Election Security, with Dr. Charles Stewart (Episode 20)

John Ellis talks with MIT political scientist Charles Stewart III about the mechanics, vulnerabilities and public perception of American elections. They revisit the disputed 2000 contest, the administration of the 2020 vote, and the reasons false claims of fraud gained traction despite extensive verification. Stewart explains why machine counting is generally more reliable than hand counts, why California takes longer to certify results, and how election officials measure performance. The conversation then turns to the 2026 midterms, including federal efforts involving voter files, mail ballots, and identification requirements. Stewart warns that close races could produce widespread challenges, pressure on local officials and disputes over seating members of Congress. He also discusses AI, election funding, and why he remains confident in the country’s fact-based institutions today.

News-Items.com

Hosted by John Ellis

Produced by Dale Eisinger

Transcript

SPEAKER_00 0:00

Hello and welcome back to the News Items Podcast. I'm John Ellis. I am the founder and editor of two Substack newsletters. One is called News Items, the other is called Political News Items, and you can find them both at news-items.com. My guest today is Charles Stewart III. Mr. Stewart is the Keenan Sahen, Distinguished Professor of Political Science at MIT, where he has taught since 1985. He's a fellow of the American Academy of Arts and Sciences. His research and teaching areas include American politics, congressional politics, elections, and American political development. Professor Stewart is an established leader in the analysis of the performance of election systems and the quantitative assessment of election performance. Since 2001, Professor Stewart has been a member of the Caltech MIT Voting Technology Project, a leading research effort that applies scientific analysis to questions about election technology, election administration, and election reform. I could read the rest of his CV, which is impossibly impressive, but I won't. I just uh we'll cut right to the chase. And thank you very much for joining us, Professor.

SPEAKER_01 1:13

Oh, it's my it's my pleasure, and um I'm glad you didn't keep going.

SPEAKER_00 1:18

I want to talk about three elections. Uh 2000 election, which featured hanging Chads, uh the 2020 election, which sort of blew up under sustained assault by President Trump and those loyal to him, and then the upcoming 2026 elections, midterm elections, which have created their own controversy, I guess, or controversy has been created about them. Uh but going back to the 2000 election, I don't recall the the charge of the you know the election was stolen or the election was rigged or any of that. I I remember it as being just exceptionally close, and it came down to that weird thing where we were looking at hanging chats. Is that is that correct?

SPEAKER_01 2:04

It's you know, that's a I'd say it's like 75% correct, and a part of the maybe memory has been clouded by subsequent events, right? You know, going into 2000, that was a close election. And I think one reading of it could very well have been, well, this is a close election, and it was a coin toss. And, you know, so there were hanging chads and the rest. It could have been something else, but it was it was close. I think in retrospect, it became something else for elements of the political universe. First of all, for Democrats with the Bush v. Gore decision, it became a marker for the politicization of the court and a question about the legitimacy of the outcome. And you'll probably recall the bumper stickers that went around about the time re-elect Gore in 24 or in 2004, right? So there was a bit of this kind of election denial illegitimacy that got that got salted in there. Furthermore, there started to be a narrative, mostly on the left, that the voting machines were raped. And I would say, in fairness, 2000 awakened America to the understanding that voting technology could be consequential, could break down sometimes, and we needed to do better in it. But then that that shone a spotlight on the process and anything having to do with election technology. So you may recall also that going into the 2004 election, the president of Deebold, which was the manufacturer of one of the largest manufacturers of voting machines, said this, you know, really unfortunate thing. Um he was, I guess, the the chair of Bush's re-election committee in Ohio, right? And he said some inpolitic things about, you know, kind of counting the votes in Ohio, which was interpreted as being proof that Dee Bold was going to rig the election. So, you know, there's been elements of this kind of the system is rigged since 2000. I think what has changed over time is that politics continues to be close. And that was the politically important thing about 2000, I think, is that represented really the tightening of electoral competition. So that elements of election administration now are under the microscope. And if you lose, there's a tendency for people who lose to blame the process when they lose. And man oh man, that's so easy to do nowadays. Yeah, if you if you want to be a sore loser.

SPEAKER_00 4:39

It's interesting because in 2004, Robert F. Kennedy Jr., no less, wrote an article for Rolling Stone magazine in which he asserted, yes, that uh that the outcome in Ohio was fixed. And that became uh back in the days when Bobby Kennedy Jr. was a uh, I guess on the left. It it's weird that he's now the Secretary of Health and Human Services, but we'll move right on from that. You you've done quantitative research on on the accuracy of human counting, because one of the things is people say is, you know, we should just go back to counting the vote by hand. And your research shows that actually doing it by, quote, technology, end quote, leads to greater accuracy. And you point out that people get tired and they make mistakes and the machines can run, you know, more or less 24-7 and don't make mistakes. Is that, you know, w if that's the case and you you're, you know, you're you prove that it is in fact the case, what is it about the machines that is it just is it just that the machines can be manipulated by dark forces that that causes such you know agita about their use?

SPEAKER_01 5:52

To be flip about it, I mean I I think one problem is that too many people have watched TV shows in which the young 14-year-old hackers figure out how to get into the Pentagon and set off the missiles, right? I mean, I think in popular culture there is a sense that everything is connected to everything else and that it doesn't take a whole lot of effort, maybe just kind of, you know, even a 14-year-old can do it. And so I think that that is kind of the cultural context in which we live. So yeah, I mean, I I don't think it's any anything more, anything more than that. And so then when you get into something as complicated as running elections, it's hard to keep kind of all the pieces together. There are parts of the election system that are connected to the internet that are, you know, a big threat surface and have been have been penetrated from time to time. Voting, you know, voting roll systems, election night reporting, there are cases where they've been hacked, if you will. That's separate from the voting machines. Um and then, you know, we can go down all sorts of rabbit holes, but then it kind of gets into yes, there are vulnerabilities, but what's the opportunity to exploit those vulnerabilities? You know, if I left my if I if like I left my wallet in my office while I go to get a cup of coffee, somebody could come in and steal that wallet. If I lock my door, someone could pick the wallet, pick the lock. But what's the chance and the opportunities and all, you know, and what's the payoff to doing that? Very, very low. And I think that's kind of the thing that's hard to all fit together. Is first of all, what are really the opportunities out there? And then secondly, how likely is it for nefarious actors to overcome both physical and other types of protections against these systems? It's a complicated system, and it's easy to overlook the degree to which the system is protected.

SPEAKER_00 7:45

Aaron Powell When you look at the administration of elections, what what are the key things that you're looking at? Uh what are the markers that you look for in a well-managed election as opposed to uh a less well-managed election?

SPEAKER_01 7:57

Aaron Powell Well, you know, those markers, by the way, are international markers. It's not just what we look at in the United States. But, you know, we can think of kind of three big buckets of things. There's registering voters, there's capturing the vote, and then there's tallying and verifying the vote. And so you look for people having access to voter registration and systems that keep those registration roles accurate and clean. When you look at capturing votes, you look at actually a wide variety of things. In the US, this is really complicated because we insist upon some people being able to vote on election day, some people being able to vote in person before election day, and some people voting by mail. And each of those have a set of security protocols associated with them. And we can talk about those. But it involves, you know, kind of making the long list, and this is what election officials do a lot of. They have a lot of playbooks and checklists, like airline pilots do. And you have to have systems to verify that that voter is the one they claim to be, that the that the vote is accurately captured on a ballot or electronic device, and that the scanner that scans the device, which is gonna be for 98% of votes, that it accurately records the device, and there's mechanisms to ensure that. And then on the counting side and the verification side, that's more of a, I would say, a procedural concern. Are there checks and audits in the same that we have way we have financial audits during the Canvas period? Do you basically do the double entry bookkeeping, the equivalent of double entry bookkeeping that you do in finances? And at the end of the day, when there are disputes or challenges to elections, do those disputes get resolved based on the facts, or do they get resolved based on politics or other, you know, non-fact-based systems? And we can kind of go through any particular election and judge any particular election using those criteria. And some elections do better in the U.S. and some do worse. Um I think on the whole, you take a random election for anything in the United States, it it runs it it runs well enough that the that the right winner is called at the end.

SPEAKER_00 10:06

One th question that a lot of people have asked me, this is a sidebar, but one question a lot of people ask me, and I've I've never had a proper response or a uh well-informed response, and I'm hoping I can get one from you. Why does it take so long to count the vote in California?

SPEAKER_01 10:23

Some of which are, I would say, a matter of choice, and some of which are a consequence of all the things they do in California that are not done elsewhere. Okay. And I have to actually quickly say California has a one-month period for certifying a let the election. But if you judge it by how how long does it take California to count all the ballots to the point that you know who won the election, it's about a week. And that's longer than other states, but it's not a month. The whole month period is a mop-up operation in which, for instance, California has a very long period for quote unquote curing problems with mail ballots or registrations. So you have onesies and twosies of people coming in to verify that that's their signature or to show some ID, maybe that their ID was challenged, something like that. California gives a long time to clean up that stuff. Okay. So that's that's why it's a month rather than two weeks or even a week. California could certify their elections in a week. They could, right? The other thing that I this is where I all needle California. You know, if you if you give a county a month to certify it their election, there are some counties, not all of them, maybe in in but a small number of them that will say, well, got a month. You know, why should I why should I pay the overtime? Why should I do all these other things that cost money when I don't have to spend that? And so, you know, so and and to that degree, that's an administrative choice that some counties make. But I would say the large counties like Orange and LA and Santa Clara, they're counting as quickly as they can, and they're counting about as fast as other states. They rely a lot on mail ballots like Washington and Oregon and Colorado.

SPEAKER_00 12:15

Aaron Powell One thing about the 2020 election, and we we'll talk in some detail about this, but the thing that struck me about it the most was how well it was done given the pandemic, given, you know, the the just the disruption of of civic life that the pandemic caused. Is that accurate? Was it a well-run election, in your view?

SPEAKER_01 12:39

Yes, my colleague at Stanford Law School, Nate Persilian and I wrote a an essay for the Journal of Democracy after the election entitled The Miracle and Tragedy of 2020, of the 2020 election. And the miracle side was precisely what you're talking about that given everything facing election administration and the you know, everything facing election administration in 2020, I think it would be a finalist for well, certainly even before degrees of f of of difficulty, it would be a finalist for the best run election in American history. You add degrees of difficulty and it's you know it gets a 9.9 or something. I mean, it's um a really well-run election. Trevor Burrus, Jr.

SPEAKER_00 13:19

And yet we had from immediately, really, from Fox News calling Arizona at whenever they called it, 11 o'clock or 11.30 on election night, President Trump launched a, you know, it's rigged, it's it's it was stolen from me, et cetera, et cetera. Aaron Powell Why did that message resonate? Was it is it just that the politics are so polarized that people w on on the losing side, if you will, on the Republican side, on the MAGA side, whatever, w wanted to believe it? Or I mean given that it was one of the best run elections in American history, it seems surprising to me that President Trump's message that it was rigged uh resonated the way it did. Do you have a view on that? Aaron Powell Yeah.

SPEAKER_01 14:03

And I start actually by pulling away from 2020 and just thinking generally about voters in the election process. Look, I mean, voters don't walk around thinking actively about election administration. It's the last thing on their mind. They base their judgments on election administration from what they hear from trusted sources. And we have a system and a set of practices and norms in the United States such that people run for office and when they lose, when they lose fairly unsquarely, they give you concession speeches that are gracious and they move on. That's one thing. That did not happen this time. But the other thing that kind of made the election skepticism so powerful is that there's been some very good research done about expectations about the outcome of an election. And in 2020, in 2020, President Trump's supporters had very high expectations that he was going to win. Higher, slightly higher, but nonetheless higher than one normally sees going into an election. So they're primed for him to win. And so when he then, on Wednesday, rather than either conceding graciously or saying, who knows if we won, we're going to challenge this. I mean, there's kind of a normie way of saying, hold on to your horses, we're going to make sure every vote counts. He didn't do that. He confirmed what many of his, actually most of his supporters believed was that he had actually won. And there were kind of facts or factoids around that kind of confirmed that if you wanted to go along. One was as you mentioned Arizona. In retrospect, Fox News called Arizona before they should have. For sure. Right. And secondly, Pennsylvania was close. And anytime there's a close election, it's going to take a while. You're back to, okay, this is close. We got to, you know, make a list and check it twice, and we have to make sure that all the ballots are counted and counted their proper way. And it took until Saturday for Pennsylvania to finish their work and get, okay, we've counted them all, and this is this is the winner. So you get these two things going on. You have Arizona. My goodness, Arizona is a red state. What's going on? And look, it's getting more red. And then Pennsylvania, it's getting more blue. What's going on there? They're, you know, they're so you know, it's got normal human psychology running up against a violation, a really important violation of no legitimate concession speech. Compare what President Trump did in 2020 to what Al Gore did in 2000, who, one, you know, said, you know, during the counting, it stops now. And then when he was vice president, presiding over the certification, he shut down efforts to challenge the election. That's one scene, high contrast to 2020. And that's about norms. That's, you know, that's about norms.

SPEAKER_00 16:53

That is the major norm, really, in American politics, American governance, is that democracy depends on the consent of the loser, right?

SPEAKER_01 17:03

Aaron Powell Oh, absolutely. And I think that, you know, one of the things, kind of going back to 2000, one of the things that people in the election administration world, especially in the election kind of technology world, have worried about or thought a lot about is that how do we convince the losers that they've lost? And that's given rise to work, say, on auditing and other types of techniques in which you demonstrate. You don't just trust, right? Reagan, trust, but verify, right? You verify the outcome of elections. And I think one of the reasons why people like me are so confident about the outcome of the 2020 election is that those verification mechanisms were used. And so we don't have to trust that the elections were accurate. They were verified that the counts were accurate.

SPEAKER_00 17:53

You did major work, I guess, or or investigative work on the outcome in Pennsylvania. Can you tell us about that? Because that was, as you say, that was where Trump was ahead, Trump was ahead, Trump was ahead, and then Urban Vote from Philadelphia and less so in Pittsburgh turned it blue. And you looked at that fairly closely. Tell us about what you found.

SPEAKER_01 18:14

Well, a couple of things there. One, as I alluded to earlier, there was the matter of just getting all the votes in. And what we know from most states, and Pennsylvania is a good example, that that there's a thing called the blue shift that's been observed in elections for the last 20 years. And that this just refers to the tendency in most states to over time for the election results to kind of trend in a democratic direction as you look at the unofficial results coming in. And part of that is that cities like like Pittsburgh and Philadelphia, but particularly Philly, they're big, they're complicated, they take longer to count votes. It was complicated further in 2020 by the fact that voting by mail had become so politicized, and so that mail ballots, absentee ballots, were much more democratic than in-person election day ballots. And so you had and those took longer, and they took longer in Pennsylvania because of the rules they have about when you can start to process those ballots. So big city, a lot of mail ballots, heavily democratic. It's guaranteed to kind of supercharge a blue shift in a state like Pennsylvania, where if it's close, you know, it's gonna be consequential. You're not gonna notice the blue shift in in Massachusetts, for instance, or in California, right? But it's there, right? Nor are you gonna watch the blue ship the red shift in Washington because it's so democratic, it kind of doesn't matter, right? And so you're you're so so you're watching it. So that's that's really one of the things should that should also be said that that that we just pointed out in Pennsylvania, there were charges that, well, they were just dumping ballots in Pennsylvania, or in in Philadelphia rather. And um one important thing that people forget about Pennsylvania in 2020, and this continues to haunt the Democrats, is that they had a real problem turning out Philadelphia voters in 2020. There was a big, there was a big hit to turn out in Philly in 2020. Right. And so the other thing that kind of undercut this idea of just kind of, you know, dumping ballots in that city is that if there were ballot dumps, they were really bad about it because you know the number of ballots went down, they didn't go up.

SPEAKER_00 20:34

So one thing about the 2020 election that was good in terms of the outcome is that President Biden won the popular vote by a substantial margin. And if you go back to Gore, won, I think, by one million votes nationally. Clinton, I believe, won by three million votes nationally, Biden won by seven million votes nationally. If the outcome had been different, and you know, the 2020 election electorally, I think less than 100,000 votes in five states would have would have changed the outcome. Is that a is that going to be a problem for elections going forward where in you know in the in the aggregate, if you will, the blues win, but on the electoral system, the reds win. Is that is that something that you see as a future problem? Because it seems to me that that if you focus your campaign on just getting to 270, you can you can you know you can not worry about winning nationally, you can just just win the electoral system. Is that something that you think is going to be a major concern going forward or we don't know? I don't know if it's gonna be a major concern.

SPEAKER_01 21:43

I mean, I think it might be the t the pimple on the end of our nose in the sense that it's a kind of a flaw that many people see in the electoral system, but it's again not on not top of mind. And I mean there's there's enough people who will allude to what you just said is that okay here's the rules. Get to 270 and you win. Right. And there's one of the ways of doing that is just kind of, you know, distribute your resources in an optimal way. So you win 270 even if you lose the electoral electoral vote. And those are the rules of the game. And so I don't see a big a lot of energy around a true national popular vote. Interestingly enough, following the 68th election, we almost got a constitutional amendment to go to a national popular vote.

SPEAKER_00 22:29

Is that right? I didn't know that.

SPEAKER_01 22:31

Yes, yes. Democrats and Republicans alike, they were both freaked out by George Wallace, right? And that united them. And we don't have a third party to freak out the two major parties right now, which I think is actually probably the thing that's going to keep any sort of resolution to this issue off the table.

SPEAKER_00 22:50

Aaron Ross Powell Moving ahead I think a lot of people are confused about exactly what the Trump administration is doing to influence the outcome of this year's midterm elections. There's the Save Act, there's the Postal Service intervention, if you will can you give us sort of the three or four things that the administration is doing that seem unusual? I guess is the right word. Yeah I'll keep it to three or four. I just made a list of 40. We'll extend the length of the podcast and at hour two fifty we will we'll ask the next question. 40 I mean really you have a list of 40 really 40.

SPEAKER_01 23:29

And I have to say I've I've I've I've I've just as an aside I've categorized them into um kind of different sorts of things that we've observed. A lot of it's rhetorical and some of it's is but the last category is is crazy. And there's an adjective before that crazy that I I will spare your listeners. Please but crazy and which I think are low probability but alas we live in a world in which you have to write those things down. Right. Right. I think you know what we're observing, we can think about them in as kind of material efforts that are unusual that the Trump administration is trying to do. And then we can kind of think about the rhetorical effects of them or the strategic effects of them. But just materially the unusual things that they're trying to do are hoover up all of the voter files around America. They are trying to interject the Postal Serving voters and election officials to the point that the Postal Service will decide who is allowed to get a mail ballot. And the third thing I would point out that would make kind of make material difference in this election if they could do it is impose a form of photo ID that no state in America that has a photo ID law would countenance.

unknown 24:50

Okay.

SPEAKER_01 24:50

So altogether those things so for instance the Postal Service effort right now there's a there's a rulemaking and then there's litigation to create this system that would require states to send their voter list to the Postal Service. The Postal Service will do something to bless the list and then you could potentially be legally liable if you send a ballot to somebody not on this list. That happily has been put on a hole by a federal court so that's not going to happen but it's one of the things out there. Similarly the Department of Justice has gone to the states trying to get voter lists unredacted not only with your name and address which anyone can get but your social security number your driver's license number presumably because they want to be ready to perform data matches to challenge voters in this election. And then finally they're trying to impose this um this voter ID law voter ID laws are very popular and I've predicted that eventually every state will have one. Some will be more stringent some will be less stringent but they're so popular as 20 years from now I think every state will have one. And not a number of states have them now but the the photo ID this photo ID law would basically say you have to show a driver's license. If you don't have a driver's license or a passport tough noogies. Every state that has a photo ID law has a list of things and has a backup if you don't have any of those forms of ID has a fail-safe effort. States have thought this through carefully and so I think at the bottom line so that's the material those are the kind of three material things that the that the administration has tried to do. They've been thwarted by the federal courts or by Congress this is the Save America Act. The Senate is not going to pass it. The courts have thwarted the other two things. Despite as an aside reasonable people in the middle could come to agreements about many of these things that the Trump administration is trying to push extreme versions of. Yes I think the the the short the short version of this is imagine 2020 but with a threat surface that involves four times as many states and 40 times more seats in the federal government. By which I mean there will be a narrative in fact we can hear this right now there's no way in the world the only way Republicans cannot prevail in the 2026 election is if the election is rigged. Well we know the fundamentals of political science whether you like it or not based on data going past back a century and a half given the unpopularity of this president and the and the bad economy that we have Republicans are going to do poorly in this election justly or no, they're going to do poorly in this election. They're going to lose the House and they may even lose the Senate. Okay? It's going to be a bad day for Republicans. So what's going to happen on Wednesday morning? A lot of people are going to be saying it was rigged. Why was it rigged? Because we had non citizens because people weren't required to show their ID. There's a lot of fraud in a lot of ballot box stuffing. There's a lot of non citizens because people aren't required to show their citizenship when they when they're they vote. California is going to take five years to count these ballots there's something nefarious there. You can kind of see the froth that can be um that can be shinned up in the days after the election. And this is going to motivate challenges to close results and maybe even to non-close results. Now what we'll see is the courts bat them down and county commissions batting them down and all of that, but it will be a degree of chaos that I think will go beyond what we saw in 2020. And because you know there's about 30 or so House seats and you know kind of in play. Yeah and there's about a half dozen Senate seats in play. Right. Right. That's when I say the threat surface is enormous. And so it doesn't take a couple of judges to do the wrong thing. It doesn't take a couple of county commissions to do the wrong thing in which you know we're really in kind of a never never land of some people being challenged not being challenged. If you're watching the news it looks like the country is just melting down. Right. Then when you get to January then when you got to January 3rd, January 4th probably is when they're they're going to convene, there'll be a couple of red meat folks who will get up you know Democrat from California will show up. They've defeated a Republican and and there's going to be a challenge to giving the oath of office to that person. And so that's kind of the headline version that's the reader's digest condensed version of how kind of the setting up the predicate then gets reinforced throughout the campaign season and then in the host election.

SPEAKER_00 30:03

Yeah. And I I mean we are looking if you do you know presidential approval rating view of inflation view of of the economy you you are looking at president's approval rating right now is about 35 to 37 percent. So in the in the math of political science that would you know the average turnover in the House would be 26 is a little bit less now obviously because of redistricting and so on so forth. But nevertheless, you're looking at a gain, a likely gain between 10 and 18 seats on the on the Democratic side this time around. And so you know the the that brings into play if it is a blue wave, so to speak, then you have districts that ordinarily are plus five to plus ten Trump that are suddenly those congressional districts are suddenly razor close and maybe the Democrat wins. Is that that's the that's where we would likely see the most agita, right?

SPEAKER_01 30:58

I th I think so and the importance and and and the thing to keep in mind and I think a lot of folks who live in you know urban areas like I do have lost track of that there's a lot of people out in the country. I'm I don't want to be pejorative here. It's kind of out beyond the center city who are agitated about the state of our elections. And it doesn't take many of them and they're organized. They go to seminars they buy the pillows out in the out in the out in the and out in the lobby they are exercised about things like this. And they're going to go to the meetings of the of the of the Board of Elections. They're going to be protesting that small county elections office that's never seen anybody even say boo to them before. And so that's going to be part of the dynamic I think that a lot of election officials are worried about and democracy protectors are preparing for is if not real violence, then real you know, anger expressed at the process and it will be dealt with by people who are not used to being yelled at, not used to being shouted at, not used to getting people pounding on their windows or yelling obscenities at them at the grocery stores. And I I think that will be a a you know a very unfortunate political dynamic.

SPEAKER_00 32:19

I hope it doesn't happen and I'm usually optimistic about these sorts of things but it's the sort of thing that needs to be prepared for one thing I noticed from you know doing research on on your work is that you you are can't say this definitively but you are very, very concerned that there's sort of no way back from this from this agita about election outcomes that that it's poisoned the wells so badly that it can't be unpoisoned. Is that accurate? Aaron Ross Powell I I do I do worry about that.

SPEAKER_01 32:52

And you know looking back over American history, you know how you get out of a situation like this isn't very edifying, right? I mean the kind of the progressive movement arose out of you know that generation's disdain for the degree of division that led to the Civil War, for instance. Right. So you know that took that was a long time and it took a horrible experience in this country to kind of get back into kind of some sort of try to get across the aisles. Or you know the Great Depression. I mean it's like these real cataclysms that end up breaking political divisions. So I don't see how we get out of this. And being a technocrat in elections I mean I think it's a second level concern if you're you know the big concern is about just the nature of democracy. But it also means, you know, like everything, there's imperfections and there's work that needs to be done to improve elections. And that's why I got into this business a quarter century ago. And this sort of agitat gets in the way of building a better, a more perfect election system than we have now.

SPEAKER_00 34:02

Aaron Ross Powell So every question eventually comes down to AI, the concern people have about what AI will do to X, Y, or Z. How concerned are you about AI hacking the election or disrupting the election? I'm not quite sure where that question ends up, but you know what I mean.

SPEAKER_01 34:22

Yeah I I know what you mean and and I I think I and I think the trouble in really kind of getting a sharp question in this in this area reflects the fact that everybody in elections and campaigns are trying to get their arms around this issue. This is how I parse it. There's the campaign and then there's the administration of the election. And we have to assume that because campaigners are strategic animals taking advantage where they can find it, the AI is going to be the new tool in the toolbox and it's going to be good for use for good and used for evil. At the same time on the election administration side my sense is that election administrators are worried about kind of kind of misinformation and those sorts of things that they now might have to deal with. But you know on the other hand the typical election office that's running elections has one staff member, one permanent staff member and it might surge out you know for the election for the poll workers, but they have one permanent staff member and they have to do all these things associated with elections. And the tools that come from AI to draft notices, to kind of do mundane job of say communicating with voters and getting the paperwork out and doing all these sorts of things AI is being seen as actually a great tool for underresourced election officials. And so I think it really is a two-edged sword that those of us, most of the world that lives in the in the in in a world of the campaign, it's like buckle in and see how this tool can be used because people are being paid millions of dollars to take advantage of the technology. On the quiet side, election officials are cautiously optimistic that AI actually will be they'll be able to use it to their advantage in the same way that people are learning to use AI to do mundane tasks and to expand their capabilities.

SPEAKER_00 36:22

Who is that in most states a county vote is how it's counted except for New England, right, which is town by town by town. So is there one person by county uh in those 44 states I guess that do it by county? And how does that work in the New England states?

SPEAKER_01 36:41

Yes. So the data come from a survey two things. One a survey conducted at Reed College now at Portland State College. They've been doing really interesting survey work of election officials and then the Election Assistance Commission also does data collecting from local officials. So there's a bit over 3300 counties that run the run elections. On top of that, as you said, um New England plus the upper Midwest so Wisconsin and Michigan, you also have authority at the local level. In fact Wisconsin has something like a quarter of all of the local election officials in the nation in that one state because it's done at the Hamlet I mean it's it's really highly decentralized. And so when I say one person I mean that of these 10,000 jurisdictions it's not just the counties it's of the 1000 jurisdictions most have a that's called the town clerk. They don't and they have no support staff at all. Right. And they may very well be doing the dog licenses and the building permits and all the other things in in that town right that person did not go to college and um is a well that's probably the the the relevant thing. People oftentimes will talk about you know middle aged women which I think actually is kind of irrelevant here but it's you know basically you know high school educated folks who are doing a lot of things on their own. There are now that's most officials most voters on the other hand are in like the 200 largest jurisdictions in America. LA County for instance LA County has more voters than all but seven states. They have a helicopter I mean they have all sorts they have hundreds of employees so I don't want to give the the impression that but you know if you are you know if you were the town clerk in the middle of Wisconsin you would like the capability that Dean Logan the the chief of elections in LA County has. You would like to be able to communicate with your voters about the complexities of the law communicate with them about why there are these anomalies. Tell them how to better fill out their ballots. There's all sorts of things they would like to do. And so and I make this point there's one person you know if you were in charge of it all you would like the capability that a Dean Logan has in LA County Aaron looking at the 2026 election and looking ahead to the 2028 presidential election how are you going to study that? Are you going to do that as part of the project at MIT or are you going to sort of form a consortium to you know do a larger survey if you will we're doing a lot of things here at MIT and the MIT election data and science lab. Part of our mission is to do original research ourselves but core to our mission is um creating collaborations, creating collaborations with election officials to do research and bring best practices to the field. And then also collaborating with other academics in this area. So we are right now standing up a program called the Election Performance Central which we also did in 24. It's a successor to the Healthy Elections project that Nate personally and I ran in 2020. So we at so we at MIT will be writing reports about battleground states, talking about the data there, talking about the processes there. We're doing that in part to inform the public, but we're also flexing our muscles to get used to using this data, learning about the facts in the states, getting ready for what happens on the Wednesday after election day. We're also putting together a coalition of um it'll probably be 40 to 50 political scientists and law school professors around the country who will have their pulses on states, we'll have a Slack channel, which is the thing that kids do these days to communicate about what's happening in those in those states um we'll be getting together regularly to you know to support each other, to compare notes, and again to be ready to move on the Wednesday following election day when things begin to heat up.

SPEAKER_00 40:48

Aaron Ross Powell And does the system generally speaking obviously it it probably varies state to state and county to county but is is there enough money to support this work or or is that one of the concerns that you know the budgets aren't getting bigger to ensure the accuracy and quote fairness end quote of the elections?

SPEAKER_01 41:08

Aaron Ross Powell There's great concern about the resourcing of elections. And I'm kind of in the middle of the road on this issue because it's it seems because it's clear that election administration which is really a local a local government service is underfunded for the needs. Everything is underfunded in in in local government but given the need they just can't win out against the schools and the police and the fire department and the roads. We can understand why that is a lot a lot of reasons for that. And so local governments don't spend enough it seems easy I talk about a problem called how hard can it be ism that funders generally don't appreciate the complexities in the administration. Just go to the Xerox machine run off a few ballots and you know how to count things. Yeah. Don't bother us. So so there is that there is a legitimate question about who should be paying with this. Certainly the election activists would like the federal government to pony up everything that's needed. I think the federal government should probably pony up more but if we take seriously the decentralization of our electoral process I'm in favor also as we figure out what's the appropriate federal role in supporting elections, paying for it, we need to be pressing state legislators and county commissions also to fund state and local elections. If states are going to push back against the federal government and say, we're in charge, you're not the boss of me, we're in charge, then they're not in a good position to say, ah, but we want the federal government to pay for this. Give me the money. But give us the money. We want the money but we don't so I mean I so there's this it's more complicated than many people think. I think again I think there's a middle range there where a deal could be worked out but we're not in that world where reasonable people can get together and strike a deal.

SPEAKER_00 42:57

Aaron Ross Powell One last question I have is it given the the you know what you described as going to the gro uh an election worker going to the grocery store and getting you know yelled at and pizza boxes to election officials homes and so on and so forth. Is is there are you concerned that people just don't want to do that job anymore that you'll have you know you'll have poll workers like here in Southworth, Connecticut where I live, I think there are I think Fairfield has nine voting places, right? And nothing's going to happen in Connecticut because nothing ever happens in Connecticut. But it if it were a 50-50 state and you thought well gee if if I you know if the election is really close and it goes one way or t'other I'm going to come under threat. I don't want to do that anymore. Is that a a real concern in in the states that we all think of as the battleground states?

SPEAKER_01 43:54

It's certainly a concern. Election officials express that concern. At the same time the the research that has been done asking election I mean poll workers themselves about why they do things, why why they volunteer, why they come forward. It's actually technically not volunteering, but why they come forward. And their experience when they do the work they come forward for highly patriotic and civic duty reasons. Thing number one thing number two, they do not experience the types of vitriol and negativity that we read about. And so their lived experience working the polls is not what you would worry about if you were listening to podcasts or watching the news and those sorts of things. And which by the way is also the experience of voters. And this is part of my research as well. When they go to the polls they don't see troops they don't see violence they just see quiescence. And so you know I I think at that local level yes we're worried but what oftentimes is mentioned over threats is poor pay, long hours, those sorts of things I mean it's kind of workforce issues. Right. We need to pay people more we need to train them better. We need to give them better working conditions. And that's probably as important as I I'll put it this way that's something that at least state governments county governments can control. They can't control what the president's saying but they can control these things. Aaron Ross Powell we've taken a lot of your time um what questions didn't we ask you that we should have I can think of a bunch of questions you shouldn't have asked and you didn't remember I have a list of 40 things um that uh that'll be the next podcast we can do the other 37 exactly um invite me invite me back on uh on November 4th and um and we'll see how things go. I just wrote you know I I just wrote actually you should have asked me um or is I I'll volunteer the answer. Okay. I'm known I'm known as the optimistic guy in this area. And I've been talking to a number of reporters over the last several weeks when I kind of talking about what I anticipate and a couple of them have said no I've been talking to you for 20 years and I've never heard you talk like this before. What's happened and and have things fundamentally changed? And my answer is I experienced the 2020 election in the after aftermath and I know what we have to be prepared for. But I continue to be as optimistic as I've ever been about fact-based institutions of this country. And um so the the reasons for optimism is that those institutions so far have proven to be robust and the the ball that we should keep our eye on really ends up being the performance of the election officials, the judges, the county commissions, the county boards of elections, those people whose whose decisions and actions will be determinative, not the flack that's thrown up around them that's a great way to end.

SPEAKER_00 46:57

Professor Stewart thank you very much for doing this today and we will get back to you on November 4th or 5th and see where we are look forward to it. All right thanks again