Transcript: Dems Will Win the House. The Senate is the Big Question

This is a lightly edited transcript of the October 15 edition of Right Now With Perry Bacon. You can watch the video here or by following this show on YouTube or Substack. Perry Bacon: Good morning. I’m Perry Bacon. I’m the host of Right Now on The New Republic. We have a great guest today. It’s G. Elliott Morris. People probably already know who he is, but anyway, he was at FiveThirtyEight and The Economist before, doing great analysis of data and politics. He’s now doing his own newsletter, Strength in Numbers. He’s also helping run a site called 50 Plus 1, which has some great forecast and modeling of the election. Elliott, thanks for joining me. Great to see you.G. Elliott Morris: Yeah, Perry. Happy to be here as always. This morning I’m on my second cup of coffee during our call. You might see me pour a third.Bacon: All right. So I’m going to go through a bunch of different races, because you’re the man to ask for this kind of thing. So I want to start off with the House. I’ll only ask this broad question first, which is, you’ve been discussing this a lot, but I don’t think we’ve had it on Right Now, and this is just asking very explicitly: my liberal friends, and most of my friends are liberal, are very nervous about 2016, 2020, 2024. Their perception is the polls are always wrong and the polls are always inflating Democratic vote share, and they’re very nervous about being disappointed again. So should they be panicked by the polls showing Democrats leading by a fairly substantial margin?Morris: This is a good question.Bacon: Maybe panicked is—I framed the question poorly. Talk about—Morris: No, I think you should be as panicked—so I was at an event with a senator recently, and they asked this question, “Are the polls right?”Bacon: Like a United States senator asked people this question. That’s interesting.Morris: Yeah. A US senator asked the polling nerds in the room, Hey, are the polls going to be right? And there were some people in the room that are like, Hey, we’ve made a lot of changes over the last couple years or the last decade. They should be more accurate than they were in 2016, for example, when they weren’t weighting by education, which was just, in hindsight, kind of a rookie error. It’s just one of those things that wasn’t necessary before, so pollsters didn’t do.And I wasn’t super convinced by this. My response to the senator was, they’re about as good as they have been historically, which is to say you should expect them to be off by about three percentage points, potentially in every single race by about the same amount. Which is the type of error that we would not have reacted so harshly to, first off, in a year where there’s a blowout, when it just doesn’t matter, or when people weren’t conditioned to expect so much precision from the polls.So I think the thing that has changed over the last 15 years—and it was truer in 2016 and less so now, because there’ve been so many errors—is that people just expect the polls to be super accurate. And they’re not. But they’re not necessarily worse than they have been historically. There were some cases of some really huge errors in the 1980s. So if you want me to give some comfort to your super blue listeners, the comfort is they’re almost certainly going to win the House, and you can do a lot of things without winning the House.Bacon: Let me stop you there for a second to ask that. Because you do these forecasts, so presumably you think the polls are not useless. So you’re saying the polls might be 3 percent off, but the forecast accounts for that, correct?Morris: Yeah. So the reason why the forecasting is useful is that it explores what could happen if the polls are wrong. If they are off in the amount that they have been historically.Bacon: Off in the past.Morris: In my mind, we’re doing the hard work for you, the reader, or in this case the viewer, by taking the polls today and shifting them three points in one direction, count up the seats. Today Democrats lead in 54 states according to the polls. If you shift all the polls three and a half points, 3.3 to be exact—that’s the historical error for Senate polling this far ahead of an election—you can get 49 seats for the Democrats if you shift in the Republican direction, or 56 for the Democrats if you shift in that direction. That is the range. Most of these outcomes include Democrats controlling the Senate, but not all of them.Bacon: Let’s go through the details of that. Your estimate is the Democrats have a 99 percent chance of winning the House, on 50 Plus 1. The median outcome is 49 seats gained by the Democrats, so 242. I looked up VoteHub today. They have it at 231 seats and an 84 percent chance. Split Ticket has it at 234 seats and an 89 percent chance. So in other words, the 84, 89, 99—the analysts agree the House, it would be an almost unprecedented error in forecasting and polling if—the Democrats are overwhelming favorites to win the House. Is that correct?Morris: Yeah. And just to say, I think The Economist is at 95 and Nate Silver is at 93 or something. I don’t subscribe to Nate Silver so I can’t see, but I have someone who does subscribe tell me the number every Friday. So that’s as of last Friday. It could be even higher by now.So it’s not hard to believe, like 90, 95, frankly not hard to believe. Democrats only need to pick up three seats. In the generic ballot today, the test of the House popular vote, they’re doing around plus eight or plus nine in terms of margin, which would be an 11 percentage point shift since 2024. The last time there was a shift that large was 2018, and that year Democrats gained 42 seats. So they’re going to gain less because of gerrymandering. But maybe just 35. OK, they only need to gain three. So it would take a 9/11 style political event.Bacon: We’ll pick a different analogy, but I get the point you’re making.Morris: I know it can be toxic, but actually that was the last time—2002 was the last time the incumbent party in the White House gained seats in the House. That’s what I mean by that.Bacon: I see. Got it.Morris: So it would be two out of the last 100 years or so of House elections is what we’re talking about. And obviously we have a lot of leading indicators. There’s no signs of that sort of political environment. Like in 2002, it was easy to see this coming. Democrats weren’t leading by 10 points in the generic ballot.Bacon: Senate, you have Democrats winning 71 out of 100. This is a 71 percent chance, I think, and you have a median outcome of 52 seats for the Democrats. VoteHub has it at 59 percent Democratic, likely to win 52 seats. Split Ticket at 65 percent and 52 seats as well. So that’s around where Hillary versus Trump was. That’s in the range of Trump’s potential win in 2016 when he was the underdog. So in other words, people should not be banking on—the Republicans holding the Senate is a realistic outcome of the November elections, correct?Morris: Yeah. The way to read the polls today is that if they have an error that is historically average, in the Republican direction, Republicans hold the Senate. If there is a larger than average miss, then Democrats might only be picking up two or three states, maybe even one. And those sort of errors happen, and they are basically impossible to predict ahead of time.The reason I like the forecasting is that we are doing a lot of this hard number crunching for people behind the scenes to prepare optimistically, right? To prepare them for potential polling error that might happen on election day. Instead of saying, “Just look at all these polling averages, look who’s up, look who’s down,” I think that leaves a lot of information on the table, since there’s a lot of uncertainty we can talk about with the election forecast that you can’t get by just looking at the polls. That’s the entire reason the forecasting exists in the first place.Bacon: So the Senate, we should be looking carefully at, and that is something that, like you said, a normal polling error. I think that’s a good phrase, actually.Morris: Yeah. And it’s not hard to see how this would happen. So Democrats are leading by three points in Texas. I think a little bit more, four points maybe, in Ohio. They’re leading by a couple of points in Maine, where Susan Collins has historically beat the polls, and they’re leading by one point in Iowa. They’re leading by two points in Kansas. I don’t know if I believe that, but we are not in the business of putting our thumb on the table. We can come back to Kansas. So there’s one seat for Democrats that I think is a pretty sure thing. That’s North Carolina. Former Governor Roy Cooper is super popular in the state. He’s up 5.1 percent in the polls.Bacon: One seat they’re going to flip, just to be clear what we’re talking about. Ossoff’s going to win, but one seat they’re going to flip, yeah.Morris: So one Republican to Democratic flip. That gets them to 48, and then all these other states I’ve just listed, those are within three points. Those are within the historical average polling error for Senate races. Don’t be surprised if the polls are wrong. They’ve been wrong before, and in this case, if they are as wrong as they have been on average historically, it means Republicans hold the Senate.Bacon: Interesting. OK, so let’s go through some races specifically now. So I looked up Cook Political Report and went through some of their likely, toss-up, et cetera, and their breakdown. So let me go through these three. So Cook Political says likely—which, likely versus lean, we’ll come back to that. So likely Democratic Senate, Georgia, North Carolina, and they’re saying Minnesota is likely as well, but maybe not solid.So talk about Minnesota. Do you think there’s an outcome—how likely is it? Because the Republicans are hinting Minnesota, there’s a lot of talk that the Democratic candidate is not particularly strong there. Do you see any realistic chance of an upset in Minnesota, where the Republican could win?Morris: A lot of people tell me—Bacon: Are you saying yes?Morris: No, I’m not going to give you a yes or no answer, Perry. Nice try.Bacon: Fair enough. I want to make sure—you said “Yeah,” so that’s why I was making sure.Morris: Oh, OK. No, I’m not saying yes.A lot of people tell me Peggy Flanagan’s a bad campaigner, she’s a bad fit for Minnesota. I think this is a lot of rationalization of the polling numbers. It is hard for me to know whether or not that’s—I don’t really think we can quantify what makes a good campaigner, a bad campaigner, very well. So much of this is subjective. And also, it really doesn’t matter until people start counting the votes. So maybe it’s true, maybe that’s an explanation for her lagging poll numbers. I don’t really know.The fact of the matter is, the polls in Minnesota are closer than you would expect. Peggy Flanagan’s leading her challenger, the former sportscaster there, by four points in the polls. This is a state where, in a year Democrats are leading nationally by eight or nine points, you should expect Minnesota to be two points to the left of that. So you should expect Peggy Flanagan to be around D plus 11, and she’s at D plus four. So it is true that she’s underperforming, and that race could go the other direction.However, what we know from studying polling accuracy historically is that usually when the polls are wrong, they err in the direction of the fundamentals. That would mean in a blue state, in a blue year, the polls have historically underestimated support for the Democrat. Or in your Kansases, your Iowas, your Ohios, your Texases, if the polls are wrong, typically in states like that, the Republican would tend to overperform. And that’s because we have other information from the fundamentals that give us clues about elections that’s independent of the polling data. You want to average these two things together, historically speaking.So our forecast model is around the likely Democratic zone in Minnesota. We have Peggy Flanagan at about an eight out of 10 chance. So slightly more optimistic than, to take your example, Clinton’s odds of victory in 2016. But a 17 percent chance happens one out of seven times. There’s 30 elections this year in the Senate, so it wouldn’t be totally surprising to see that race go the other way. And if it does, then the Democratic path to a Senate majority gets really hard.Bacon: So lean—Cook Political lean Democratic, so less than likely. They have New Hampshire—Morris: And by the way, lean means something around 70, 75 percent chance in Cook’s ratings historically. And likely is around 95.Bacon: OK. That’s not what they say. That’s what you’ve calculated, right?Morris: That’s what I’ve calculated based off their historical ratings. I don’t think they give the probabilities like this.Bacon: I don’t think they say that.Morris: Yeah, I don’t think they agree with the quantification of the ratings in general.Bacon: They have lean, which is, like I said, less than likely, as you’re saying. Lean Democrat, New Hampshire. So do you see that as a place also where we could get an upset? What’s your sense of New Hampshire?Morris: Yeah, New Hampshire is a weird one. The polls were really close there about a month ago, but—sorry, I’m blanking on Pappas’s first name. Help me out here.Bacon: Chris.Morris: It’s not Chris. Oh, are they both Chris?Bacon: I think—Morris: Chris Sununu is John Sununu’s brother. Yeah, Chris Pappas. I was like, surely they couldn’t both be Chris.Bacon: Yeah.Morris: They could. It was a really close race earlier in September, for example, back when the primary was going on. It’s possible that there were a lot of Democrats that would end up voting for Chris Pappas that were supporting his challenger, that wouldn’t admit that, and now they’ve come around.But the race isn’t looking nearly as close as it was previously. And the quantitative forecasts like ours at 50 Plus 1 that were hedging on those close polls and saying, this state looks like a state that would be more favorable to Democrats—I think that hedging was right, and the race raters weren’t saying this race is a toss-up, I believe, even though there was a lot of conversation online about maybe New Hampshire is going to fall to John Sununu again. It doesn’t look that close.Now, again, seven-point lead in the polls—leads of that size have evaporated with about a 5 to 10 percent chance, this close to an election. So it’s not impossible that John Sununu could win. But yeah, this is one of those states I think people got a little out over their skis on a month ago. It’s not as close as it seemed.Bacon: So toss-up. They list seven states: Alaska, Iowa, Kansas, Maine, Michigan, Ohio, Texas. So the ones I want to talk about right now are Michigan and Kansas as toss-ups. What’s your sense of Michigan and Kansas?Morris: Michigan. If you ask someone from the Cook Political Report this question, or from election Twitter posting universe, they’re going to say, “Oh, Abdul El-Sayed is too progressive to win in Michigan,” or maybe he’s more progressive than your moderate candidate in Michigan, so he’s underperforming expectations. I think we could save the hand-wringing in the moderation versus progressive debate until the actual votes are counted and we see how much he underperforms by.There’s also a lot of other descriptive factors about Abdul El-Sayed that might be making him underperform in Michigan. There’s other experience factors that could also be factoring in here. So I don’t want to anchor us to any particular explanation now, especially one that is exclusive of other explanations, like your first Muslim senator winning in Michigan. That seems like, maybe some white voters in Michigan might not be so happy about that. Let’s at least leave that on the table as a possibility now.But this is slightly surprising to me in terms of the Mike Rogers of it all, rather than the Abdul El-Sayed of it all. Mike Rogers underperformed by a couple points in 2024 when he ran and lost to Elissa Slotkin.Bacon: Let me ask you. What’s surprising is the race is close, or you’re saying that Cook is calling it close? Because those are different questions.Morris: So it is surprising that Cook is calling it a toss-up, given the polls show consistently that El-Sayed has, I think, a three to four-point lead, four-point lead. So they’re saying, “Oh, these polls are...” The literal read of that is that they’re betting on some amount of polling bias. Which, they have their right to. Maybe they’re seeing some internal data that we’re not seeing.But from a 30,000-foot perspective, I would have also expected Mike Rogers to be doing worse, as someone who recently lost an election in Michigan. He underperformed Donald Trump in Michigan in losing the state. And that is a state that has been hit hard by tariffs and, just like everywhere else, by price increases and the affordability crisis that Donald Trump has publicly taken basically a lot of credit for, in pushing for inflationary policies over the last 20 months.So if Abdul El-Sayed loses there, it’ll be a combination of factors. Maybe he was too progressive for some members of the Democrats. Maybe he was too Muslim for other swing voters. Maybe Mike Rogers did better in the debate, right? We can add up all those factors. But yeah, it’s close enough that it could go either way. That’s the state where a normal polling error means Mike Rogers wins the race.Bacon: Kansas. I’ve not read all these polls. I’m just very dubious of calling Kansas a toss-up, but I’ll be open to your perspective, and if I’m wrong, I’m open to being told that.Morris: Our model says Kansas is a toss-up. I wouldn’t put on the internet Perry Bacon is wrong about Kansas. You can disagree with the polling error, or you can disagree with the polls in a state that red, with that many working class white voters. You could make the case that the polls are undercounting Republicans there. My whole point is that we don’t know about that until the election.Bacon: But your look at the data suggests it’s a toss-up, though. The polling is close, is your point.Morris: The polling suggests a toss-up, with Hamilton—I’m just looking up our polling so I can give you the precise numbers for your audience. Hamilton leading by two points now on average against Roger Marshall. Some polls have it at one. New York Times has it tied.Bacon: Hamilton’s the Democratic candidate.Morris: Yeah, Adam Hamilton’s the Democrat. So, like I said, these are the types of people who haven’t been answering the polls in recent elections, so maybe those polls are wrong.I am surprised that Cook calls it a toss-up, because the whole point behind the qualitative race handicappers is that they take a holistic view of all the indicators, that they don’t hew so closely to the polls. So that makes me think that they are seeing some other internal data, or with candidate interviews, that also suggests that the race would be close. And typically your best predictions for elections come in averaging the polls and the handicappers and the fundamentals. In the case of Kansas, these are all suggesting basically a tied up race in an environment that’s this Democratic. So maybe it actually does make sense. This is not the type of state we would have all expected to be competitive a year ago, though. So I understand the hesitation from people, the skepticism.Bacon: So Texas is called a toss-up here. My sense of looking at the polls—Morris: Cook is calling it a toss-up, just to be clear.Bacon: Toss-up. What I’m asking is, because in this case, you’re saying a three to four-point polling miss is typical. My sense is the polls are showing Talarico ahead by more than three or four points. Are they just assuming Texas is Republican, just the way I would? Is there some reason to think this is a toss-up beyond the history of Texas? Because the data would seem to be in the Talarico-is-the-favorite zone, as far as I can tell.Morris: Yeah, exactly. And our forecast had Texas at lean, right? A 60 percent chance for Talarico yesterday. And today it says toss-up, because he’s gone from 60 percent chance to 59 percent chance. There’s really no difference between these takes in terms of the numbers of it all.But your read is correct. If Cook is seeing the same data we are, Talarico up three on average, and they’re calling it a toss-up still, then they are averaging it with the fundamental indicators. That would be stuff like Ken Paxton holds a statewide office. This is a state that’s really red. Maybe you would consider that Ken Paxton has been plagued by some amount of scandal, although it’s not rising to the level of, he’s sleeping with his staffers, right? The type of scandal that does typically rock the campaigns. These recent leaks of him talking about how bad his campaign is doing does not help. You should probably not be allowing so many donors in the room with their phones at that point, I would think.Bacon: But you’re saying that your polling average in Texas is about three, and if it’s three, then that gets in the normal polling error. That makes sense. I’ve seen polls saying the poll rates higher.Morris: But to call it a toss-up, you have to bet on the polls being biased in the Republican direction by about that much. Which is not a bet that historically would have worked out. So Cook is saying, “We actually will take that bet. We do think the polls are maybe overestimating Talarico because of some other data we’re seeing, campaign finance, whatever.” There is a lot of money being spent on Ken Paxton’s behalf. And you know what? People wouldn’t spend a billion dollars in elections if it didn’t matter. So maybe we can expect that to have a difference in the last two and a half weeks, three weeks, wherever we are.Bacon: How do you see Maine?Morris: Yeah, Maine’s great. It’s a good example of the tension where the polls and the fundamentals disagree. In Maine in 2020, Susan Collins beat the polls by about eight percentage points against Sara Gideon in that race. This is one of the larger polling errors of recent time in a Senate election.So it’s notable that if you’re expecting that level of polling error again, the polls to underestimate Susan Collins—and she has beaten expectations four elections in a row—that’s a fair enough expectation. Let me just get to our Senate forecast here for Maine. Then you would be expecting her to be leading in the fundamentals view of the race, right?Let’s just unpack that for a second. If Susan Collins typically beats the polls by eight points, and historically she’s beaten the average challenger by more like 12 or 13 points above what we would have expected based on how blue the state is in presidential elections, you would be expecting her to win by four or five points this year. So the polls then showing Troy Jackson up three points, that’s an example where, contrary to in Texas or in Kansas, the fundamentals and the polls really disagree.So like I said earlier, those fundamentals—you should expect polls typically to miss in favor of the fundamentals, for the election results to be somewhere between the polls and the fundamentals. And that argues for a more skeptical view of the surveys in Maine. So they show Troy Jackson up two points. Maybe you can mentally adjust this in your head and say, if you’re a Democrat, always two points behind. I typically tell people just always subtract two from the side you’re betting for in the polls, and then you’re prepared for if they’re wrong. You can make your campaign decisions in the world that you’re behind, that can prepare you for victory better.But I really have no idea on the ground. So I’ll just say a couple things about the issues. I know you have more questions. ICE and immigration have been a huge issue in Maine. Jackson has been dogging Susan Collins for her votes for ICE and DHS funding in these debates, and there’s some high-profile incidences of violence and death at the hands of ICE in Maine. So this is one of those states where immigration is maybe more on the minds of voters than it is elsewhere. And affordability, and especially gas and diesel prices, matters in a state like Maine in the winter, where most people heat their homes with oil.So I would expect, based on the issues, that Troy Jackson comes out ahead. The fact that it is so close suggests, maybe our storytelling about Susan Collins as the enduring moderate has some staying power in 2026.Bacon: OK. So Cook likely Republican—I’m going to go through these three states quickly. Likely Republican: Nebraska, South Carolina, Florida. Do you think there’s any chance Democrats win, or I guess not Democrats in the case of Nebraska—is there any realistic chance Republicans lose Nebraska, South Carolina, or Florida?Morris: Yeah, there’s a chance for sure. Our model thinks that there’s about a 20 percent chance that Democrats take either Florida or Nebraska, and maybe closer to 30 or 40 percent in South Carolina. Just based mainly on the surveys in South Carolina, that is.And this is maybe a word about the forecasting models here. One of the reasons why the quantitative forecasts are so optimistic is that they’re simulating different elections. In one election, they take the polls and they add five points. In one simulation, they take the polls and they subtract five points for the Democrats. Sometimes they also will explore state by state errors in the surveys. And if you do that enough, then on average in these states, South Carolina, Florida, and Nebraska, in these simulations, you should expect the Democrats to pick up one of those in many of these simulations.I think on average these are adding about a third of a seat to the Democrats’ total. So that pushes up their expected seat number on average, which increases their probability of victory. Because it’s, oh, there’s some edge cases in these states. We don’t really know what’s going on in South Carolina. So maybe Democrats pick up that seat in this simulation. If you do that 40,000 times, like we do at 50 Plus 1, like Nate Silver does, then these models are accounting for surprises in these edge likely seats. And so that adds to their total. Whereas if you’re just taking a straight count of the qualitative race raters, you would basically wipe all these likely Republican states off the map.South Carolina’s a really interesting one. I don’t have to tell you, this is a really Republican state. Annie Andrews seems like a type of Democrat that might be able to win in South Carolina, and maybe one dynamic here is both of the nominees are women. So there’s potentially less latent sexism fueling voters’ decisions in a Southern state, because both of the nominees are women. Also Lindsey Graham wasn’t exactly beloved for his support of the war in Iran at the time of his demise, and that could be weighing on his—I’m sorry, is it his sister or his sister-in-law? It’s his sister, right?Bacon: Yes.Morris: Yeah. It’s not implausible that in a super Democratic year where there’s a lot of anti-incumbency bias, someone like Annie Andrews could win there. I am personally, after seeing the polls, a lot less optimistic on behalf of Angie Nixon in Florida than I am for Annie Andrews in South Carolina.Bacon: OK. So governor, Cook toss-up: Wisconsin, Georgia, Nevada, Ohio. Talk about Wisconsin and Georgia, because the sort of Twitter, Bluesky discussion assumed that Georgia, Keisha Lance Bottoms is in big trouble because she’s being outspent by a lot, and assumed that Wisconsin, the Democrats picked the quote unquote “normal candidate” and they’re going to be fine. So do you see Wisconsin and Georgia as toss-ups?Morris: I see them as toss-ups. Yeah. In Wisconsin the Democrat is ahead by two points in the polls. That’s within that historical polling error range we’ve been talking about. In Georgia, it’s the opposite. Bottoms is behind by one and a half or two points in the polls. These are both states where if the polls are wrong, they could go the other way. Toss-up seems reasonable, especially in states that are—and Georgia’s definitely to the red of the nation as a whole.Wisconsin, at the state level, is a little redder. But a weird thing about Wisconsin is that it gets so much investment that it doesn’t tend to move as much as the overall national popular vote does. So I’m just going to say two sentences about this. In between 2020 and 2024, the national popular vote moves six points to the right. Wisconsin moves a point and a half to two points to the right, if you’re generously rounding on Donald Trump’s behalf. And it was similar from 2016 to 2020, it moves a point, whereas the nation as a whole moves by about two and a half.Bacon: So it’s been very divided and very polarized, or something like that.Morris: Very divided, yeah. And it gets a lot of investment, so there’s a lot of voter contact that helps the Democrats reach the people that they’re always trying to reach, who can fall into the sort of low engagement category in other states. So Republicans that have been talking about, “Let’s just get Trump out, and we’re going to activate all of our low turnout propensity Republicans”—there’s a lot fewer low turnout propensity Republicans, in terms of composition of the population, in Wisconsin, because it gets so much investment from the parties. The voters there are really high turnout propensity just in general.So it’s kind of a weird state. I would expect it to be close, even in a big blue kind of year.Bacon: Every election there is close.Morris: Every election in Wisconsin is close. Yeah. I had the former chair of the state party, Ben Wikler, on the Strength in Numbers podcast a while ago, and he lays out a much better explanation than I’m giving you here, in terms of how the party actually makes its investments in voter contact. That explains why it is less elastic, why vote choice there is more stable, and why it’s always close, regardless of what the nation as a whole looks like.Bacon: OK. Lean Republican governor’s races, Florida and Texas. And the reason I brought this up is because those would be earth-shattering. Like, Talarico beating Paxton is explainable in part because Paxton seems like a corrupt, crazy person. But either one of those races flipping at the gubernatorial level would be really big news on the map. How do you see those races?Morris: I think those ratings are good. I’m from Texas, and it’s not going to be a mystery to your audience, I lean to the left. So I think that Greg Abbott has done a bad job as governor. And this isn’t super punditry, there’s some good examples of this. There’s a lot of dissatisfaction on two policies in particular, data centers and the education system, which both cross party lines.So in the Texas education system, the public schools have been closing, which really impacts a lot of those lower engagement, maybe lower turnout propensity voters who aren’t super high income and can send their kids to—Bacon: And he’s pushing vouchers very hard in rural—rural voters often like their schools even though they’re conservatives on other issues.Morris: Yeah, they do like to go to public school, especially if there’s not a Christian academy in their tiny town of 2,000 people, which is where I’m from. So that is potentially an issue that crosses party lines, and it’s an incumbent performance issue. You don’t want the schools to be closing and people to be mad at you for not having the schools open. This was a chief criticism of Joe Biden and his response to COVID, was that the schools were closed. That’s potentially bad for him.The big one is data centers. Greg Abbott has seen how toxic the data center push has become and completely flip-flopped on this. The issues here might be enough. They might be anti-incumbency enough, anti-leadership enough, for him to not get that fourth term. And four terms in a row, that is a lot. Even in a state like Texas. I think Rick Perry only had three, although he bowed out, so maybe he’d go to four. I don’t know.And basically the only way to get enough Republicans to vote for Gina Hinojosa, his challenger, is for that anti-incumbency bias to sway a lot of Republicans either to vote for Gina Hinojosa or to not turn out to vote.The other factor you might ask me about next is, does Ken Paxton’s bad campaign weigh on Abbott? It’s interesting, Ken Paxton is saying that’s going the other way around, that Greg Abbott has been bad for his campaign, in some of this leaked audio, if I’m remembering correctly. It’s like a fun blame game being pushed around here by both of the Republican nominees.It’s possible that there’s some voters who just won’t turn up at all because they don’t have a Senate candidate to vote for. I think the likelier thing is that there’s an undervote. That there’s a good number of Republicans that turn out to vote in Texas, and they vote for Greg Abbott, and he gets in, but they leave off voting for Ken Paxton, because they don’t find it morally justifiable, or for whatever reason they don’t like the guy. This is what a lot of voters said that they would do in, I believe, a New York Times focus group of Texas last week. So there’s some anecdata that this might be brewing, an undervote in Texas.Bacon: Last thing, I’ll make my own comment here, which is, it looks like Roy Cooper and Ossoff are going to probably win by more, and Cooper particularly is more a center moderate. He might win by more than the Democrats in Minnesota, Maine, or Michigan do. So I think you’ll hear a lot of commentary, “This proves moderation is coming.”Morris: Oh my God.Bacon: And you’re already seeing that.Morris: Yeah, the groundwork is being laid for this. People are sending me stuff about this morning.Bacon: I do want to say one thing I’ve been thinking about. I’m on the progressive side, I wouldn’t hide otherwise, so that’s part of it. But I do want to note that it looks like one of the biggest gaps is going to be between Keisha Lance Bottoms and Jon Ossoff, and Keisha Lance Bottoms is very close to Joe Biden, quintessential moderate.So my view is, different candidates do better. Roy Cooper has been elected like 1,000 times in North Carolina. He’s a very good candidate for reasons that are not just about ideology. Keisha Lance Bottoms has the same ideology, but maybe is not necessarily as popular, has not run statewide. So my point is only, elections are more complicated than just ideology, not that ideology might not matter. I’d be curious how you respond to that. And also, this sort of micro fixation on who won, comparing individual states and so on, is a little bit tricky. But I want to just see how you respond, and we’ll close there.Morris: So there’s another great reason why Democrats might be doing poorly in Michigan and Minnesota. Or in—oh, the New Hampshire governor is a Republican. Kelly Ayotte, right?Bacon: A Republican in this case, yes.Morris: Yeah. So definitely this applies to Michigan and Minnesota, however. They have Democratic governors.Bacon: New York too.Morris: New York’s definitely—this also applies to New York. They have Democratic governors.If what we’re seeing is a broad anti-incumbent backlash at the federal level, anti-Donald Trump—and this is voters are really mad about prices, about the federal government still being involved in the war in Iran—and they’re just reflecting that anger at the party in power. Who’s the party in power in Michigan and Minnesota? It’s Democrat.So maybe I’ll put it this way. That is a potential explanation for why El-Sayed and Peggy Flanagan are doing poorly compared to expectations in Michigan and Minnesota, and why the Democratic nominee for governor of Michigan—I’m forgetting her first name. Is it James Benson, or Benson—Bacon: It’s Jocelyn Benson.Morris: Oh yeah, the Republican’s name is James.Bacon: John James is the Republican. Jocelyn Benson is the Democrat.Morris: Why she might be doing poorly compared to your expectations based on the national environment, and how left-leaning or right-leaning compared to the nation as a whole Michigan is. Actually, that race, the governor’s race, could end up being right on the dot. But that is a potential explanation.And people have been asking about the moderation versus progressive thing for so long, because I have put myself on the side of the empirical political science on this, which is that we just don’t have enough evidence—in out of sample tests in the future—that a progressive or that a moderate would significantly outperform the average candidate that would be put up by the Democratic Party in these races. Sure, of course you don’t want to run AOC in Kansas. That’s not my argument. But compared to the average type of candidate you might get in these places, there’s just not enough elections to really say either way, especially in our modern age of high polarization.And what I’ve been trying to invite people to do is to see elections as products of multiple features of candidates, the way that voters see them, rather than just seeing politics in this one-dimensional moderate versus progressive, or thinking broader left versus right dynamic, that we just know from voter psychology, a lot of voters don’t possess the same lens or the same conceptualization of ideology.Bacon: People voted for Obama and Trump in consecutive elections. That would tell you that they’re not thinking about ideology in some strict way, or even race, or even other things.Morris: Or also that their ideology can change. That the features of ideology can change. What makes people a moderate or progressive changes over time. And just the very fact of that change means we should be less confident in our narration about, I think, why people are making their decisions and the ideological conceptualization.But I would invite you, if Abdul El-Sayed loses, to view that loss not just as a, “Oh, he supported Medicare for All, he lost.” But like I was saying earlier, there’s a lot of other features of the man, Abdul El-Sayed, that might not be palatable to your average white Republican, potentially swing voter or low propensity voter in Michigan. And elections are multi-factor.Bacon: That’s not reassuring either, but it is more complicated. The fact that he’s Muslim—we have some evidence that being Muslim is challenging, and that is not ideal for the country I live in, but it’s more complicated than just Medicare for All. Probably it’s more important than Medicare for All in some ways.Morris: It’s also just truer. It’s just truer to the voter psychology and how voters are seeing these candidates. If you watch Fox News, the narration about Abdul El-Sayed is scary. His photo’s always up there. He’s generally brown looking, but they make him look super brown compared to really white Mike Rogers right next to him.Bacon: It’s not a fixation on, Pete Buttigieg supports a public option and El-Sayed wants Medicare for All. It’s not really in that nuanced way.Morris: Yeah, the way that you and I are placing candidates potentially on the left-right spectrum, we just know that’s not how many voters are engaging in politics. Actually, most voters.I will leave you with one point of data, which is, in a survey that I have yet to release, about 30 percent of the public cannot name the party that is currently in control of the House of Representatives, and cannot say which party is the more conservative political party in America. They pick the wrong option. They either say Democrats are in charge or Democrats are more conservative.Bacon: You said 30 percent of likely voters?Morris: That’s just a political knowledge thing.Bacon: It’s just a political knowledge thing. The likely voters number probably is better on that than this, which is all people, right? So many people don’t vote in this country.Morris: This is all people. Yeah. Maybe the likely voters have slightly higher levels of political knowledge. We can check that. But there’s still going to be a significant number of people out there who lack these key concepts that we are all using in the commentariat to explain politics. So a lot of voters don’t view politics in that way.Bacon: Elliott Morris, great to see you. Congratulations. MSNBC is going to start using some of your data and analysis on their website and on television. That’s a great coup for you.Morris: MS Now.Bacon: Yeah, MS Now. Glad you corrected me there. And also I encourage people to subscribe. It’s Strength in Numbers. It’s one of the best Substack newsletters that I can think of. I learn a lot from it. I encourage other people who do Substack letters to be as good and add information, not just opinions, the way Elliott does. So congratulations on all your work, and great to see you.Morris: Oh, thanks, Perry.Bacon: Bye-bye.