this is about 17 minutes in, I do not want to delude Your Honors, it's going to get worse. So even the government understands that the backlog is going to get worse. I'm not sure which way that cuts, though. I mean, that cuts in favor of impossibility. On the other hand, it could cut in favor of they're not faithfully going forward, and we're not in a position to make factual findings on that, and I'm afraid that the district court didn't. Maybe the district court's not in a position to do it either, but when you have a Congress saying you must do something by a certain date, and then you have Congress clearly not providing enough money for the kind of hearings that everybody originally contemplated, at least, I mean, this is not Congress versus the courts. This is not Congress versus the agency. This is Congress versus Congress, and Congress can decide amongst itself. But Congress, although we've talked about Congress and the RACs, we've talked about Congress and the budget, the one thing that hasn't come up is Congress and the statute that pertains to consent settlements between HHS and the providers. So this is 13, 42 U.S.C. 1395, triple D, F5. And this is the statute that says that the HHS can compromise claims with providers and it can do so using less than a statistically valid sample of claims. So there is no clearer indication in the statute itself that the agency is in a position to settle these claims if it wants to. If the agency, at the end of these phased processes, comes in and demonstrates to the satisfaction of the district judge that it was actually, factually impossible to comply, then the agency presumably will seek and obtain a modification of this mandamus. But right now, all it's doing is saying that it predicts it can't do something. The answer is it has to try. So if the agency engages in reasonable settlement offers negotiations and they break off with parties, you would concede then that with respect to however many of those appeals where that happens. So let's suppose they engage in reasonable negotiations that cover 400,000 claims and they can't settle them, but the district court finds as a matter of fact that their last and best offer was a reasonable offer. Then you would concede that they should get basically credit for having resolved those 400,000 claims and they can go towards counting towards eliminating the backlog. I think if the agency is able to offer that proof and if the district court accepts that proof, that's absolutely right. If the agency reaches out to 400,000 hospitals and says we're going to offer you 68 cents on the dollar and the hospitals say no thanks, we're holding out for 99.5% on the dollar, and the agency comes in and submits an affidavit saying this is the offer on the table, this is what we were attempting to do, this is the set of circumstances we were targeting, and this is the response that came back to us, then the agency might be able to make a good case to the district court to modify the maintenance. But all of this conversation that we're having now in May of 2017 runs the risk of doing what the NRDC versus train court warned of, which is presupposing impossibility, speculating about it rather than asking the agency to do its part in good faith to try to settle these claims, which it plainly has both the statutory and the regulatory processes to do. But didn't the agency make a big settlement offer at the end of 2006 that was rejected, that was expected to reduce the spec log and that was rejected? No, I don't think it was rejected. I think that it was not accepted. What the agency would say, I think, is that it wasn't accepted in the numbers that it expected. In part that's because, as I understand it, a lot of the hospitals that were subject to that offer actually had taken the prior deal, had taken the 66 cents on the dollar deal that had been offered that took care of 380,000 claims. The last point I would like to make is one just about this statutory impossibility. One of the government's primary arguments here is that it cannot pay claims, that the statute instructs it not to pay. There are two responses. The first is the statute also clearly allows for compromise of those claims. And the second is, again, in Mr. Saltzman's argument of last week, one of the points that he made to the panel was most of these appeals have to do with documentation requirements. Documentation requirements are not a statutory requirement. They are in the regulations and they are in the policies and they can be altered. The agency needs to try. I agree. There's not a lot in the district court's decision, granting mandamus, about how escalation and the possibility of escalation should weigh in with the balance of the equities. Because, as I understand your argument, you say, well, escalation prejudices the claimants because A, we can't supplement the record, B, we can't get a hearing, and C, we don't get de novo review at that stage. And if we escalate, the review before the board would be, I guess, without a hearing, without supplementing the record, and it would be, I guess, a deferential review instead of a de novo. And deferential to the agency on any subsequent judicial review without a record. That's right. But, you know, it's not nothing. And if I'm balancing equities and if one of your complaints is, well, we can't supplement the record, well, you're the claimant and you should know what you're supposed to put in. And if you, like, don't have an opportunity to put more in to supplement your claim, it's kind of the brakes. And with respect to lack of a hearing, lots of things are decided without hearings. And so the difference in the standard of review, that's something also, but it's not like this isn't an empty remedy here, but it doesn't seem to have been accounted for by the district court at all in the decision below. Right. Well, the first thing is that the standard of review, of course, is the kicker. If you're up on judicial review, if you've needed to escalate all the way up to get a judge to hear your case and you're on abuse of discretion review or deference to the agency, I should say, that puts the escalator appellant in a particularly poor position. But more to the point, this argument about escalation, of course, was the core argument that the government proffered in the first AHA appeal, where the government said there's an adequate alternative remedy. And if you look back at this court's first opinion, one of the things it says is, in any particular unusual case, which is what escalation is designed for, maybe it's a good remedy. But where you have an industry plaintiff like the American Hospital Association bringing a systemic claim, the answer cannot be you all can escalate all of your appeals. But we also have in the record a declaration from the chief ALJ that there are large companies out here that deem it good business practice to appeal every single rejected claim. That's a pretty significant and, I would say, damning fact that's in the record. Judge Wilkins, I think what I would say to that is... So if they're appealing every claim, then perhaps the equities say that the ones that are most important to them, they try to settle. And if they can't settle them, then maybe the equities say that they escalate and take their chances. But that the... It's not like the agency has created the circumstance completely. There is some ability of your clients to control the inflow here. And without there being some disincentives for inflow, especially inflow of bogus claims, then I don't see how we're appropriately balancing the equities here. Right. So with respect to bogus claims, everyone agrees, including us, that 42 CFR 401.601F essentially strips out from any need for compromise claims that indicate that they're fraudulent. So a truly bogus claim essentially is in a category by itself. But how is that to be figured out under the pressure of getting 30 percent of the claims done by the end of this year? I mean, all the proposals that you've suggested have to do with statistical settlement or something like that. But none of those actually determine whether any particular claim is bogus, to use a word of legal art, apparently. I didn't realize it was, but okay. So how is this consistent with that CFR? Basically, they have to pay in order... There's no way they're going to be able to pay these settlements or make these settlements without not actually looking closely at specific claims. Isn't that right? I took Judge Wilkinson's hypothetical to be if you've got a supplier or a provider that is essentially flooding the system with bogus claims. I understand that, but leave that aside. If you read that CFR as saying they can't settle a claim that on the merits they've evaluated is bogus, doesn't that require them to be able to evaluate claims on the merits? No, I don't think it does. I think what it does is it permits them, if there is an indicator of fraud on the face of a series of claims, to set it aside. But again, we're talking about the need to settle just 30 percent of these claims. So unless the agency is going to commit to the court... This year, just one year. That's correct, but 30 percent in one year. If the agency is going to say to this court that 100 percent, essentially, of the claims that are pending right now are all bogus, surely there are 30 percent of claims that are facially appropriate, that may suffer, as Mr. Saltzman said last week, from documentation problems, but that can be resolved by settlement. So the idea that we are going to get into a situation where there's this perverse incentive, I think, is not true. The other thing I would say is the Fair Fund amicus brief does a good job of pointing out that where you're talking about this adverse incentives question, there are statistics, of course, that will aid both the agency and the company in its good business practices in deciding how much on the dollar to offer that company to make its claims go away. If the company has, for example, a significant error rate and it is significantly getting pared back in the appeals process, that company presumably would be willing to accept something less on the dollar and to accept it sooner than four years from now. Can I ask you, this statutory section that you're talking about, consent settlement, do you think that Congress intended that the consequence of the inability to meet the deadlines is that you have to go into this kind of settlement procedure that you're talking about? I mean, that's what we're saying. The consequence of the failure is to look at, is to do these kind of broad settlements. Well, that is what we've been reduced to saying because the agency hasn't come forward with a solution. And remember, we're now talking about a case that's over three years old. And one more sort of finer point on that, Chief Judge Garland, you mentioned a few minutes ago that we had suggested X, Y, or Z. It's not our burden to suggest or establish the routes that the agency can take to cure this. But setting all that aside. But there has to be possible. It's not enough to just, as the judge said in his early first opinion before actually issuing the order in this case, that he doesn't have the power to wave a magic wand and say it's going to be done. So unless there is a route that's possible, it's a little hard to uphold an order that just sets deadlines. But the reason I asked you about the consequence is really the same question that Judge Wilkins was asking you. The previous panel held that, not that the escalation provision of the statute doesn't bar mandamus. That is, it doesn't prevent jurisdiction for mandamus. But it did not say anything about whether the escalation provision should be balanced or how it should be balanced. And I ask this question because it is entitled, Consequences of Failure to Meet Deadlines. And Congress said that the consequence of the failure was escalation above the ALJ. And when I look at it, it didn't say the consequence is consent settlement or anything else. Right. And so even if that doesn't affect the jurisdiction of mandamus, don't you think it should be a heavy consideration in any court's decision about whether to issue mandamus or what the content of the mandamus order should be? I believe it was a consideration, which means that your review of Judge Boasberg's decision, that that was not sufficient to count against mandamus, of course, is subject to abuse of discretion review. So unless the panel is prepared to conclude that the district court abused its discretion in concluding that mass escalation, which, after all, would just kind of move this bolus of claims all the way up to the district court, if you could meet the amount and controversy requirement, and with the deferential standard of review that comes along with it, the district court clearly concluded that that was not sufficient to count against mandamus. The district court at this stage has simply ordered the agency to try to come into compliance 30%, 60%, 90%, 100% over the next four years. If the agency, again, wishes to come back with actual proof, not just speculation but proof, then the district court at that stage can assess that proof and decide whether it's going to modify the injunction. But this is not the time to have that conversation, because all the agency is telling you right now is that it thinks it can't do that. The agency should be instructed to try. That's the point of mandamus. Fair enough. Further questions? Take your times up. We'll give you another two minutes. Thank you, Your Honor. There's just one point I want to make, and it's really picking up right where Ms. Stetson left off, which is the entire premise of her argument today is that the agency needs to start trying to fix this problem. And I think what that overlooks is the agency has been struggling against this backlog since well before the mandamus order came down in December of 2016. The reason I bring up the decline in the backlog today as opposed to at the time of the last appeal isn't because I want to give a false sense that everything is fine or everything is fixed. It's to show that the agency is struggling against the backlog, that the agency has not adopted what Barr Labs called a position of utter indifference at this timetable. There are significant constraints on the agency, but the agency, as is detailed from JA 95 to 105, has adopted a number of initiatives. Settlement conferences, the use of attorney adjudicators to provide on-the-record review in cases where there can't be a hearing, statistical sampling, a wide array of initiatives. Okay, the list goes on. Yes. If the court were, if the district court, say we said that deadlines themselves are not appropriate, but we also didn't think status reports were appropriate, what are the other kinds of mandamus requirements that could be considered by the district court? Could the district court order, for example, I appreciate that the court did not want to do this, but that court order that the agency eliminate RAC to the extent of its discretion because it doesn't have discretion to eliminate the deadlines, the statutory deadlines. So the RAC program is itself statutorily mandated. Now the agency could ratchet it back to the extent of its discretion, whatever that is. So I think that would be problematic, Your Honor, because I do think the agency needs some degree of flexibility to balance competing priorities. So if, for example, and again, this is purely hypothetical. I thought your position as reflected in your brief was that even if RAC were completely eliminated, it would still be impossible. Isn't that your position? That is my position, Your Honor. And in light of that fact, and it is precisely because that's true, that I think, for example, if eliminating RAC would lead to a marginal improvement in the backlog, let's say it was 5 percent better, but it would expose the Medicare program to billions of dollars in additional fraud, I think Bar Labs and the Fourth Circuit's Cumberland County decision and even this court's decision in the prior appeal, which says the agency has some degree of flexibility to balance priorities, would suggest that the agency is best positioned to do that. That being said, and we do think status reports here would be sufficient to ensure that the agency doesn't adopt a position of defeatism to make sure the agency continues to work on this problem. But if this court were to want more, at JA-146, for example, what the agency proposed were certain scalebacks of the RAC program, reductions in look-back periods and various things that we think would be unadvisable, we think would strip the RAC program of too much efficacy without sufficient benefit to the backlog. But those are changes that we could live with. So that is something that we put forward as an affirmative possibility. I do have one question about the RAC. Are we going to see another bump in the backlog through renegotiation of contracts? I mean, can't you do something, or maybe you already have, to extend the contracts? Because as I understand it, that was part of the cause of the huge multi-hundred-thousand-dollar increase after a delay or a pause of a couple years. So here's the timetable on that, Your Honor. So in 2013 to 2014, the RAC program was a significant driver of the backlog. More than half of appeals originated through the RAC program. The last two years have seen a drastic reduction, down to less than 10%. Now, the drastic reduction is due to two factors. One is the agency implemented a number of permanent changes to the RAC program. The other is that while contracts were being renegotiated, there was a slowdown in RAC activity. Now, in October 2016, new contracts were finalized. So there is going to be more RAC activity. But those new contracts involve several new provisions that are designed precisely to prevent what happened in the past. So one thing from the start of this litigation that plaintiffs were asking for was that RACs be given stronger incentives to be accurate. And in fact, these new RAC contracts, as you'll see at JA-144, have precisely provisions that provide economic incentives for accuracy and which will require RACs to maintain certain accuracy rates. So the agency does not at all expect a repeat of 2013 to 2014. Other questions? We'll take the matter under submission. Thank you. Thank you.