How many noncitizens are on U.S. voter rolls?

President Trump has a habit of sowing doubt about America's elections. Now, the administration is mobilizing government agencies to make wild claims about massive numbers of non-citizens voting. It's the NPR Politics Podcast. I'm Tamer Keith. I cover politics. I'm Miles Parks. I cover voting. I'm Jude Joppy Block. I cover the intersection of government and personal data. And I mentioned a wild claim. Here is one. The US Department of Homeland Security says 250,000 non-citizens are on the voting roles in just four states. Miles, I want to do some basic fact checking here. Where is this number coming from? I wish I could tell you. Honestly, and I think the four states who are accused of the Department of Homeland Security having all these hundreds of thousands of non-citizens on their voter rolls wish that they knew more about it as well, which is a big part of what we've been reporting on the last couple weeks. You know, there was this speech when this was announced was this nationally televised speech by President Trump about a little over a month ago now, where he unveiled what was supposed to be an election security bombshell. And it really was the culmination of two years of the president has been putting people closely associated to his causes and to the idea, frankly, of pushing misinformation about elections into different positions of power, where we were expecting numbers like this for the last two years, this is the first time that we saw where the Department of Homeland Security was saying. They had this huge proof of this, this kind of bombshell report. And yet, even weeks later, there's still just very little information we know about where these numbers came from or how many of them are actually non-citizens on the voter rolls. Yeah, I know that when the President gave that speech, immediately afterwards, Miles, you were talking about how it is very likely that these numbers are going to come down. Right, I mean, this is a tale as all of this time, frankly, with voter rolls, is that when you do any sort of analysis, I will say data matching in the election space is notoriously fraught, and so what almost always happens is somebody decides that they're going to look for XYZ on the voter rolls, and they do some sort of analysis, and they come out with an answer that's like, well, we found six million dead voters on the rolls, and then once you actually dig into those numbers, and you do investigate, you know, and make sure your methodology is accurate in that you're matching on the correct criteria. Full social security numbers, for instance, are a good way to match voter to voter and make sure you're not going to get a bunch false positives. Once you do that analysis, we see it in red states and blue states. Every single time you see a giant number get whittled down and whittled down and whittled down, which almost certainly is what would happen with this number, though we don't really know again exactly where it came from. Jude, do you have any information about the methodology that the Department of Homeland Security used? Did they describe that? Do you have any hints? Well, we didn't up until we did a public records request for a meeting that Nevada Secretary of State officials had with DHS, and this was the first meeting they met virtually with officials from the Department of Homeland Security to talk about this data, and we got a copy of that video meeting. And they talked through Nevada's questions, including questions about methodology. In that call, DHS did give more information about what they did to arrive at the number that they publicized for Nevada, which was nearly 16,000 non citizens that they were alleging were on Nevada's roles. In that call, they made clear that they had used Nevada's publicly available voter roll and that they had then tried to match voters to US citizenship and immigration services data, and so they were using their own data to try to see if they could match things like name, address, date of birth, phone number to the publicly available voter roll. And in that call, what was really interesting was that even though this nearly 16,000 number got publicized, they acknowledged that was a ceiling that they had not yet really drilled down on those numbers and that they plan to do a manual review of each one to really find out if it was the same person, they acknowledged that there's things like a junior and a senior that live at the same address, but they're not the same person, and they might have different immigration statuses. Things like that that make this data messy and hard to decipher, and so they were acknowledging this is a ceiling, it's not the actual final number, and after that conversation, there was follow up where Nevada was asking to review what they had so far, and so far, DHS has sent them information for just 185 alleged non citizens voter IDs for them to look at. Now, DHS still says they have a lot of work left to do, they they mentioned in their correspondence that they have like 14,000 left to drill down on, but what's the takeaway here is really that these numbers are in no way final, and it gives it gives an indication of what the willingness to put out information that's not yet final as well to the public. Yeah, I mean, that is a remarkable distance. We should say that the New York Times first reported this story, but you two matched it and and got audio of this call, which is fascinating. Well, it was interesting to hear the back and forth. Also, you know, the person at the Department of Homeland Security, who is running this operation, is a woman named Heather Honey. We've mentioned her on podcasts before she's a long history in the election denial space before President Trump brought her on to work on election integrity issues at the Department of Homeland Security, and I just want to play a little section of her talking to the Nevada officials here she is. We are an open up here. We are, you know, and acting in very good faith, you know, and we just want to help you. And I thought that was so striking to hear her. She's basically convincing them we are acting in good faith. We're an open book. We're going to be transparent. But then Nevada, even at the smaller list of voters, these 185 voters, Nevada then followed up and said, okay, totally we want to investigate these. Can you tell us how you came to this number? And they said, no, we will not show our work at all. We're just going to give you this list and expect you to work off of it. And so we also heard from officials in Pennsylvania and California to other states that were mentioned in this huge announcement. And they said they had similar interactions with the Department of Homeland Security that basically they asked basic questions about the methodology. And they wanted a detailed list of voters and how they came to these conclusions. And they have not gotten that yet from DHS. It is truly remarkable here that federal officials would go out and make these public announcements and a ledge, massive fraud, announce it to the American people. And then when pressed by state authorities, really not have much to back it up. Miles, what's going on there? I mean, if you think about it this way, this is something that David Becker who runs the Center for Election Innovation and Research talks about a lot. Which is that President Trump has had the full weight of the federal government behind him for two years now. And he still has not been able to find the proof of the fraud that he has been insisting occurred for six years now. And so I think the thing to remember, I think pushing this a little bit forward is that this is also not the end of this sort of data. And I say, I'm doing the quotation marks this data analysis that the day after Trump's speech, Mark Wayne Mullen, the Secretary of Homeland Security promised that they were going to do more of this sort of analyses and release it around the election. Let's listen to that. We will pursue maximum pressure on this to let you know we will be proactively looking at early voting. And then after post election, we will continue to scrub all those that did vote. And so we're I think essentially working with the assumption that we are going to get more of these giant numbers. I think this is not the end, I guess, is what I would say. Right. And we're actually already seeing that there's a strategy of just putting out a lot of numbers about allegations of non-citizens voting. So even beyond the DHS allegations in the four states, just last week, the census put out a report, a highly unusual, unauthored report. Our colleague Hansi Loh Wong was the one who led the reporting on this. And this report is alleging that that about 24,000 non-citizens voted in the 2020 election based on an analysis of 128 million voter records. So it's still a very tiny percentage. But on social media, Trump claimed the report proves that over 24,000 non-citizens voted illegally in 2020 and that the final number is going to explode. And what's really interesting about this is our reporting found that this analysis was not done by career civil servants at census, that the team behind it included individuals affiliated with America First Policy Institute. That's a think tank that was started by people who were part of Trump's first administration. So this is a very unusual report that former census officials have raised some concerns about voting experts are really raising questions about these numbers. Which reminded me a lot, Jude, of what we were talking about earlier, of Heather Honey leading this other investigative effort at DHS. And then you've got this Trump-aligned think tank working on this census report, where I think what I was hearing from election officials a lot the last two years, is that they felt that people who had predetermined outcomes. We're going to basically investigate with a predetermined outcome and so I think we are already starting to see that. And I believe the president also in that post said this proves that I won in 2020. Which it didn't. I mean, that was the other striking thing is that if you actually even just open- this report, which, again, the methodology, very murky, unclear how they came to this conclusion, but the actual number of non-citizen voters on the roles that that report found were not enough to even overturn the 2020 election, which I thought was striking. - All right, well, we're gonna take a quick break and we will be back with more in a moment. Welcome back and a quick note, we wanna hear your questions about the midterms coming up. And when I say here, I mean, please send us a voice memo with your name, where you're calling from, and what your question is to [email protected], that's [email protected]. We are not in the business of making predictions of who will win this race or that, but we wanna know what you wanna know. - I know, I feel like the sillier the better also. I mean, I'm happy to answer serious voting rights questions as well, but if you wanna ask, like, if all the 2028 candidates were carnival food, for instance, what carnival food would they be, I would take that. I'm just thinking of, like, what kind of, I'm just, no, too far. - I'm sorry, no. Ask us your questions about national trends and politics and things like that. - I think the people wanna know who's a funnel cake. - Oh, I'm definitely a funnel cake. Okay, if you send us a memo, we might use it on an upcoming show. So in the first half of the pod, there were a lot of numbers. And I'm so confused. First, it's 250,000 people on the voter rolls in four states. Then it's 24,000 people who voted. Can you just tell me what these numbers have to do with each other or don't have to do with each other or how this makes any sense? - It's a little hard to know because we don't know where all the numbers came from, right? But I do think that what you can say confidently is that every time a state or government entity has looked at the idea of non-citizen voting, the numbers are minuscule. They are not zero. And so it is not out of the realm of possibility of taking the census report, for instance, which alleged more than 20,000 non-citizen voters out of 128 million. I think that number is closer to the ballpark of what experts expect to the number of non-citizen voters to be not exactly, but in terms of being much fewer than 1%, probably fewer than 1/10% is somewhere in the ballpark, but we don't know exactly how these numbers interact with each other because we don't know the methodology. - So just to be clear, guys, you asked both the Department of Homeland Security and the Census Bureau about their methodology. What did they say? - Yeah, so we asked both census and DHS about the methodology for the numbers they were presenting, but we didn't get answers from either one. - Alrighty then. So Miles, I want to talk about where this might be headed. You talk to voting officials a lot. What are they worried about when it comes to all of these numbers that the Trump administration is putting out? - I don't think they know exactly what they're worried about, frankly, is that they're worried about, they're preparing for a whole range of different outcomes this election season, but they do all kind of go back to what I keep referring to as the predicate, which is something that Steve Bannon said, leaning up to the Trump speech a few weeks ago, which is this idea that whatever options that the federal government wants to use to potentially interfere in the elections process, there has to be the sense that there is something wrong, right? That there is some sort of widespread fraud. And so these numbers, I think, the fear of election officials is that these numbers continuing to release different reports that have things that were traditionally credible sources of information attached to them that they could be then used to lend credibility to whether it's, there's been talk about a potential national elections emergency, that President Trump could implement legal experts or very dubious that he would actually be able to take control of any elections infrastructure. But the one specific fear that I heard about from one state election official in detail was this idea that these numbers could be used to get a warrant to confiscate ballots or election materials in the time right after election day. This is a fear that popped up after the FBI raided the elections office in Fulton County, Georgia earlier this year, took possession of ballots and election equipment from the 2020 election. And then it turns out that the search warrant that justified that raid was based almost exclusively on misleading or misinformation about the 2020 election. So this election official was saying that they were very worried. If you were able to find a friendly judge and you could use this 250,000 alleged non-citizen voters from the Department of Homeland Security, then you could potentially find a judge who would allow you to confiscate ballots or election materials in the time after election day, which could really, really potentially mock up the process. - Make it very hard to know what the original ballots had said. And so I think also with this, though, it's important to keep in mind, though, is that it is a caveat that if you can find a friendly judge who would be supportive because at the end of the day, there is a high bar for these warrants. And the Georgia case did raise a lot of awareness about the issue. And so it was almost like a practice run. So there are people in the election space that are hopeful that it would be pretty hard to do that strategy again. - And this happens in the context of midterm elections that are coming. People will be voting in November. History would tell us that the party in power, the president's party is likely to lose seats in the House, possibly in the Senate as well, could lose control of one or both houses of Congress. And that would change the course of the presidency, change the course of his second term. And it's something he is making very clear he wants to avoid. - Absolutely. I mean, the stakes are incredibly high for the president in this midterm election. I do think, to Jude's point about the potential of election interference, there's a lot of legal experts who have a lot of opinions on what's likely, what's likely to work. And then I also think it's worth noting that a lot of the Doomsday scenarios also rely on a really close election. I think if this is an election where let's just say, potentially Democrats win a 30 or 40 seat majority, the experts and officials I talk to think it's less likely that some sort of one of these Doomsday scenarios comes to pass versus if the House has decided by one seat that happens to be in California or in Texas, then the likelihood that either party would try to do everything possible to win that one seat becomes higher. - And the other thing to point out is that individual races are often quite close when it comes to Congress or like state house or we get down to mayor or city council and they can be extremely close. - Exactly. And I think that's what's kind of funny about the 2020 election in President Trump's fixation on it is that the states that he lost, he did not lose by 18 votes or 25 votes which are traditionally the margins where you can argue about, well, if those 10 ballots weren't dropped off by an unauthorized person, whereas like losing by more than 11,000 votes in Georgia is a little bit harder to argue, whereas if you do lose, let's just play that scenario out. If you lose this one house race by 31 votes or something like that, then there becomes a lot of different ways that you could contest that election. - All right, we are gonna leave it there for today. I'm Tamar Keith, I cover politics. I'm Miles Parks, I cover voting. - And I'm Jude Jofy Black. - And thank you for listening to the NPR Politics podcast. (upbeat music)