[GENERAL FRANCISCO] Mr. Chief Justice, and may it please the Court: In March 2018, Secretary Ross reinstated a citizenship question that has been asked as part of the census in one form or another for nearly 200 years. The district court's invalidation... [JUSTICE SOTOMAYOR] I'm sorry. [GENERAL FRANCISCO] ... of that decision was wrong for... [JUSTICE SOTOMAYOR] I'm sorry, it's not been a part of the survey, which is where he reinstated it, since 1950. And for 65 years, every Secretary of the Department of Commerce, every statistician, including this Secretary's statistician, recommended against adding the question. So it may be that 200 years of asking a citizenship question in other forms may be true, but not on the short survey. That's what's at issue here. [GENERAL FRANCISCO] Well, but, Your Honor, it has been part of the census for the better part of 200 years, initially as part of the overall census itself... [JUSTICE SOTOMAYOR] But don't we put... [GENERAL FRANCISCO] ... that went to all individuals. [JUSTICE SOTOMAYOR] ... don't we ask the question in context? And for 65-odd-plus years, everybody said don't add it? [GENERAL FRANCISCO] No, Your Honor, because, in fact, it was included in the long-form census until the year 2000. [JUSTICE SOTOMAYOR] That is a different... that's the whole issue. It could have been... it's still included... or it could have been included in the ACS form, which is still being done. And that's being done every first, third, and fifth year. [GENERAL FRANCISCO] Yes, Your Honor, but it has been part of the census in one form or another for a very long period of time. It has a long pedigree. [JUSTICE SOTOMAYOR] So the question now before us is... [GENERAL FRANCISCO] The question, therefore, is two things. First, we think that the Respondents' claims are not justiciable because their injuries aren't fairly attributable to the government or subject to APA review. And, secondly, the Secretary acted well within his discretion when he determined that reinstating the citizenship question would provide the best evidence of citizenship. [JUSTICE GINSBURG] Can we just go back a bit to your opening? Why was the citizenship question dropped in 1960 and remained off for all the decades after that? What was the reason for dropping it? [GENERAL FRANCISCO] Your Honor, in 1960, it didn't appear in anything, and it was moved on to the American Community Survey. [JUSTICE GINSBURG] But... but why... [GENERAL FRANCISCO] That was part of an overall movement of most of the demographic... I'm sorry, onto the long-form census, not the American Community Survey. And that was part of a larger process that moved a large number of demographic questions off of the short form and onto the long form. We no longer have a long form, so then the question is do you reinstate the long form or do you, in fact, move it back onto the short-form census. [JUSTICE GINSBURG] But didn't... didn't... [GENERAL FRANCISCO] It was eminently... [JUSTICE GINSBURG] ... didn't the Census Bureau give a reason why it was dropped? [GENERAL FRANCISCO] Well, they generally wanted to move all of the demographic questions onto the long form. We no longer have a long form. And the problems with using the American Community Survey are well known. So you basically... [JUSTICE GINSBURG] But there was nothing... there was nothing in 1960 to the effect that the Census Bureau found that putting it on the short form would depress the count of non-citizens? Nothing like that? [GENERAL FRANCISCO] Well... well, sure, Your Honor, but that's because they thought that, along with all of the other demographic questions in the census, had an overall impact of... on... on... on overall census accuracy. And that underscores why we don't think this is really subject to judicial review, because, really, what you're saying is that Congress... courts would have to review every question on the long form to determine if the informational value of the question outweighed the impact on census accuracy, because, at the end of the day, if you add any particular question onto the census, you're always trading off information and accuracy. And 141(a) doesn't provide courts with a basis for evaluating that determination. So that's why we think this isn't subject to APA review at all. But we also think that the Respondents don't have standing here, because they're injured if and only if, first, you have third-party action; secondly, you have third-party action that's illegal; and, third, that third-party action is based on speculation that the government will itself violate the law. [CHIEF JUSTICE ROBERTS] On the... on the illegality, is... is that a... a predictive factor? In other words, do we, as our cases have often said, do not assume illegal behavior in establishing standing, but is that simply predictive? In other words, we doubt people are going to engage on a regular basis in illegal behavior, and, therefore, we don't think their injury is... is tangible or likely, or is there something special about the fact that it's illegal activity? [GENERAL FRANCISCO] Well, I think, Your Honor, in... in the Court's past cases, it has often been used as a predictive factor, but I also think that when you put it all together, it breaks the chain of causation for Article III causation purposes. [CHIEF JUSTICE ROBERTS] I mean, it is true that if people go 60 miles an hour in a 55-mile-an-hour zone, that's unlawful. But you wouldn't say that they're not going to do that in forming public policy... [GENERAL FRANCISCO] And... [CHIEF JUSTICE ROBERTS] ... or considering standing. [GENERAL FRANCISCO] ... I... I agree, Your Honor, and that's not our argument here. I think that Clapper is a good analogy. In Clapper, it was quite clear that the plaintiffs in that case suffered an injury in fact, because they declined to use their cell phones to communicate with their clients and their clients declined to use their cell phones to communicate with them out of a fear that their phone calls would be intercepted. No question that there was an injury in fact. But what the Court held was that that injury wasn't fairly attributable to the government because it was caused by the plaintiffs' fear... [JUSTICE SOTOMAYOR] I... I... I'm sorry, you're talking... [GENERAL FRANCISCO] ... that the government would intercept their calls. [JUSTICE SOTOMAYOR] You're talking proximate cause, which we've never used. [GENERAL FRANCISCO] No, Your Honor. [JUSTICE SOTOMAYOR] We've used determinate or coercive effect. In Wayfair, one of the reasons we found the stores, not the states, to be injured is because consumers fail to pay taxes, an illegality under the law. And we said that's why the states were being harmed. In NAACP versus Alabama, we held the NAACP had standing, even though it was their members who would be injured by other people, an illegality, harassment, but as the just... as Chief Justice Roberts said, it's predictable. There's no doubt that people will respond less... [GENERAL FRANCISCO] Uh-huh. [JUSTICE SOTOMAYOR] ... because of the census. That has been proven in study after study. One census surveyor described an incident where he walked into a home, started asking citizenship, and the person stopped and left his home, leaving the census surveyor sitting there. So, if you're talking about prediction, this is about 100 percent that people will answer less, so... but I don't know that it's prediction. It's an action by the government will be a... will be a cause of this, not proximate necessarily, but that cause will cause harm. [GENERAL FRANCISCO] Right, two responses, Your Honor. First, we're not talking about proximate cause. We're talking about an analysis that was similar to what this Court used in the Clapper case, where the Clappers said that even though there was injury in fact, even though the government's actions were in a very real sense a but-for cause of that injury in fact, it wasn't fair to attribute that to the government because it was based on the plaintiffs' speculation that the government would intercept their telephone calls. [JUSTICE SOTOMAYOR] But this is not... [CHIEF JUSTICE ROBERTS] You... [JUSTICE SOTOMAYOR] ... the plaintiff acting; this is third parties acting. [GENERAL FRANCISCO] Yes, and that makes it, we think, even worse because the Court's cases have generally said you don't rely on third-party standing. But... [CHIEF JUSTICE ROBERTS] You... you said you had two responses? [GENERAL FRANCISCO] Yeah. The second is that on NAACP against Alabama, that was a case where the NAACP was being directly regulated by the statute that they were challenging. They were forced to disclose their private membership lists, and that was their injury. There was nobody that stood in between them and the disclosure of their private injury... private membership list. But turning to the agency reviewability argument, there really is nothing in 141(a) that provides courts with a basis to review this decision. The language is quite similar to what the Court addressed in Webster. Every... the addition of any particular demographic question is always going to be a tradeoff between information and accuracy. And I'd... I'd... I'd urge you to look to the 2000 long-form census that had highly detailed questions about not just citizenship but things like your commuting time, how many bedrooms you had in your house, whether you suffered from certain health conditions. Under Respondents' position, courts would have to review each one of those questions to determine whether the informational value of that question outweighed any potential impact... [JUSTICE BREYER] On the main form... [GENERAL FRANCISCO] ... on census accuracy. [JUSTICE BREYER] ... on the main form... on the main form, suppose the Secretary puts in a question about sexual orientation. Suppose he puts a question in about arrest record. Suppose he says, I'm going to have the whole survey in French. [GENERAL FRANCISCO] Uh-huh. [JUSTICE BREYER] In other words, we have no role to play no matter how extreme? [GENERAL FRANCISCO] Your Honor, you certainly do have a role to play, and I think... [JUSTICE BREYER] All right. Well, then that's the question. [GENERAL FRANCISCO] ... and I think your examples... [JUSTICE BREYER] That's the question. [GENERAL FRANCISCO] Yeah. [JUSTICE BREYER] That's the question. And in this case, there's a statute, and the statute says that the Secretary... at least on this form, the main form, he shall use administrative records, unless... it says to the maximum extent possible. Don't ask direct questions. Use administrative records, because they want to keep it short. [GENERAL FRANCISCO] Right. [JUSTICE BREYER] To the maximum extent possible. So I have two rather technical questions in what I think is the heart of this case. It's a technical case. All right. The first question is the Secretary, I gather from the record... and we've looked at it, my office, pretty carefully... is told by the Census Bureau in three studies that if you ask this question on the regular form, you will get back fewer answers. And they extrapolated to do that from the other surveys and so forth, and those extrapolations, you know, holding for everything constant, showed that the non-citizens often didn't say they were non-citizens. [GENERAL FRANCISCO] Uh-huh. [JUSTICE BREYER] And some didn't return it at all. Now I haven't seen any evidence to the contrary. So I'm asking you where the evidence is on that. [GENERAL FRANCISCO] Sure. [JUSTICE BREYER] And as to the second, same question, the second is that several surveys, including Dr. Abowd, told the Secretary: Mr. Secretary, if you add the question to the census, the short form, the direct form, you will discover that even the information you want about citizenship is worse than if you just look at the administrative record. Now how can that be? [GENERAL FRANCISCO] Sure. [JUSTICE BREYER] Well, A, they say 13 million people, it'll be a wash because you won't get information either way. They won't return it and you don't have it over on the administrative part. But, as to 22 million, which you highlight in your brief... [GENERAL FRANCISCO] Yeah. [JUSTICE BREYER] ... what about them? And as to them, what Dr. Abowd says he says, and I saw it in the record, he says, as to those 22 million, I'll tell you what, you just look to the census returns and you're going to find it not that accurate because some are not going to tell the truth. [GENERAL FRANCISCO] Right. [JUSTICE BREYER] So go look at the administrative returns, and they won't be there, but we'll model them. Now the question is, which is more accurate as to citizenship? [GENERAL FRANCISCO] Sure. [JUSTICE BREYER] The models over here on the administrative part or the answer to the questions on the census part? And here is what Dr. Abowd says: 22 million, he's asked, if you follow your practice, you'd use a survey response, not model it; is that right? That's right. And in your opinion, that would be less accurate than if you just went with modeling over on the administrative part? That's correct. And the conclusion of the Census Bureau remains that adding the question over here, even if you use the administrative part too, produces worse citizen... worse data on citizenship than just using the administrative data alone? That's the question. [GENERAL FRANCISCO] Yep. [JUSTICE BREYER] Answer from the expert: "That's correct." So I read that, and, you know, the judges below have listed 14 other examples or other examples, many other examples, and... and... but that's the most direct. So where in the decision memo did the Secretary address that problem? [GENERAL FRANCISCO] So, sure, Your Honor. [JUSTICE BREYER] Both problems. There are two problems. [GENERAL FRANCISCO] Yes, yes. And I'd like to address that evidentiary issue first and then I would like to come back to your question about Section 6(c) of the statute as well. If you look at the Joint Appendix page 148, the... the Census Bureau staff specifically told the Secretary... and here I'm quoting from it... that it "cannot quantify the relative magnitude of the errors across alternatives"... and he's talking about Alternatives C and D... "at this time." So what he was saying was that I don't know if the response error from asking the question is going to be more or less than the prediction error if I use... [JUSTICE BREYER] No, he said just what you said. He said, I cannot quantify it. [GENERAL FRANCISCO] Sure. [JUSTICE BREYER] And that means he can't put scientific numbers. Of course, they said that they wanted two years to test it, but they can't quantify it. But we do have three studies, and those studies look at what happened when you asked this question before, and what happened when you asked this question before... [GENERAL FRANCISCO] Right. [JUSTICE BREYER] ... is the response rate fell. [GENERAL FRANCISCO] And so, if I could complete my answer, what the Bureau staff told him was that they didn't know which one would be better or worse. So what the Secretary... [JUSTICE SOTOMAYOR] I'm sorry. [JUSTICE BREYER] Where does it say that? Where does it say that? [CHIEF JUSTICE ROBERTS] Maybe you could, if you don't mind, maybe you could complete your answer. [GENERAL FRANCISCO] Sure. That is specifically, Your Honor, at page 148 of the Joint Appendix, where it specifically says and explains that it cannot quantify the relative magnitude of the errors across the alternatives, Alternatives C and D, at this time because it didn't know if the response errors from asking the question would be more or less than the prediction errors from the model. So what the Secretary knew was two basic things... [JUSTICE SOTOMAYOR] I'm sorry. They... they have prediction models. They say multiple times, at least three, if not more, that Alternative D, which was the Secretary's alternative, and their Alternative C, so everybody's clear, C was simply to use administrative records, D was the Secretary's idea of adding the question to the survey plus administrative records. And on the prediction models, which is what scientists can do, each and every time they said D would be less accurate than C. Now you're asking... [GENERAL FRANCISCO] Except for the one time where it mattered, Your Honor, in the key differences... [JUSTICE SOTOMAYOR] But comparative... [GENERAL FRANCISCO] ... memo, where they specifically said that they did not know if C was better than D. And so what the Secretary knew... [JUSTICE SOTOMAYOR] No, no, no, no, that's not what he said. [GENERAL FRANCISCO] That is exactly what he said. [JUSTICE SOTOMAYOR] He said you can't... you're... the words "comparative errors" have a different meaning than you're giving it. Comparative errors are I'm comparing this type of error to that type of error and what they compare each other to. You can't do that to a scientific certainty. But you can have predictive models, which is what they did, and they... [GENERAL FRANCISCO] Right. [JUSTICE SOTOMAYOR] ... showed you, time and again they told you, you add the survey question, it's going to be less accurate than just relying on administrative records. [GENERAL FRANCISCO] Your Honor, I... [JUSTICE SOTOMAYOR] So how do you... how do you take or pluck out of what they say in one sentence, if you're the Secretary, and rely on that one sentence and ignore the wealth of statistics, graphs, testimony, proof, control studies of how... how these response rates came about and decide that that one sentence is enough to justify ignoring everything else? [GENERAL FRANCISCO] Because, Your Honor, I respectfully disagree with your reading of the administrative record. If you read through the key differences memo, what the Bureau staff is telling to the Secretary... and, look, there's no question that the Bureau staff preferred not to have this question on the census... but what they were telling the Secretary was that they couldn't tell which model would be more or less accurate. But they did give him specific information. [JUSTICE KAGAN] But, General... [GENERAL FRANCISCO] They told him that... [JUSTICE KAGAN] Please finish. [GENERAL FRANCISCO] They told him that if he actually asked the question for 30... for 22.2 million people for whom no administrative records existed, he would have got actual answers at 98 percent accuracy. And that the alternative, their preferred alternative, was to use a statistical model to estimate citizenship, not just for the 22.2 million but for 35 million, but they had not yet constructed that model and didn't know what the error rate in that model would be. [JUSTICE KAGAN] But I... I... [GENERAL FRANCISCO] And so the... [JUSTICE KAGAN] ... I think, General, I mean, 98 percent sounds awfully high, but it's kind of irrelevant too. The question is whether, if you used the model, it would be greater than 98 percent. It would be 99.5 percent. [GENERAL FRANCISCO] Right. [JUSTICE KAGAN] Because then the Secretary would have no basis for saying that you should use the question rather than the model. And as to that, as I think my colleagues are suggesting, there is a bottom-line conclusion from the Census Bureau, and the bottom-line conclusion is that Alternative D, which is the proposal that the Secretary eventually took, would still have all the negative cost and quality implications of Alternative B, which was simply adding the question alone, and would result in poorer quality citizenship data than Alternative C, which is just using the administrative records plus the modeling. So there is a bottom-line conclusion from the Census Bureau. And it seems as though what the Secretary needs is some... I mean, a Secretary can deviate... [GENERAL FRANCISCO] Uh-huh. [JUSTICE KAGAN] ... from his experts' recommendations and from his experts' bottom-line conclusions. [GENERAL FRANCISCO] Right. [JUSTICE KAGAN] But the Secretary needs reasons to do that, and I searched the record and I don't see any reason. [GENERAL FRANCISCO] Sure. And, Your Honor, so I want to just finish what I was saying instantly before because I think it... it responds to your question, and then I'd like to expand directly in response to your question. What the Secretary concluded was, in the face of uncertainty, he'd rather go with the bird in a hand and ask the question at percent accuracy than an unknown and untested statistical model. And that's, after all, the same preference that the Enumeration Clause itself makes, a prefer... preference for actual counting over estimation, because actual counting is less efficient. [JUSTICE SOTOMAYOR] But not census... [JUSTICE KAGAN] Well, if I can just add to the question, and... [GENERAL FRANCISCO] But to go... to go to your question... yes. [JUSTICE KAGAN] ... and let me just add to it a bit because you said, you know, an unknown and untested statistical model, but here are his experts in the Census Bureau saying we are confident that we can produce a statistical model that will produce more accurate bottom-line results, and... and, again, this bottom-line conclusion is the same. They know what kind of statistical models they can build. And this is the bottom-line conclusion. And where is the reason that the Secretary gives as to why he rejects that? [GENERAL FRANCISCO] So there are a couple of reasons, Your Honor. First of all, although they had a high confidence that they could create a good statistical model, they were not able to tell him that they thought that that model would be any better or worse an estimation. They never were able to say that would... it would beat that 98 percent number. So, in the face of that uncertainty, he reasonably chose... [JUSTICE KAGAN] I think what I read you is them saying that they could beat the percent number. [GENERAL FRANCISCO] No, Your Honor, I don't think that's what they're saying there. I think what they're saying there is that if you ask the citizenship question, it will make the model a little bit less accurate because you're going to have fewer people for whom there are administrative records, but you actually have to use that model for a much smaller number of people because you have actual answers from 22.2 million at 98 percent accuracy. So that's what they're saying. But what they couldn't come down with a conclusion on is whether it would be more or less accurate to ask the question... [JUSTICE BREYER] All right. When you answer this question... [GENERAL FRANCISCO] ... or use the model. [JUSTICE KAGAN] So can I just ask... [JUSTICE BREYER] Would you answer this question? Wait. I am trying, go back to my question, to write down the pages. I'm not going to, you know, resolve this right now. So, when Justice Kagan asked the question or I do, you've mentioned page 148. Insofar in answer to her question or to mine, if you could give me a few things to read to show that he did consider it, to show that he did give a reason for rejecting it, that would be helpful to me. Sorry. Go back to you. [GENERAL FRANCISCO] So, Your Honor, if you look at Secretary Ross's decision memo, the pre-... the decision memo in the Petitioners' appendix, I would look to a couple of things. First of all, I would look to his discussion on pages 555a with the problems with administrative records. The bureau is still evolving its use of administrative records, and the bureau does not yet have complete administrative records data set for the entire population. And that points out why he preferred asking the question and getting actual answers from 22.2 million people at 98 percent accuracy, because you simply didn't have administrative records for 35 million people, and the bureau had not yet figured out how to do that estimation. He then goes on to say on that same page, more than 10 percent of the American population, some 25 million voting-age people, would need to have their citizen age imputed by the Census Bureau. And so he was making clear that he'd rather go with actual counting than imputation. And he pointed out that by proceeding with his preferred course... this is at page 556a of the Petitioners' appendix... this may eliminate the need for the Census Bureau to have to impute an answer for millions of people, specifically about 22.2 million people for whom the bureau... [JUSTICE KAGAN] But... [GENERAL FRANCISCO] ... told him he would get actual answers at 98 percent accuracy. [JUSTICE KAGAN] ... if I may say, General, when you think about those statements, I mean, it's just conclusory. It's just like, well, this would eliminate the... the... the need for modeling because we could ask a question. But the question is, why is asking a question better when you know that asking a question is going to result in lots of non-responses and in lots of false reporting? And so you can't just go back to I'd rather ask a question. You have to say why you'd rather ask a question and what benefits it has to ask a question. [GENERAL FRANCISCO] Right. [JUSTICE KAGAN] And if I may say... I'll just finish here, General... I mean, a lot of your argument... your briefs are extremely well done. [GENERAL FRANCISCO] Thank you. [JUSTICE KAGAN] But a lot of your argument just do not appear in the Secretary's decision memo. And... and the fact that SG lawyers can come up with 60 pages of explanation for a decision, that's all post hoc rationalization. The question is, what did the Secretary say? Where did he say it? When did he say it? What does it mean, other than just ipse dixit and conclusions? [GENERAL FRANCISCO] Sure. Your... Your Honor, I... I'm tempted to pocket the compliment and sit down, but I won't do that. [LAUGHTER] [GENERAL FRANCISCO] I think the Secretary fully acknowledged that there was an upside to the request, and the upside was the one that the Department of Justice set forth in its letter, that having citizenship data would help improve Voting Rights Act enforcement. He fully understood there was an alternative, using administrative records, and he analyzed that alternative in the language that I just read to Justice Breyer, and he understood there was a downside, that adding the citizenship question would potentially increase self-response... decrease... increase the number of... decrease the number of self-response rates. But he found two things with respect to that. First, he found... and all of this is in his letter... that he could mitigate that to at least a certain extent with follow-up operations, perhaps not entirely but at least to a certain extent, and, secondly, to the extent that materialized, it was the product of illegal activity. So he considered the benefits. He considered the alternatives. He considered the costs... [JUSTICE GINSBURG] Which... which letter... [GENERAL FRANCISCO] ... and he concluded that the benefits outweighed the costs. [JUSTICE GINSBURG] General, which letter are we talking about? [GENERAL FRANCISCO] Secretary Ross's decision memo in March 2018, Your Honor. [JUSTICE GINSBURG] The memo, not... [GENERAL FRANCISCO] The memo, yes, Your Honor. [JUSTICE GINSBURG] ... not the letter from... [GENERAL FRANCISCO] Yeah. [JUSTICE GINSBURG] ... Department of Justice. [GENERAL FRANCISCO] Yes, Your Honor, the memo. But the Department of Justice's letter is the one that articulated the Voting Rights Act rationale that formed... [JUSTICE KAVANAUGH] Can you explain... [GENERAL FRANCISCO] ... the principal benefit. [JUSTICE KAVANAUGH] ... can you explain how it would improve Voting Rights Act enforcement? [GENERAL FRANCISCO] Yes, Your Honor. One of the critical elements of Voting Rights Act enforcement is something called Citizen Voting Age Population, or CVAP. Right now, everything for CVAP comes from the census, with the exception of citizenship. So population, age, race, all of that comes from the census, except for citizenship, the C in CVAP. So a large amount of voting rights litigation focuses on expert witnesses who try to fill in that missing C and try to estimate that missing C through imputation based on the American Community Survey, which goes to just one in 38 households. And the Department of Justice wanted to get all of the same information from the same database so that critical feature of voting rights litigation, CVAP, all came from the same place. [JUSTICE SOTOMAYOR] General... [JUSTICE KAGAN] And to just... [JUSTICE SOTOMAYOR] ... how do we know that DOJ couldn't do Alternative C, rely on the administrative records? The one thing that we do know is that the Secretary went to the Department... went to DOJ at the beginning and asked them for help in adding the citizenship question. They initially said no. At least their lower-level people said no, told them to shop it to DHS and see if DHS wanted the citizenship question. DHS said not our work... [GENERAL FRANCISCO] Uh-huh. [JUSTICE SOTOMAYOR] ... go back to DOJ. They go back to DOJ, the people they're in touch with, that are not low-level, but they're not the highest level, say no. And Secretary of Commerce speaks to the head of DOJ at the time... [GENERAL FRANCISCO] Uh-huh. [JUSTICE SOTOMAYOR] ... and the head of the DOJ says: We'll give you anything you need. They do a letter, the letter says the ACS is not enough. What the letter doesn't say is, ah, if you supplemented with administrative records, which 6(c) lets you do and tells you you should do to maximize the extent possible that the actual count is accurate, and we do know that there will be less people being reported, which is the whole purpose of the survey... [GENERAL FRANCISCO] Uh-huh. [JUSTICE SOTOMAYOR] ... how many people there are, and nobody doubts that there will be less people reported. That's a maximum need of the census survey report, not citizenship. Let's not confuse the two things. The enumeration is how many people reside here, not how many are citizens. That's what the census survey is supposed to figure out. [GENERAL FRANCISCO] Uh-huh. [JUSTICE SOTOMAYOR] DOJ needs citizenship. But, when the Bureau asks DOJ, not the Secretary, to meet so they can discuss why the administrative records are not good enough, they say, we don't need to. So tell me, in that sequence, how does the Secretary know the answer... [GENERAL FRANCISCO] Sure. [JUSTICE SOTOMAYOR] ... to that question? [GENERAL FRANCISCO] Sure. So a couple of responses to that, Your Honor. I'm going to start with the 6(c) issue because I know that was of interest to Justice Breyer as well. And under 6(c), under my... my friend's on the other side's position, you actually couldn't even ask the citizenship question on the American Community Survey. And you also couldn't ask about sex and age on the census itself since all of that information is all also available in administrative records. But the reason why administrative records are insufficient under 6(c) for any of these purposes is for the simple reason that you don't have them for 35 million people. In terms of the Department of Justice's request and the Census Bureau's alternative proposal, it simply wasn't responsive to the Department of Justice's request for two reasons. First, administrative records didn't solve the problem that the Department of Justice was trying to solve, which was getting all of their CVAP data from the same source and covering the same time period. Administrative records come from a different database and cover a different time period than all of the other information used to construct Citizen Voting Age Population. [JUSTICE SOTOMAYOR] That doesn't tell you why it's not good enough. They may have wanted something. [GENERAL FRANCISCO] Sure. [JUSTICE SOTOMAYOR] But they would then be introduced with one database that has been, according to the chief statistician of the Bureau, introducing multiple layers of uncertainty. [GENERAL FRANCISCO] Right. [JUSTICE SOTOMAYOR] Uncertainty about or an undercount of people because they already say that undercount's going to be at a minimum 5.8, less people are going to respond. You're going to have a lesser number that are going to group with the administrative record. You're going to have 9.5 million that conflict between their answer and the administrative records. And we have to change the bureau's use of that information to be able to use the administrative record. And the Secretary doesn't ask, if we change that, what else will it affect? [GENERAL FRANCISCO] Right. [JUSTICE SOTOMAYOR] So he doesn't know that. And we now have error in the unreporting population of at least 500,000. So something the chief statistician of the bureau tells us is, unlike our simpler prior models, this introduces more uncertainty at every single level of the calculus. And so that data is going to be more suspect, more prone to cross... to less reliability, and less accurate. And so, if the bureau... if the Department of Justice refused to listen to that, how can the Secretary conclude that he's complying with 6(c) fully? Because it says to the maximum extent possible, and how can you be possible if you don't even ask why? [GENERAL FRANCISCO] So... [JUSTICE SOTOMAYOR] This seems like he thought of something, I want to add a citizenship question, I don't know why, but this is a solution in search of a problem. I've got to find a problem that fits what I want to do. [GENERAL FRANCISCO] So, Your Honor, there's a whole lot in that question, but I think I will start with where you ended. And if you really think 6(c) is a problem, then we really cannot ask the citizenship question on the American Community Survey since that is just as subject to 6(c) as the census is. [JUSTICE SOTOMAYOR] No, what it says... what it says is to the maximum extent possible. [GENERAL FRANCISCO] And here the Secretary is using the administrative records to the maximum extent possible because he's combining them. [JUSTICE SOTOMAYOR] But the ACS is not the survey. [GENERAL FRANCISCO] He's combining them with the administrative records and the self-responses and using administrative records where they're available, using self-responses at 98 percent accuracy... [JUSTICE SOTOMAYOR] But the problem... [GENERAL FRANCISCO] ... where administrative records are not available. [JUSTICE SOTOMAYOR] ... but the problem is you can't confuse the survey, which is really the question of 6(c)... [GENERAL FRANCISCO] No. [JUSTICE SOTOMAYOR] ... or the focus. [GENERAL FRANCISCO] It's not. [JUSTICE SOTOMAYOR] Because... [GENERAL FRANCISCO] 6(c) applies to all census instruments, not just the census. It fully applies to the American Community Survey. And it likewise applies... [JUSTICE SOTOMAYOR] But the ACS is not used... the ACS is not used for the citizenship purpose. [GENERAL FRANCISCO] But, Your Honor, your question... [JUSTICE SOTOMAYOR] You already said why it wasn't. [GENERAL FRANCISCO] ... your question is about 6(c). 6(c) applies to the ACS and it applies to the census. And sex and age information, which we ask on the census, is also available in administrative records. Indeed, administrative records are more accurate with respect to sex and age because, presumably, your birth date and your sex don't change over the course of time, whereas your citizenship status does. So, if you really think that 6(c) is a problem, we can't ask it on the ACS and we can't ask sex and age on the census, so that's why I think that is plainly wrong. It really does boil down to whether the Secretary's judgment here is a reasonable one. And in the face of two competing possibilities, either asking the question, getting answers for two-thirds of the people for whom no administrative records existed, at percent accuracy, or using an estimation model that had not yet been created and had an unknown error rate, the Secretary reasonably chose to go... [JUSTICE KAGAN] But just... [GENERAL FRANCISCO] ... with the bird in the hand. [JUSTICE KAGAN] ... General, just going back to Justice Kavanaugh's simple question about why the Secretary thought that there was a need for this data, and then part of what Justice Sotomayor was talking about was that it did really seem like the Secretary was shopping for a need. Goes to the Justice Department. Justice Department says, we don't need anything. Goes to DHS. DHS says they don't need anything. Goes back to the Justice Department, makes it clear that he's going to put in a call to the Attorney General. Finally, the Justice Department comes back to him and says: Okay, we can give you what you want. So you can't read this record without sensing that this... this need is a contrived one. Nobody had... there have been lots of assistant attorney general in the Civil Rights Division that have never made a plea for this kind of data. And... and just the way this went back and forth, I guess I'd like an answer to that simple... [GENERAL FRANCISCO] Yeah. Sure. [JUSTICE KAGAN] ... question. [GENERAL FRANCISCO] And I have two responses, Your Honor. And then, if I may, I'd like to reserve the remainder of my time for rebuttal. First, I think it is quite common for cabinet Secretaries to come into office with ideas and inclinations to discuss with their staff and discuss with their colleagues whether there is a legal and policy basis for that inclination. Secondly, there's no evidence in this record that the Secretary would have asked this question had the Department of Justice not requested it. And there's no evidence in this record that the Secretary didn't believe that the Department of Justice actually wanted this information to improve Voting Rights Act enforcement. [CHIEF JUSTICE ROBERTS] Thank you, General. General Underwood.

[MS. UNDERWOOD] Mr. Chief Justice, and may it please the Court: The Secretary decided to add this question about citizenship to the 2020 census although the record before him contained uncontradicted and strong evidence that it will cause a decline in the response rate of non-citizens and Hispanics, to the detriment of the states and localities where they live. He gave three reasons for the decision, and none of them can survive APA review. One, he said there was inadequate evidence of an effect on the response rate. But that is flatly contrary to the record. He said he could dismiss or discount any such effect because non-response is an illegal act. But that is an irrational and impermissible factor to consider on this question. And he said that adding the question would help voting rights enforcement. But that claim is unsupported by the record as well. [CHIEF JUSTICE ROBERTS] Do you... do you think it wouldn't help voting rights enforcement? The CVAP, Citizen Voting Age Population, is the critical element in voting rights enforcement, and this is getting citizen information. [MS. UNDERWOOD] Well, as we have... as has been discussed at length in the... during the previous argument, the evidence before him was that it would not give better citizenship information than... that it's the 22 million that the government points to, the 22 million whose citizenship information will be either modeled or the result of the answer to a census question. [JUSTICE GINSBURG] Was there anything that showed that the Department would have been aided in either past cases or cases on the drawing board? Any case? [MS. UNDERWOOD] There was not. And what I'd like to point out is that the comparison should be to using administrative records. 2 The Department of Justice letter, taken at face value, says the old ACS survey data that we've been using is inadequate and we need an improvement over that. The Census Bureau produced this answer, which is we can do this by linking the existing census information to administrative records. The Department of Justice never commented on that. The Department of Justice actually declined to meet with the Census Bureau people who wanted to meet about it. So there was nothing before the Secretary to say that this survey... this census information would be an improvement... there was no comparison at all from the Department of Justice about whether this would be an improvement or not. It seems to me that at least with so much question about whether this information would be better or worse than the... the use of modeling from administrative records, the Secretary had an obligation to find out the answer to that question. [JUSTICE KAVANAUGH] What if the answer was uncertain? [MS. UNDERWOOD] Well, if the answer is uncertain, then it is hard to invoke that as a reason. Now we get back to the cost in the enumeration. That is, if the... if it's unclear... we think it's worse. But, if it's just unclear whether this question will improve voting rights enforcement, that is not sufficient to pay the cost of the steep decline in the enumeration because the enumeration is, after all, the primary purpose of the census. [JUSTICE ALITO] Well, on the modeling... [CHIEF JUSTICE ROBERTS] Well, is that... go ahead. [JUSTICE ALITO] On the modeling, there was a lot of talk during the first part of the argument about... I think it's 22.6 million people who it would... it is predicted would answer the citizenship question and as to whom there is not administrative data. And there was an estimate that those answers would be 98 percent accurate. And the comparison then has to be between that percent predicted accuracy rate and whatever the accuracy rate would be for the model. And is there anything in the administrative record that shows that the model was tested and that it was possible to extract a... a predicted error rate for the model? [MS. UNDERWOOD] What we have... what we have is that this... the model hadn't been generated, but what we have is the Census Bureau saying this is like other modeling that we routinely do. We're confident that we can do it. [JUSTICE ALITO] So, if the Secretary said... has to choose between two things, and, on one, the Secretary knows there's a percent accuracy rate, and as to the other, the Census Bureau says, we're going to create a model, and we don't know how... we can't give you any statistics, but trust us, it's going to be more accurate than 98 percent, is it arbitrary and capricious for the Secretary to say, I'll go with the 98 percent because that's a known quantity? [MS. UNDERWOOD] If there were no cost to the enumeration, that would be a different question. But, when there is this much uncertainty, then it is arbitrary and capricious to take that kind of risk... [JUSTICE BREYER] I don't understand... sorry. [MS. UNDERWOOD] ... with the enumeration. [JUSTICE BREYER] I don't understand uncertainty. I thought the 98 percent... of course, it's 98 percent. Most people are citizens. [MS. UNDERWOOD] Right. [JUSTICE BREYER] The people who are citizens are not going to... you know, they'll say they're citizens. All you'd ever expect are a few percent who are not citizens. Then I have on pages J... Joint Appendix 882 through 884 Mr. Abowd's testimony, where he unequivocally says three times that in... in... not 98 percent... in respect to those people who are not citizens, the administrative model will be more accurate than just asking the census question, and, if you add the census question, then you look to it for the answer, you will discover that you are less accurate in respect to non-citizens. Now he says that. That's why I asked the Solicitor General what is their contrary to that, and he gave me things to look at. And I would say the most contrary thing, which I want to ask you about, is Dr. Abowd, at the trial, said, "There's no credible quantitative evidence that the addition of a citizenship question will affect the accuracy of the count." All right? That's what he said. [MS. UNDERWOOD] Yes. [JUSTICE BREYER] Now I'm sure in their reply brief they pointed right to that. You can't simply ignore it. And so I want to hear what your answer is to that, which the government says is contrary evidence. [MS. UNDERWOOD] If you look at that testimony in context, it is perfectly clear that what he is saying is that he didn't have enough evidence for a firm quantitative statement, meeting scientific standards. He actually defined the term "credible quantitative evidence." And if you'll bear with me, he... he said it's evidence that is specifically related to the insertion of a citizenship question into the otherwise planned 2020 census that identifies the citizenship question itself as the likely or one of the causal elements associated with changes in the outcomes and that would stand up to extensive peer review within the Census Bureau and with the scientific community. [CHIEF JUSTICE ROBERTS] That... [MS. UNDERWOOD] And... and the... [CHIEF JUSTICE ROBERTS] ... that's the evidence that he found was not available? [MS. UNDERWOOD] Yes, that's right. He said he didn't have enough to quantify accord... in accordance with peer-reviewed standards. He didn't say there would be no effect. He said, I don't have enough to give what I... what I believe as a scientist to be this term, "credible quantitative"... to make credible quantitative evidence. [JUSTICE KAVANAUGH] Can... can I ask... [MS. UNDERWOOD] So he is... [JUSTICE KAVANAUGH] I'm sorry, please finish. [MS. UNDERWOOD] And... and my point is that, at the same trial, if we're going to look at the trial testimony, at the same trial, other experts said: Maybe so, but there is enough evidence to make a different kind of judgment, not a firm scientific, quantified judgment, but a judgment. [JUSTICE ALITO] Do you think it's proper to look at the trial record on this issue? There's a lot of citation in the Respondents' brief to trial testimony. Aren't we reviewing the administrative record? [MS. UNDERWOOD] We are. Well, as for standing, we're reviewing the whole record. And as... [JUSTICE ALITO] That's correct. But as to the arbitrary and capricious review... [MS. UNDERWOOD] Correct. That's... that... that is correct, but that would make this statement also of Dr. Abowd off the record, off the administrative record. [JUSTICE KAVANAUGH] Can I... [MS. UNDERWOOD] But what I... [JUSTICE KAVANAUGH] ... can I ask a... [MS. UNDERWOOD] Yes. Please. [JUSTICE KAVANAUGH] Please finish. [MS. UNDERWOOD] No, going... going... going back to your question... [CHIEF JUSTICE ROBERTS] Wait a minute, Justice Kavanaugh. [MS. UNDERWOOD] ... while I think there is good evidence, and nothing contrary, that this 22 million would be more accurately identified by the modeling than by the census, I think it is sufficient for this purpose to treat it as somewhat uncertain, because it is uncertainty with respect to the discretionary part of what the Census Bureau does; namely, collect extra information. The core function of the census, not of the Census Bureau in all its actions, but of... on the census form... [JUSTICE KAVANAUGH] But the... the United Nations recommends that countries ask a citizenship question on the census. And a number of other countries do it. Spain, Germany, Canada, Australia, Ireland, Mexico ask a citizenship question. And the United States has asked a citizenship question, as you know, in one form or another since 1820, excluding 1840, and, again, long form at times, in more recent times, and then on the ACS since 2005. The question is, does that international practice, that U.N. recommendation, that historical practice in the United States, affect how we should look at the inclusion of a citizenship question in this case? [MS. UNDERWOOD] The same guidance from the U.N. also says to be careful to test questions to make sure they don't interfere with the enumeration. It says you need to make a judgment in context. It may be that those countries either haven't examined or don't have the problem that has been identified... the problem of depressing the enumeration that the United States has. It's certainly something to look at, but... [JUSTICE KAVANAUGH] But you agree it's very... it's a very common question internationally? [MS. UNDERWOOD] Well, it is certainly useful information for a country to have. And I'm not suggesting at all that that information shouldn't be collected. The question is whether it should be collected on the very instrument that is... whose principal function is to count the population, when we have such strong evidence that it will depress that count, make it less accurate, and make it less accurate in a... [CHIEF JUSTICE ROBERTS] Well, the principal purpose... you're... you're right, the principal purpose is to count the population, but we've had demographic questions on the census... I don't know how far back, but, certainly, it's quite common. [MS. UNDERWOOD] That's... that's correct, but we have no evidence about... [CHIEF JUSTICE ROBERTS] Sex, age, things like that. You go back and it looks... you know, do you... do you own your house? Do you own a radio? I mean, the questions go quite beyond how many people there are. [MS. UNDERWOOD] Well, I'd like to say two things about that. We have no comparable evidence about any of those other questions that they depress the count in this substantial a way and in this disproportionate a way because, as this Court said in Wisconsin, distributive accuracy is even more important for the census. [JUSTICE ALITO] Well, the... the... the response rate is very important, so can I ask you a question about that? A lot of your argument and a lot of the district court's argument seems to hinge on this prediction that there will be 5.1 percent fewer responses if the citizenship question is included on the census. But that seems... that is based, as I understand it, on the fact that non-citizens are somewhat less likely to complete the ACS, which includes the citizenship question, than are citizens. Am I right in understanding that? That's fundamentally where that comes from? [MS. UNDERWOOD] It's not about not completing. It's about not... I mean, it's not about skipping questions on a form. [JUSTICE ALITO] Not... not responding. [MS. UNDERWOOD] It's about not... not responding. Yes. [JUSTICE ALITO] That's correct. Okay. They're somewhat less likely to respond to the ACS than are... than are... [MS. UNDERWOOD] It's the ACS in one study and the long form in another. [JUSTICE ALITO] Okay. But what jumps out is the fact that citizens and non-citizens differ in a lot of respects other than citizenship. They differ in socioeconomic status. They differ in education. They differ in language ability. So I don't think you have to be much of a statistician to wonder about the legitimacy of concluding that there's going to be a 5.1 percent lower response rate because of this one factor. But maybe there's something more there. So what... what does that analysis miss? [MS. UNDERWOOD] The strong... well, a couple of things. The strong empirical evidence that is the basis for that judgment, which, by the way, has not been contested by the government, the government has other things to say but does not contest this decline... [JUSTICE ALITO] I thought they did, but, in any event, go ahead. [MS. UNDERWOOD] I'll come back to that... is a... is a retrospective review of comparing in one case for 2010 the short-form census and the ACS, and in 2000, it was to compare the short form and the long-form census. It's a comparable comparison. In each case, the longer one had a citizenship question on it. In each case, everyone, population groups notwithstanding, there was a decline from the short form to the long form. But there was a much greater decline among Hispanics and non-citizens. [JUSTICE GORSUCH] But, counsel, isn't... doesn't Justice Alito have a point to the extent that there could be multiple reasons why individuals don't complete the form? [MS. UNDERWOOD] Well, the lay... [JUSTICE GORSUCH] And we haven't... [MS. UNDERWOOD] I'm sorry, go ahead. [JUSTICE GORSUCH] Plenty of interrupting. But we don't have any evidence disaggregating the reasons why the forms are left uncompleted. What do we do with that? I mean, normally, we'd have a regression analysis that would disaggregate the potential causes and identify to a 95th percentile degree of certainty what the reason is that persons are not filling out this form and we could attribute it to this question. We don't have anything like that here. So what are we supposed to do about that? [MS. UNDERWOOD] Well, I think... I think there are a few things to say. [JUSTICE GORSUCH] And... and... and... [MS. UNDERWOOD] Well... [JUSTICE GORSUCH] ... and let me just throw in one other question. I know your light's on, but I really wanted to get it to you and I'm sorry we haven't gotten there. And that is, what do we do also... and it's totally different, so I'm really sorry... what do we do with the fact that, as I understand it, some of the Respondents and other people in litigation have complained when... when folks have relied on the ACS to extrapolate citizenship for purposes of redistricting and, in fact, argued that we should rely only on actual census data? And I understand Respondents have made that argument in litigation. So what do we do with that? [MS. UNDERWOOD] There are a lot of complaints about the ACS. The Census Bureau's proposal to use administrative records solves most of them. It's not a question of just the ACS, which is a survey about which there have been many complaints, and the... putting the question on the census. The Census Bureau is... they are data experts. There are many ways of trying to collect data. The question in this case is whether doing it on the census form is warranted, even though it causes such a harm to the count. [JUSTICE SOTOMAYOR] I thought that... [MS. UNDERWOOD] Now that brings us back... [JUSTICE SOTOMAYOR] ... Dr. Abowd's testimony, or at least the letter, somewhere I read, that they controlled for the... for all of the other reasons that Justice Gorsuch was mentioning as reasons why people would not complete? [MS. UNDERWOOD] Well, they certainly controlled for the length of the form. That was the... that was the... [JUSTICE BREYER] Well, not just that. On page 110... [MS. UNDERWOOD] Yes. Yes. [JUSTICE BREYER] ... of the Joint Appendix, it says whether the response... dah-dah-dah... that they're much greater. It says in comparable rates for other demographic variables, like sex, birth date... [MS. UNDERWOOD] Yes. [JUSTICE BREYER] ... age, race, ethnicity. So I thought that that was an effort to control for the things that Justice Alito... [MS. UNDERWOOD] It... [JUSTICE BREYER] ... mentioned insofar as their relevance. [MS. UNDERWOOD] It was. The... the only limitation on it was that they had to deal with data that already existed. Dr. Abowd wanted to do a random controlled test of this question and wasn't permitted to do so. [JUSTICE ALITO] But that's not... that's a... that's a different issue, isn't it, what Justice Breyer mentioned? [JUSTICE GORSUCH] Yeah. It's totally different, yeah. [JUSTICE ALITO] It's... it's... it's the decline in the response rate based on those variables, but not... it doesn't... it doesn't, as Justice Gorsuch says, disaggregate the many factors that could explain a decline when you're distinguishing between citizens and non-citizens. [MS. UNDERWOOD] Well, it did try to control for other properties that citizens have... [JUSTICE GORSUCH] But we don't... [MS. UNDERWOOD] ... and non-citizens... [JUSTICE KAGAN] Am... am I right... [MS. UNDERWOOD] ... have... [JUSTICE KAGAN] ... that there... [JUSTICE GORSUCH] It's fair to say we don't have this isolated, though, isn't it? [MS. UNDERWOOD] They did their best. [JUSTICE GORSUCH] They did their best. [MS. UNDERWOOD] There is some degree of isolation, enough to enable them to believe that they had isolated the factors that people thought of as plausible. [JUSTICE BREYER] Yeah, I mean... [JUSTICE KAGAN] Would it be right to say... [JUSTICE BREYER] ... there are a million factors. [MS. UNDERWOOD] Yes. [JUSTICE BREYER] There's pet dogs, you know. [MS. UNDERWOOD] Yes. [JUSTICE BREYER] I mean, there are cats. And... and so, if, in fact, there were some factors that are relevant which were not in the data because they only controlled for six other factors instead of 600, I would expect somewhere in this record someone to have written that there were these other factors that also should have been controlled for. I know what you're going to say, unless I'm wrong, you better not tell me that I am right if I'm not. But I could not... [LAUGHTER] [JUSTICE BREYER] ... find any such place in the record. [MS. UNDERWOOD] Nobody proposed, that I know of, proposed factors that might be alternative explanations that should have been tested for. [JUSTICE KAGAN] And... and would it be right to say, General, that it was the Census Bureau's conclusion, a bureau full of statisticians... [MS. UNDERWOOD] Yes. [JUSTICE KAGAN] ... that it was the citizen... citizenship question that was driving the differential response rates? [MS. UNDERWOOD] That is... that is correct. I do want to make... [CHIEF JUSTICE ROBERTS] Are there other... are there other questions on the census for which the administrative records provide more accurate information? [MS. UNDERWOOD] There is nothing in the record about that. [CHIEF JUSTICE ROBERTS] Well, then I don't want to hear about it. [MS. UNDERWOOD] Okay. Okay. [JUSTICE SOTOMAYOR] Could you answer the General's point that, if you rely on 6(c), then you shouldn't be even asking this on the ACS form? [MS. UNDERWOOD] No. The... for one thing, in order to do modeling, in order to do sampling, they need some survey data to compare it to. And so some judgments can be made, and the judgment might be made that the ACS or some questionnaire that doesn't involve harm to the count that is sampling or... or some other form of... less... less than universal questioning, that testing questions on that kind of instrument is the way to do it. But... [CHIEF JUSTICE ROBERTS] Thank you. Thank you, General. Mr. Ho.

[MR. HO] Mr. Chief Justice, and may it please the Court: The Secretary's decision rested primarily on one assertion, that it would improve the accuracy of citizenship data provided to the Department of Justice. But the administrative record revealed precisely the opposite, that it would make that data less accurate and, thus, harm the Secretary's stated purpose of Voting Rights Act enforcement. And the Secretary's explanation for his decision misstated the evidence in the record in three critical respects. First, the Secretary asserted that adding the question would maximize the Census Bureau's ability to use administrative records on citizenship. But the government has conceded that that was not true. At page 32 of their brief, they acknowledge that, if the question is added, the number of people who can be matched to these administrative records, the most accurate information that we have on citizenship, will fall by one million. Second, the Secretary asserted that adding the question would improve the bureau's imputation of citizenship for people who lack... for people for whom the government lacks any such records. But the government has conceded that that was not true either. At page 34 of their opening brief, they acknowledge that the Census Bureau determined that if the question is added, the imputation process will become less accurate. And here's why: The accuracy of imputation depends upon the accuracy of existing data. Federal administrative records are based on a person's legal documents of their citizenship and, thus, are quite accurate and reliable for this purpose. But the citizenship question is not. The evidence shows that non-citizens respond to the question inaccurately one-third of the time. So, if the question is used, the data that's used for imputation will be contaminated by those incorrect responses, making the output of the imputation process less accurate, making the data less accurate, and, again, harming the Secretary's stated purpose of improving the accuracy of citizenship information. The Secretary misstated the evidence in a third respect. He asserted that adding the question would fill in the gaps in 22 million... for 22 million people in administrative records on citizenship. But the Census Bureau concluded specifically, on the last page of their March 1 memo in the administrative record, that the Secretary's decision will not solve that problem. And the reason is, again, because responses to the question are highly inaccurate, whereas the imputation process, based solely on... on administrative records, would be more accurate. That's reflected in the Census Bureau's bottom-line conclusion in its March 1 memo, and it's reflected in the testimony of Dr. Abowd on that trial. [JUSTICE ALITO] But this takes us back, does it not, if I'm following your argument, to the 22.6 million people who will answer the citizenship question but as to whom there aren't administrative records? That's what you're talking about? [MR. HO] Yes, Justice Alito. [JUSTICE ALITO] Okay. And so then this is territory that we've... we've covered, but, if the Secretary is told here's the error rate that we can expect for those who answer the citizenship question, and on the other hand, we have this model and we can't tell you how accurate it is, but trust us, it's going to be better, is it arbitrary and capricious for the Secretary to say, I don't want to go with this model because I don't know what the accuracy of that is? [MR. HO] Justice Alito, respectfully, I think the Census Bureau said a little bit more than trust us. What the Census Bureau said was we can develop a highly accurate model for this that's going to be better than getting the question wrong one-third of the time, which is what... [JUSTICE ALITO] Yeah, well, they said, in our opinion, this would be better, but they can't quantify it. They... they don't provide a specific number; they don't even provide a range. Am I right on that? [MR. HO] They do say that it would be more accurate than responses to the citizenship question, which they do quantify as being incorrect one-third of the time for non-citizens. And if I could get back to... [CHIEF JUSTICE ROBERTS] But it is a... [MR. HO] ... Justice Kavanaugh's... [CHIEF JUSTICE ROBERTS] Go... go ahead. [MR. HO] Oh, I'm sorry. Justice Kavanaugh's question earlier about whether or not that can help with Voting Rights Act enforcement, it can't. And... and here's why: Citizenship data matters in the Voting Rights Act. [CHIEF JUSTICE ROBERTS] I'm sorry, just to... what can't? [MR. HO] I'm sorry. [CHIEF JUSTICE ROBERTS] I lost sight of the "it" in your answer. [MR. HO] The Secretary's question, Mr. Chief... the Secretary's decision, Mr. Chief Justice. And responses... [CHIEF JUSTICE ROBERTS] I thought you were talking about whether it's helpful with respect to the voting rights information. [MR. HO] That's right. Adding a citizenship question to the census, I'm sorry, is not helpful for Voting Rights Act purposes because responses to the question are inaccurate so frequently for non-citizens. Citizenship matters in the Voting Rights Act context when you're dealing with a population in which there's a large number of non-citizens. The VRA requires the drawing of districts in which minority voters constitute a majority sometimes under some circumstances. Now, under normal circumstances, voting age population data will be sufficient for that purpose if citizenship rates are high. But, if the minority group has relatively low citizenship rates, for example, as is the case with Hispanic populations in some circumstances, then you need citizenship data to make sure that you're drawing a district in which minority voters are, in fact, a majority of the population. And data that's wrong one-third of the... the time with respect to non-citizens just doesn't help you draw districts... [JUSTICE GORSUCH] Well, when we... [MR. HO] ... at that granular block-by-block level. [JUSTICE GORSUCH] Well, when we talk about the block-by-block level, one of the complaints that we've heard from the other side is that the data that we rely on from the ACS is at too high a level and that the census goes down to a more granular level. And, in fact, some of the states who are now Respondents before us have in litigation, including in this Court, argued that ACS data should not be relied upon for purposes of citizenship or other purposes, that the census data is more accurate. What do we do about that? It seems to me like you kind of put the government in a bit of a Catch 22. You say they shouldn't use the census, except for in later litigation when they have to use the census. [MR. HO] Justice Gorsuch, let me say two things in response to that. The first is that, to the extent that more granular citizenship data were, in fact, necessary for Voting Rights Act enforcement purposes, and we, I think, set forth a number of reasons in our brief why that's not, in fact, the case, but just assuming that it is, the... [JUSTICE GORSUCH] Well, we... we know states have argued this, including some of the Respondents before us. So I'm... you know, it is a thing. [MR. HO] Okay, that... that's fair. But what the Census Bureau recommended was that it could develop that block-level data either with existing ACS data or using administrative records and that that would be, in fact, the best and most accurate way to do that. And... [JUSTICE GORSUCH] So... so the states that said previously that wasn't enough now are going in all future litigation to bind themselves to accept that it is enough? [MR. HO] Well... [JUSTICE GORSUCH] Are you prepared to say that? [MR. HO] ... Justice Gorsuch, we've never taken... our clients have never taken that position, and I... I'm not aware of my organization ever taking that position in litigation. [JUSTICE GORSUCH] And how... how about the underreporting or the folks who stop and break off answering the long form and... and we're asked to believe that that's solely attributable to this question? We have a whole bunch of states that say that, in fact, the break-off rate because of that question, at that question, is something like 0.36 percent. [MR. HO] Well... [JUSTICE GORSUCH] So that it's very difficult to understand why that question would be the cause of people stopping answering, whereas another possible explanation that hasn't been explored, as I understand it at least, is the length of the form itself may deter those with less means and less time to fill them out, just as simple as that, and we don't know. And what do we do with the fact that we don't know? [MR. HO] Justice Gorsuch, the Census Bureau's conclusion was that the most likely explanation was the citizenship question. The only difference in that comparative estimate was the presence of a non-citizen in a household, and citizenship is obviously the most salient question that goes to the difference between those two populations. And the number on the break-off rates for the Internet ACS survey, which I believe Your Honor was referring to, they showed that Hispanics were actually eight times as likely to break off in responding to the ACS upon encountering the citizenship question. Now, by contrast, there isn't a shred of evidence in the administrative record that suggests that this question will not have the effect of harming response rates or will actually improve the citizenship data provided to the Department of Justice. But if I could make one other point in response to your earlier question, Justice Gorsuch, adding the citizenship question doesn't even solve that granularity problem that you referenced, and here's why: Because the Census Bureau can only produce estimates of citizenship at the block level. The government has now conceded that on page 18 of their reply brief, which is quite remarkable, because the government's rationale for asking this question has been to provide a full count of citizenship. And because of the Census Bureau's disclosure avoidance protocols, it actually can't do that at the block level. It undermines the whole rationale for adding this question, and the Secretary didn't even address it in his decisional memo, which renders his decision arbitrary and capricious under State Farm. [JUSTICE KAVANAUGH] It seems to me... [JUSTICE KAGAN] So... [JUSTICE KAVANAUGH] Go ahead. [JUSTICE KAGAN] ... I was just... what is the "it" in that sentence? What did the Secretary not address? [MR. HO] He did not address the fact that because of the bureau's disclosure avoidance protocols, it can only provide estimates of citizenship at the block level. If I could, let me explain why. The statute requires the Census Bureau not to disclose information that could result in the identification of a person's census responses. If you have 100 people living on a block and the Census Bureau says, well, there are 100 citizens there, you will have necessarily identified all of their census responses. So what the bureau does is it alters demographic totals for census blocks before publishing them. That means that that data is an approximation, it's an estimate, just like the ACS data that the Department of Justice currently relies on. And here's what's critical. [JUSTICE ALITO] Well, this is... this gets really, really technical, but... well, and your... I'm sorry. [CHIEF JUSTICE ROBERTS] No, go ahead. [JUSTICE ALITO] That's... that's fine. [CHIEF JUSTICE ROBERTS] Okay. Thank you, counsel. [MR. HO] Thank you, Your Honors. [CHIEF JUSTICE ROBERTS] Mr. Letter.

[MR. LETTER] Mr. Chief Justice, and may it please the Court: I just want to say right upfront the Speaker of the House wishes to thank the Justices for their courtesy in hearing from the House today. [CHIEF JUSTICE ROBERTS] Tell her she's welcome. [LAUGHTER] [MR. LETTER] Thank you. I'll pass that along to her, Mr. Chief Justice. I want to hit just a couple of points, but one of the ones I want to hit right upfront is something that... that General Underwood said and I think bears some emphasis, which is the... remember that the... the census that we're talking about here is the decennial census provided for in the Constitution, of utmost importance to the House of Representatives. That provision obviously is the ground... has to be the grounding for the statute that is being applied here. And so anything that undermines the accuracy of the actual enumeration is immediately a problem. So there's been a lot of discussion here, quite properly, because of the way this case has been briefed, about will this help the Justice Department and the Voting Rights Act, et cetera. And that may be a very important point, but it is not why the Census Bureau carries out an actual enumeration, which goes to the apportionment of representatives among the states and then distribution within the states. So, if there is something that undermines the accuracy of that count, even if it's important for other reasons, that is both a statutory violation and, therefore, a violation of the Administrative Procedure Act, and a constitutional violation. Now this Court does not have to reach the constitutional question because it is a statutory violation. I... [JUSTICE ALITO] But do you think that any decrease in the actual count, if... if you add any question beyond counting people, and that decreases the actual count to any degree, then that additional question is improper? [MR. LETTER] Justice Alito, I... I'm sure that... that the Court would find there's a de minimis exception. There's no... no doubt about that. So where this Court would draw that line, I don't know. What I... I can tell you, and I'm... I'm sure you know this, but I'll just... from the... this Court said in the Wisconsin case that... a question there was could a statistical adjustment be made, and this Court set the standard of what actual enumeration means and it says "a reasonable relationship to the accomplishment of an actual enumeration." And this discussion about the importance of voting rights data obviously does not bear a reasonable relationship to the accomplishment of an actual... [JUSTICE GORSUCH] Mr. Letter, I'm sure you've given this some thought, but... I know you have. In terms of assessing what a reasonable relationship is, what do we do with the history and the fact that this question has been on for what a long time was the only form in the census through almost all of our history, and it continues to be asked today in the long form or in the ACS. It's not like this question or anybody in the room is suggesting that the question is improper to ask in some way, shape, or form. And what we do as well with the evidence of practice around the world and virtually every English-speaking country and a great many others besides ask this question in their censuses? So I'm sure you've given consideration... thought... thoughtful consideration to those questions. [MR. LETTER] Absolutely, Your Honor, although I can tell that you also have. First of all, I don't know if the other countries that are listed, for instance, in the... the U.N. recommendations have an actual Enumeration Clause written into their constitution that is of paramount importance. So I'm not sure that, when... when the U.N. made that recommendation, that that matters for the United States. Second, Your Honor, the... if... there... there are other factors that would undermine actual enumeration. There is no evidence in the record here. As the Chief Justice pointed out, we're dealing with a record here. It may be that some people find questions about gender now offensive or maybe in the future that will be deemed offensive and that would undermine the accuracy of the... of the actual enumeration. We don't have any evidence on that. What we do know, Your Honor, as you've quite correctly pointed out, we have a history of this, but what we do know now is the experts right now say that this question, if it is put on the... the form, which, remember, is the only form right now for the actual enumeration, that will cause... that will make the undercount worse. [JUSTICE GINSBURG] Mr. Letter, the... Congress has the primary control over what the census will be, not the executive, and Congress has been alerted to this citizenship question for some time, and it has done nothing about it. So one question is, who should decide? Congress is silent. Should the Court then step in? [MR. LETTER] It's a very fair question, Your Honor. Two responses. One, I think that this is a very ironic point for General Francisco to be making. He has said... he's emphasized in his brief Congress knows about this, Congress should do something. This... the Court can take judicial notice of this because it's in the public record. The Secretary of Commerce has been called before Congress to explain what he did here, and Assistant Attorney General Gore, the one, you know, about the... the author of the... the request by the Justice Department, has been called to Congress. They have been declining to answer. They're not giving Congress the information it requests because they say there's litigation going on. And, I repeat, this is a matter of public record. So it's ironic for General Francisco to be saying: This is for Congress. Well, if that's for Congress, obviously, the House needs the information. And yet we're being told we can't have the information because it's... because it's only for you. [CHIEF JUSTICE ROBERTS] I thought all the information available, as I understand it, leads to only one answer, and so why isn't that answer sufficient for them to take whatever action they consider appropriate? [MR. LETTER] The... I'm sorry, Chief Justice. [CHIEF JUSTICE ROBERTS] Well, we've been told there was no basis for the Secretary to make any decision, other than the recommendation that was submitted to him by the bureau, because that's the evidence. That's the scientific evidence. And so there's no room for the exercise of any discretion. So what information... what more information does the Congress need to address the problem? [MR. LETTER] We want to know what... you... you decided otherwise. Why did you decide? As we know, his... his... his letter provides not... his memo provides not much information. This... the Justices here today have been asking these key questions. So we want to know, what made you do... what made you decide this? Was this just a political decision? [JUSTICE KAVANAUGH] Well, I thought Justice Ginsburg's question went more to why doesn't Congress prohibit the asking of a citizenship question in the same way that Congress has explicitly provided that no one can be compelled to provide religious information? [MR. LETTER] Right. And so that is something that Congress could attempt to do, yes, and that is one of the things that would be... would be asked about. But, as we know, that doesn't stop this Court from interpreting the statute and the Constitution. As we know, this Court is the final word on the Constitution. [JUSTICE KAVANAUGH] On the... on the statute, I had one question, which is you make a good first principles point about enumeration being the purpose of the census. But it turns out that the census, as you know, has been used for lots of other statistical and demographic collection purposes throughout... throughout our history. So it's not just for enumeration. And the statute that Congress has passed gives huge discretion to the Secretary how to fill out the form, what to put on the form. So how are we to think about enumeration when the history and the statute suggests that there is more than just enumeration that's at stake here? [MR. LETTER] Your Honor is exactly right. The... the census information, the census data, are used for all sorts of things that are very important. Remember, the Census Bureau does things way beyond just the decennial census actual enumeration. But, again, this Court was very clear, and... and, by the way, the other thing is General Francisco has argued no review. This Court has reviewed how the actual enumeration has taken place, I think, about five times. So there's clearly judicial review here. [JUSTICE KAVANAUGH] Well, assuming there is review and assuming it's arbitrary and capricious, as you know, it's deferential. [MR. LETTER] Yes. [JUSTICE KAVANAUGH] And the question I think here is a policy judgment that it's more important to get accurate citizenship information even at the expense, potentially, of a slight decrease, potentially, in response rates. And the question is: Given the statutes, why does that judgment fall below the standard of reasonableness in assessing the different policy considerations? [MR. LETTER] Our position is that, one, the Justice Department can get this information elsewhere, as we know. But, two, you can't undermine the accuracy of the actual enumeration in order to get information... [JUSTICE KAVANAUGH] So... [MR. LETTER] ... for the Voting Rights Act. [JUSTICE KAVANAUGH] ... the constitutional backdrop, I think, if I'm translating your argument, means that anything that would undermine the enumeration is impermissible and unreasonable? [MR. LETTER] I believe so, Your Honor. And the only thing... the only addition I would make to that is, as I said to Justice Alito, undoubtedly, there's a de minimis determination. But, again, this Court is... this Court is the expert on the Constitution. Thank you, Your Honor. [CHIEF JUSTICE ROBERTS] Thank you, counsel. Four minutes, General Francisco.

[GENERAL FRANCISCO] Mr. Chief Justice, thank you. I'd like to make four points: First, on the disclosure avoidance protocols, we discussed those at our reply brief at page 17. The bottom line is that those protocols apply to all census data, including on sex, on age, on race, that the Department of Justice uses to construct citizen voting age population. It's never been a problem before. There's no reason to think it'll be a problem now. And I... in fact, I think in the trial record, Dr. Abowd testified how it would not be a problem. Second, I'd like to point Your Honors to the key differences memo at page 148 of the Joint Appendix, the very... the one full paragraph, the first sentence says, "The relative quality of Alternative C versus Alternative D will depend on the relative importance of the errors in the administrative data, response data, and imputations." It then goes through and discusses the various types of errors. And its final conclusion is, "Unfortunately, the Census Bureau cannot quantify the relative magnitude of the errors across the alternatives." Third, in terms of response rates, the administrative record shows that the Census Bureau staff believed that there would be a 5.1 percent decrease in the initial response rates from adding the citizenship question. But, as Secretary Ross points out in his memorandum, that doesn't take into account follow-up operations. That's the response rate drop before follow-up operations. And it doesn't disaggregate between those who are going to be put off by the citizenship question itself, as opposed to those who are put off by the larger macro environment because they don't trust the government or don't like this particular administration. And that, I think, is one of the reasons why Dr. Abowd concluded in the testimony that Justice Breyer cited that "there's no credible quantitative evidence that the addition of a citizenship question"... [JUSTICE SOTOMAYOR] I... [GENERAL FRANCISCO] ... "will affect the accuracy of the count." And... [JUSTICE SOTOMAYOR] ... I thought that Dr. Abowd stated and the district court found that the follow-up process was, at best, riddled with a number of... of... of inadequacies and that it wouldn't be adequate enough to take care of the shortfall. [GENERAL FRANCISCO] Well, we can debate about whether it would be adequate enough, but in order to say that the final self-response rate would drop by 5.1 percent, you have to conclude that it's zero percent accurate. And I don't think that there's any basis in the record to conclude that it's... [JUSTICE SOTOMAYOR] So how much accuracy... [GENERAL FRANCISCO] ... zero percent accurate. So... [JUSTICE SOTOMAYOR] ... would be... let's assume it were 5.8 or something close to it. Is that de minimis to you? Is that... [GENERAL FRANCISCO] Your Honor, I think... [JUSTICE SOTOMAYOR] ... inconsequential? [GENERAL FRANCISCO] ... I think that that is largely an impossible question to answer. That is not built into the Constitution itself. There's always going to be a tradeoff. The long-form census, for example, caused a drop in self-response rates relative to... to the short form by, I believe, around percent. But my final point is one that Mr. Letter alluded to, and that is under... [JUSTICE SOTOMAYOR] That's why you keep... [GENERAL FRANCISCO] ... my friend's... [JUSTICE SOTOMAYOR] ... the short form. [GENERAL FRANCISCO] Right, Your Honor. And under my friend's on the other side's position, you are effectively empowering any group in the country to knock off any question on the census if they simply get together and boycott it. There are many people in this country who might find the sex question objectionable because it limits individuals to a binary choice. If a large number of people got together in one state and said we're going to boycott the census... [JUSTICE SOTOMAYOR] General... [GENERAL FRANCISCO] ... as long as you include... [JUSTICE SOTOMAYOR] General... [GENERAL FRANCISCO] ... that sex question, you're effectively empowering... [JUSTICE SOTOMAYOR] General, are you... [GENERAL FRANCISCO] ... them to knock that off... [JUSTICE SOTOMAYOR] ... suggesting... [CHIEF JUSTICE ROBERTS] Justice Sotomayor. [JUSTICE SOTOMAYOR] Are you suggesting... are you suggesting that Hispanics are boycotting the census, that they... [GENERAL FRANCISCO] Not... [JUSTICE SOTOMAYOR] Are you suggesting they don't have, whether it's rational or not, that they don't have a legitimate fear? [GENERAL FRANCISCO] Not in the slightest, Your Honor. I am suggesting that the risk... [JUSTICE SOTOMAYOR] So... [GENERAL FRANCISCO] ... of my friend's theory on the other side is that it countenances precisely that type of coordinated behavior that would empower groups to knock off any question of the census that they found to be particularly objectionable. Mr. Chief Justice, unless the Court has further questions? [CHIEF JUSTICE ROBERTS] We're all done. [GENERAL FRANCISCO] Thank you, Your Honor. [CHIEF JUSTICE ROBERTS] Thank you, General. The case is submitted.