Chris Hare
Description: Christopher Hare is a Professor of Political Science at UC Davis, whose research uses statistical modeling to understand voting behavior, public opinion, and political campaigns. In this episode, we cover everything from the challenges in political polling to how the decline in diverse social networks is deepening political divides in the U.S. Professor Hare also shares insights on Senate voting patterns, tracing the roots of political division back to the mid-1960s. We conclude with Professor Hare's thoughts on the power of statistical methods for political science and the importance of statistical literacy for anyone looking to understand electoral dynamics.
Website:
Publications:
Favorite Pollsters:
Marquette University Law School Poll
Ann Selzer & The Des Moines Register’s Polls
Mentioned:
Show Notes:
[0:03] Introduction to Political Science
[0:54] The Journey Into Campaigns
[2:56] Education and Methodology
[4:24] Challenges of Modern Polling
[6:40] The Impact of Polling Errors
[9:12] Polling and Data Technology
[11:01] The Role of Bots in Polling
[13:53] Understanding Political Polarization
[19:12] The Evolution of the Two-Party System
[22:48] Voting Patterns and Their Implications
[27:58] The Dynamics of Swing Voters
[35:22] Identity Politics and Moral Authority
[36:31] Historical Context of Political Shifts
[40:07] The Role of Political Elites vs. Masses
[44:55] The Influence of Political Sophistication
[53:49] Discussion Networks and Political Interaction
[56:51] Political Homogeneity and Its Effects
[1:00:10] The State of Political Discourse
[1:07:24] Strategies for Engaging Swing Voters
[1:11:56] The Future of Political Modeling
[1:19:53] Navigating Statistics in Political Science
Unedited AI Generated Transcript:
Brent:
[0:00] Welcome, Professor Chris Hare. Thank you for coming on today.
Chris:
[0:03] Pleasure to be here. Thanks for having me.
Keller:
[0:05] We'd love to start off by hearing a little bit more about what got you interested in political science and how you ended up at UC Davis.
Chris:
[0:10] Yeah, so I was a campaign junkie pretty early on. I started working for local political campaigns. I guess I was 12 or 13, so pretty young. And just always love a few things about campaigns, namely the kind of intersection of human psychology, which is endlessly fascinating. And kind of, you know, military strategy or sort of sport strategy, right, that you have limited resources and you got to decide what, how do I sort of allocate these resources and what kind of posture do I take where I maximize my advantages and, you know, look to, or at least think that I maximize my chances of actually winning the election. So I left political campaigns. That's what got me into this whole world.
Chris:
[0:55] I, you know, one of the things that always really frustrated me about campaigns was that when you're making those decisions, you rarely, if ever, have good data to go on. So, my question is about how much should we spend on yard signs? How much time should we spend putting up yard signs? Do yard signs work? Well, no one really knows. And certainly, you know, in those campaigns, no one had any idea beyond just some stories that, you know, oh yeah, there was this one time this candidate came out of nowhere and caught fire because hit up these really eye-catching yard signs. And it's just anecdotal. So that's what got me into the second half of what I do, which is statistical methodology, trying to actually use data to tell stories about the world and understand what's going on in the world and at least making an attempt to answer questions like that, which are still really difficult, even with fancy machine learning methods and statistical methodology and great data. It's still hard to get it. What is the isolated partial effect of a yard sign? But it at least gives us a better shot to be able to answer questions like that.
Brent:
[2:01] Yeah. And then when you did your education, did you start in poli-sci or did you go more straight to statistics?
Chris:
[2:07] Yeah, I did poli-sci and it took me a while to figure out, oh yeah, because I don't come from a really strong math background or interest. I was okay at it, but certainly nothing set the world on fire. But yeah as I did political science undergrad and started, slowly but making these connections of like oh yeah these statistics you know and polling helped me a lot too like polling is a classic example that you always use in any intro to political science methods course because it's, uncertainty right you grab out a thousand individuals and you want to extrapolate to a population of 330 million, in order to do that you've got to be able to, come up with things like margins of error, you got to rely on probability theory
Chris:
[2:53] to make that kind of statistical inference. So yeah, as soon as I started putting together like, oh, here is like the magic eight ball that can tell me, you know, about at least give me some insight into what works and what doesn't during a campaign, I'll suck it up and learn the math and, you know, and do the derivatives and whatever, just as in service of getting a better substantive understanding of what's going on during political campaigns.
Keller:
[3:22] And diving a little bit further into the methodology, what are some of the big issues we have with polling now? And what are some of the potential solutions to make them stronger?
Chris:
[3:30] Yeah. So this is a very fundamental problem, which is garbage data, which garbage is obviously that's a little bit harsh, but it's very hard to get people to answer polls for a few reasons. Some people are just less willing to do that. And of course, right, landlines are, I mean, just, yeah, great. Yeah, exactly. Not even less common, just increasingly obsolete. And so to the extent that even if you have, you know, millions of people who have landlines, which you do, is that really representative? And so you've got to figure out other ways to get people, you know, get the share of the population that is not willing to answer polls, which is both very large and not representative. It's not random who has a landline and who doesn't. And it's not random who's willing to answer a poll and who is not.
Chris:
[4:20] So, you know, we used to have response rates, you know, that you would just random digit dial. And in the payday, you had like 20% response rate. You had a one in five chance of getting someone to answer. It's not like 4%.
Chris:
[4:33] And so it's really expensive to do these polls, increasingly expensive and time consuming. And at the end of the day, there's just this question of, who are these people that are doing it? We know that they're not random. And we also know that because the thing is, like, if they were random, it would just take a lot longer to do polls, and we could live with that. If it's not random, then we got to do reweighting. So reweighting is something that, you know, modern survey, you know, polling research has always.
Chris:
[5:05] Done, right, that you have a target in the population of 52% of voters are going to be female. And in your sample, you collected 47% of responses from females. And so you just got to add a little bit of a weight to that. Nothing controversial there, nothing funny. But now we're having to do that increasingly by political targets, right? Republicans and especially Trump voters being less likely to answer polls, distrust of media, lack of social capital, lower educational attainment, whatever the reason is, what matters is the end product, which is just it's hard to get them. And so you've got to put a weight on that. Problem is, is how much of a weight do you put on that, right? You don't want to overweight it, but you don't want to underweight it. And that's been really the crisis over the last eight years in response to 2016 is trying to fiddle with those weights to get the right mix that we can turn this data, which is decidedly non-random into something that is more randomish or at least representative of the American electorate. But that involves a lot of subjective decisions about what is the actual electorate going to look like? People that turn in ballots, how many of them will be self-identified Republicans or how many of them will have a favorable opinion of Donald Trump or how many will be college graduates?
Chris:
[6:30] If we knew that then this wouldn't be a problem but of course we don't know
Chris:
[6:34] that we won't know that until after the election so uh it's a real catch 22 uh 22 issue that we.
Brent:
[6:41] Have yeah i think you briefly mentioned it before but just overall like most polls are what a thousand to a couple thousand people and we extrapolate that out to represent the 330 million exactly yep yeah um some even smaller some.
Chris:
[6:54] Are you know 700 800.
Brent:
[6:55] Yeah i think that's really funny because people put so much weight on polls just because it gives an easy number to read, but it's actually like understanding a bit of the background, I think is helpful and why we had outcomes like 2016, where everything went one way, but then the election went the other.
Chris:
[7:12] Exactly. Yeah. I mean, I'm a defender of polls in that, you know, pollsters are doing the best they can with data that is increasingly garbage. And, you know, polls most of the time sort of work pretty well. I mean, it's like a kind of a fuzzy picture. Like it's still, it's not right or wrong, it'll show some things, but it's not going to show the level of detail you need in a very, very tight election, which, oh, by the way, every national election now is razor thin. So we need the level of precision that gets us down to like 75,000 voters in Pennsylvania. We don't have that. We didn't have that before, and we definitely don't have it now.
Brent:
[7:44] Do you have a favorite pollster or any groups that do the polling best?
Chris:
[7:47] A few. The Seltzer Group out of Iowa, Ann Seltzer and her company are pretty famous. They have a really amazing background they they do an amazing thing of uh like in 2020 they had a poll and it was like a small technical issue where they they didn't dial like like some of their sample wasn't truly random and they said we're just taking the whole poll out like we we care so much about our reputation that we're not going to put this product out even though that's going to be embarrassing, um and is also going to cost us the chance to be able to put out this poll uh that they have such standards that, willing to do that, which, you know, that's always a good sign because mistakes happen.
Brent:
[8:33] And with both proposals.
Chris:
[8:34] He just got swept under the rug. So they're good. The Marquette University poll, Charles Franklin is a pretty famous political scientist. He's been at a few different places, but he and some others run that. Those are really good. The problem is that they're few and far between. They don't do regular polling. And so, yeah, I mean, really, the best answer that we found is you can make do with a bunch of so-so, you know, C, B level polls if you combine them together and cross your fingers that the
Chris:
[9:10] errors cancel out, which to some extent they always do. You know, one just oversamples Trump voters, one undersamples it. There's definitely random error. And if you can just do – I mean, that's how Nate Silver made his name is just averaging. He does a little bit fancier than that, but I mean, really the power is just, you take a bunch of polls and you combine them together and hope that that noise at least partially cancels out.
Brent:
[9:31] Yeah. And then do you know if big tech firms are like starting to help out researching the polls? Because I think they have a lot more data on like, who are you watching more? What political advertisements are you engaging with? And those types of things. Are we going to start, like maybe not this election, but in the future, start pulling from that data?
Chris:
[9:46] Yeah, it's a great, it's a great question. I mean, the other thing that we're starting to do a little bit now is use things like ChatGPT or other large language models to simulate respondents and ask them, you know, imagine you are this person, you have these characteristics, who are you likely to vote for? And so you're seeing more of that, with all kinds of data. I don't know, but it is at least interesting and you could see it going in that direction. But now, I mean, your point is much more tangible, which is, yeah, if you're the social media companies, you're sitting on gold, you're sitting on real data. We don't know. I mean, my suspicion is they found ways to make more money out of that than they would from releasing polls. But you're absolutely correct that they, you know, if you had your hands on that data, you could come up with some pretty useful pictures of the electorate. It's really hard. It used to be actually better, I don't know, 10 years ago, I guess, Facebook or Meta would be more willing to share their data with social scientists. There's still some of that, but it's a lot more guarded and really valuable stuff. It's kind of this Pandora's box, this mystery of, they have all that data there, presumably they would use it, but to one end, I don't know.
Chris:
[11:02] They certainly don't seem to be releasing any kind of public polling, and I don't think they have a huge incentive to do that, but they certainly could.
Keller:
[11:11] Is there an issue with the polling of bots or people that are maybe paid actors going in and going through the motions or intentionally putting in data that might be completely false?
Chris:
[11:21] Yeah. Thank you for mentioning that. I should have mentioned that. Yeah, that's another, not only do we have declining response rates, but how do you deal with the fact that no one has landlines anymore. Well, obviously, you do internet survey research, which can be great. Obviously, that, you know, it's kind of funny. I'm old enough that I can remember, right? It used to be the internet was like, oh, yeah, you were the sort of, like, that was the outlier. And that landline, you know, John Q. Public had the landline. That was how you got real American voters. And the internet was like this weird, unrepresentative thing that that has since of course crossed over um uh so like given the ubiquity of of the internet and the number of people that just take surveys um which is really interesting to me like who signs up to do these things i mean it definitely ain't the money um is it is it i mean is it just a nice little side hustle that'll get you a few few extra bucks is it that you have just this innate need to you just like filling out surveys like making your voice heard um, So there's whole questions about how representative is this group anyway, because it definitely ain't random, and maybe even more non-random than the old school landline. The other thing is that, yeah, of course—.
Chris:
[12:42] We'll pay, like if I want to do a thousand person survey that would be short, that's going to be, you know, 3,000, 5,000. I mean, that's the academic rate. So that's a little bit of a lowball. But the point is these things cost money and that you definitely have an incentive to, you know, for these firms to use bots. Um, or, or even if it's not the firm themselves that they can be victims and individual participants just have, you know, a server farm somewhere, uh, in Kyrgyzstan that's, that's running day and night to just randomly fill out these surveys or, or even worse, like create some intelligibility, um, uh, to that. You know, we can catch stuff like straight lining where you just answer.
Brent:
[13:21] You know, support, support.
Chris:
[13:24] Support, support. um but yeah as they they get clever that's kind of an arms race um and it's getting harder and harder to pick out the just fake responses um so yeah i mean it's it's considering all that i think the polling industry is doing pretty well um but i mean at the end of the day we're still left with the fact that like you can only do so much with uh with data that's fraught with all those kinds of issues.
Keller:
[13:47] Yeah and like polling is obviously used to like model political situations part
Keller:
[13:51] of that in the U.S. right now is polarization. So we definitely want to dive into that a little bit further. I think to start that, could you just give us a definition of mass conflict extension, kind of what that dynamic is, and maybe like a timeline of where you think that's kind of started to grow to progress where we're now?
Chris:
[14:08] Yeah, yeah. So, you know, polarization, obviously, this is, you know, the big topic and has been for, you know, 20 years in American political science research. It started out, you know, that, you know, that it was exaggerated. And then people said, well, okay, yeah, the elites, you know, members of Congress and party activists, yeah, they're polarized, but the public isn't really polarized. And now that has shifted into, okay, yeah, the public is polarized, but how? So problem with that is so often the case with academia is a definitional debate. And you've got, you know, like a dozen different definitions of what polarization is, none of which are wrong and none of which tells the complete story. I focus, and appreciate your looking over the conflict extension stuff, because for me, that's a really overlooked part of polarization. Conflict extension is, I think, a pretty common, you know, straightforward, common sense notion of how, you know, it's not just how much you're disagreeing, but how frequently are you disagreeing, right? How many different topics are you disagreeing upon, right? And so You have a friend that you fervently disagree with.
Chris:
[15:17] Movies, right? You love Tarantino and he hates Tarantino. Um, and you really, really, you know, you can be polarized on that or you can, you can not care about it that much. So, I mean, that is a kind of polarization for me though. What's more sort of behaviorally significant is are you also disagreeing about where do you go to eat? Where do you go to vacation? Right. Where do you hang out? Um, uh, which bar do you go to or do you go to a bar or not? You know, I mean, And it's the more that you're disagreeing about. Yeah, I'm a big sports nut. And so I tend to think of it in those terms that if I can find someone that. So I'm a Saints fan.
Brent:
[15:55] Probably don't.
Chris:
[15:56] So like Falcons probably are number one dislike team. I mean, a Falcons fan. I'm not so nuts. On the other hand, I went to University of Georgia for PhD. So if I know that they're a Bulldog fan as well, it's like, well, OK, that kind of softens the blow of the Falcons thing. But if we're fighting about college football and pro football and NBA, that's really significant, even if on each of those individual issues, we're not disagreeing that much. The fact that it's repeated lines of conflict, and that's a big part, at least I argue, of the story with American public opinion and the way in which it's polarized over the last 20, 30 years is that you used to have these what we call cross-cutting cleavages or cross-cutting divisions that you had the – like me, I'm a Saints fan, but I'm also a Georgia Bulldogs fan that break up the lines of conflict. So it's not just about, ooh, I hate Georgia teams. Well, I like some Georgia teams, but I hate others.
Chris:
[16:56] And so, you know, with our political parties, America's a big heterogeneous country, geographically, culturally, everything else. And we only have two parties, which is kind of odd. And so in the past, right, the one thing that has helped us in dealing with things like polarization is that you had, like, by necessity, these cross-cutting cleavages that come up. Like, the Democratic Party couldn't just be a Southern party or it couldn't just be a Catholic party. It had to be a mix of a whole bunch of races, geographic bases of support, and so on and so forth. So that when Democrats were fighting Republicans, there were all these shared interests, right? There were some, right, you had some white representation in the Republican Party, you had some white representation.
Chris:
[17:50] What's happened, though, is that increasingly, you know, religious, gender, and definitely policy lines, so demographics, policy, sorting is occurring. Democrats are more uniformly, you know, take similar stances ideologically, are more and more demographically similar. Same thing with the Republican Party. And so you have these groups that are, you know, there are all these fault lines or possible tensions with or active tensions that are butting up against each other at multiple pressure points. And so one of the things I've probably seen some of this – talking about things like affective polarization, that it's emotional. Democrats and Republicans just like each other less. Well, okay. I can certainly buy into that. But there's a pretty, I think, straightforward policy explanation for that, which is Democrats like each other less – Democrats like Republicans less because they are disagreeing with them about a whole bunch more things. And there's less common ground. And same thing, of course, Republicans and Democrats. And so that, for me, is the contribution of conflict extension here, which is it's not just about the extent of polarization, but also about the
Chris:
[19:06] kind of consistency of polarization across a whole bunch of different points of difference.
Brent:
[19:13] Yeah. And then before we dive deeper into like polarization itself, I think you briefly touched on it there, where it's like the two-party system is kind of inherently like pushing us towards polarization. Could you give us some history on like maybe where that two-party system came in, how in like the 60s and 70s, like polarization really started to develop and like all of those like factors that kind of preamble to where we are now?
Chris:
[19:33] Yeah, it's a pretty crazy story because you definitely zoom back out far enough and you start to see these cycles of polarization and depolarization, even though across virtually that entire time span, few exceptions, things like air of good feelings, we've had a two-party system. So the reason we have a two-party system is certainly not by purposeful design, but rather we have a certain set of electoral institutions that make it very difficult for third parties to enter in. So it's not just about things like ballot access, although that isn't helping third parties, but this Achilles heel, this really fatal flaw of third parties of being the spoiler. So, yeah, you have two candidates running and then you have some third candidate and you can think about it as like businesses along a street. There's a one street town and the metaphor being that that could represent something like ideology, but we just think about it as physical distance. You got to go to the drugstore, you go to the one that's closest to you. Well, the candidate should pick points or the stores should pick points that optimize their.
Chris:
[20:51] Or excuse me, when I say optimize, minimize their distance between themselves and everyone else. And so what that predicts is actually convergence to the center. You should see them both place their businesses right at the median point on that street, one just a little bit to the left and the other just a little bit to the right, which we actually see all the time, right? Drugstores, there's a CVS, and then Caddy Corner from.
Brent:
[21:11] That is a.
Chris:
[21:11] CVS. Now, we don't see that, but what we do see is that you have parties that represent well enough these two groups and there's really no place for a third party to come in and take votes without supplanting one of the other ones right so you could come into the left of the left-wing party right like a green party candidate but um it's then you're just splitting that vote uh right so the the spoiler effect in a majoritarian system where all you have to do is win um right if you win by 1% or you win by 40%.
Brent:
[21:48] The winner takes all.
Chris:
[21:50] Exactly, yeah, first past the post, majoritarian winner take all. Yeah, so versus like a proportional system where you get 15% of the vote, you get 15% of the seats. So that's really the, that's not the only explanation, but that's, it makes it really, really, really difficult. And you can show with game theory, irrational for people to vote, And so we have this situation where we have two parties, which should converge to the center, and there are those forces, but there are other kinds of forces that make them pull out. And that's the mystery of why did you start seeing in the 60s and 70s this polarization? Why would the parties start to move further apart when it's, as a technical matter, in their interest to move closer to the center in order to gain more votes?
Brent:
[22:39] Yeah.
Keller:
[22:40] And looking at that widening gap, how did you, when you were studying politicians' voting patterns, how did that inform your understanding of why that gap has
Keller:
[22:47] been widening leading to today?
Chris:
[22:49] Yeah, yeah, it's great. And even that, like, it's a great question that even that only gives kind of partial answers, because it's definitely, you know, what you saw was that you, the new members of Congress that were entering in the 60s, 70s, 80s, 90s, that they were almost always more extreme than the representatives or senators that they were replacing. So you almost always had changes that were towards the extremes. And so it's more of a story of generational replacement versus conversion.
Chris:
[23:27] And so, yeah, then the question is, so why did districts start electing more liberal or more conservative politicians? And that almost has to – like the most proximate cause of that has to be things like party primaries. Now, this is kind of a turtles all the way down problem because it's like, well, then why did primaries start preferring more ideologically extreme candidates? And then you could talk about things like social forces or economic forces like income inequality or changes in the role of party activists. So you can keep going with that, but it definitely seems like it's a story of new members of Congress are elected more on the basis of ideology than electability, and also in terms of service to the district. Right that that and this is also like part of a social cultural and economic change washington dc used to have a much bigger discretionary budget right you as a member of congress you could get a whole lot more pork spending and pull that into your district as entitlement spending has become larger and larger share of of the the money that we take in social security medicare medicaid um Uh.
Chris:
[24:38] We, um, well, Medicare, not the Cade, we, we, there's less port to go around. And so when you're, it used to be right, reelect me, I bring the bacon home, uh, versus, you know, we built this dam and we built this levee and we built this, this road project. Um, the South was very, I mean, it's not the only one, but, uh, uh, you know, we built, Members of Congress from the South that were there for 50, 60, 70 years, their entire lives. And they kept getting reelected because they would bring the jobs and the money back. As that has become less tractable, then you start thinking about, well, why am I voting for this person in Congress? And it has to be more of an ideological representation story. I don't think that that by any means is the entirety of it. But that is one force that is just kind of an example of how you could imagine these things from outside forces. We have an older population. We had to start spending more on Social Security. That can have these very unforeseen consequences that pop up, like polarization.
Brent:
[25:41] Yeah. And then could you dive a bit more into what the findings of seeing the voting patterns of the senators were and how the Republican Party might be a little bit more diverse but still for the right while the Democrats might be a little bit more unified and those types of things that were influencing that divergence? Yeah.
Chris:
[26:00] So, you know, there's a whole methodological underbelly to this that I won't bore you to tears with, but, you know, different ways of estimating what those scars are. And you can estimate a model where you say that a legislator has a single left-right liberal conservative position throughout the entirety of their career, but you can also have these dynamic models that says, oh, they can shift around. There's even sub-variants of that. Yeah.
Chris:
[26:28] So avoiding all of that, the findings are still pretty consistent, which is, yeah, we definitely saw this gap open up during the mid to late 1960s, and that has continued essentially unabated since then. With Republicans moving to the right more quickly than Democrats. So there's asymmetric polarization in that both parties are getting more extreme, but Republicans have moved at a faster clip. But although the Democrats in the last few cycles have started to catch up a little bit, and you have these – what really ends up happening is during these wave elections, 2010 is a really good example.
Chris:
[27:09] It was a slaughter for Democratic incumbents that year. It was on the hills of the Affordable Care Act. But most of the Democrats who lost represented these marginal districts, oftentimes in the South. They were blue dog Democrats. They were the moderates. And so the Democratic Party is not even necessarily that they're moving left after 2010. It's just that all their moderates got replaced with Republicans who, by the way, were also pretty conservative. So, you know, there's a little bit of a story of like movement, but it's also a story of who is going to be the most vulnerable, who's going to switch in years. You know, 2018 was a really good year for the Democrats. Who got voted out that year? Wasn't Republican hardliners, it's Republican moderates. So you have these centrist members that bear, that had to sort of pay for the sins of their party, the ideologically extreme sins of their party.
Chris:
[27:59] And so you have this – some people call it a leapfrog pattern in terms of being solid liberal and then solid conservative because there are just fewer and fewer options in between.
Chris:
[28:13] So, yeah, I mean that's the basic story. It happens to track with some interesting pattern. The patterns of polarization we see in Congress happen to track – it could be coincidence. But some things like income inequality. As income inequality has increased, so has polarization. And it's pretty remarkable the extent of how well they correspond with each other. So some people focus on that. I tend to focus on – it's not that I deny any of that, but I think it's a little bit overlooked in political science, so I tend to emphasize it – the rise of cultural conflict, namely secularization. That as you have this real delayed reaction in America that post-industrial countries like in Western Europe tend to become less religious over time. That definitely happened in Western Europe first. It took America a little bit longer, but now we're really seeing that in a place like California that, And so there's all kinds of backlash effects to that as well, and it opens up a whole host of cultural-type concerns between the secular and the more traditionally religious.
Chris:
[29:21] And that, you know, again, in the conflict extension view of it, even if Congress isn't spending most of its time debating something like abortion, once you throw abortion onto the agenda, that just changes the tenor of politics, right? That just changes it into a more, you know, winner take all, good versus evil, right versus wrong, versus, you know, taxes where, okay, I want 34%, you want 38%, we'll meet somewhere in the middle.
Chris:
[29:46] Debates that could be fiercer than that, but at least kind of lend themselves a little bit more to compromise. Debates like abortion do not. And so the entry of those issues on the agenda, even if they don't dominate the agenda, they do, right? They change the kind of flavor and the feeling and the vibes around political conflict and why people's identities are more closely tied to Democrats and Republicans. You know, why people are less willing to swipe right for someone who's an opposite party of them, which is pretty remarkable, the amounts, but that is true. You know, people are just not willing to cross-date or cross-marry. Is that really just about, like, people care about politics that much more, or is it more a story of, well, politics has become imbued with all these other sort of signals about your values and your worldview and your, you know, about you fundamentally that become a more useful signal. And so you see all these, you know, people less wanting to be friends with or date or certainly marry members of the opposite party. That's why I care so much about conflict extension.
Brent:
[31:01] I know some part on it, like religious people, I hear arguments in this when they talk about politics where they view like God as being like the top authority and they'll have like ultimate control handed over to him. Whereas like if you're not uh religious you might have to start replacing who's that top authority in your life and then that gets into the role of the government and then do you think that part is playing a role in why so much of the identity is now being like influenced by the government and they're switching towards identity politics more and more even though it might not represent as big of a part of the legislature.
Chris:
[31:39] Yeah i think that's a great insight i i happen to agree with it um i think it's really good inside whether i agree or disagree i happen to agree with that you know yeah i mean we are uh i have a book it's probably somewhere over here but moral believing animals uh was the uh the name of it and it was by a um a sociologist and anthropologist and basically says you know look we humans are kind of weird creatures we're the only animal that at least that we know of uh that's aware of our own impending mortality um right and so we also have to live with that uh right we have some untold number of years to sort of figure that out. And his argument, Christian Smith is his name, and a few other people have done that, we had to create stories, we had to create meaning. And that necessarily implies this moralizing situations. And Americans seem like we have a strain in us that's especially like from our founding imbued within us, like this Puritan kind of belief of not only can we do better, we should do better, or even more than that, we have an obligation to do that and, you know, move towards greater justice, whatever that represents, instead of, and that's what being an American is about, that it isn't just about, you know, ethnic identity or cultural identity or geographic location, but it's about this very optimistic belief in the ability of, human society to move forward, to create actual progress towards justice.
Chris:
[33:08] That's fine, but yeah, so what that means is that really, if you buy into that, it's very hard to be truly objective in the sense of, yeah, you need an authority. You need a moral authority of some kind that historically, traditionally has been religion, but it doesn't have to be religion. It can be things like identity. It can be other kinds of things.
Chris:
[33:31] You know, secular beliefs and things like equality or, you know, a focus on oppression and which groups are oppressed versus others or, you know, you take a Marxist reading of that. And so, yeah, it seems like humans are just not – we are not built to be morally neutral. We have to have some sort of a moral vision and try to implement that. Of course, that doesn't mean we are – you know, we'd be horrible monsters. But horrible monsters don't tell themselves that they're horrible monsters, right? Horrible monsters justify to themselves in kinds of wicked and crazy ways. It seems like that's really speaking about something important about human nature. And so, yeah, I mean, if that's the case, in a lot of ways, that's positive. Like, oh, great, America wants to be better. The problem is we don't all agree on how to get there and what is better, you know. I mean, and that's the importance of all the, you know, people talk about, like, transgender rights and like, well, that's not really the bread and butter concerns of most people. And that's a side issue. Well, maybe. But, you know, this speaks to something very fundamental. I mean, one side believes that, you know, this minority group is being unfairly oppressed and denied something very fundamental, which is their ability to express themselves and their underlying identity. You know, and the other group, the more religiously traditional ones, sees this as something that is very damaging to a social order and a conversation.
Chris:
[35:00] That has evolved over time to deal with, you know, all kinds of problems and difficulties in crafting a free society that can operate together. So, you know, things like family values or traditional family values and all of that. So the point isn't to adjudicate there, but say that they're both coming from
Chris:
[35:20] a place where they think that they have the moral high ground. Yeah. issue versus the group that, you know, again and again. I mean, how often do you see a car with a bumper sticker that, you know, talks about my body, my choice or Planned Parenthood, but then also happens to have a, you know, anti transgender. You just don't see it. Obviously, those people exist, but that's definitely not the majority and it's certainly not the majority majority or even a significant block of those who were active in politics. And so, that's the whole conflict extension thing is that there's just greater consistency and constraint across these myriad conflicts, many of which are, you know, high heat, high temperature topics.
Brent:
[36:08] Yeah.
Keller:
[36:09] And have we seen that with the shift from religion to politics as kind of like the ideological home of a population, are we at a point in history where we've seen other countries get to this stage and have another, whether it be a group, whether it be the church, going back in full force and taking that kind of, that ideological and that moral guidance back? Or where does it go after politics?
Chris:
[36:31] Yeah, I mean, you know, it's, of course, in our own history, right? We had this war. I mean, that's the big peak of that sine wave, which, you know, is obviously undesirable, but also, like, seems a little bit far-fetched. I still tend to think that that's a little bit far-fetched as well, but that is a solution, which is you just have to fight it out every once in a while. I don't think that's the case. I hope it's not the case. But I mean, that is – if we're just looking for an obvious – okay, how have we solved this problem in the past?
Brent:
[37:09] Well – It's human nature.
Chris:
[37:10] It's probably human nature, and it is a solution that has been shown to actually work in the real world. So, yeah, you know, other, I mean, you have, like, in British politics, I'm not super familiar with, like, other comparative systems, but...
Chris:
[37:29] To the extent I know anything about it, it's probably in British politics. And even there, you have those kinds of cycles as well. The problem is, is that when you have periods of depolarization, like in British politics, you have an era of one party dominance. So Margaret Thatcher comes in in 1979, and the Tories just kick ass for, God, how long were they in the wilderness? Was it like 18 years, I guess, was 97? I mean, they just, labor couldn't do anything. And that forces you, that time in the wilderness forces you to.
Chris:
[37:59] Rethink things, right, and get your house in order, and certainly in that case, the Labor Party had to move closer to the center, Tony Blair, and all that, which a lot of Labor Party members absolutely hated, but they sucked it up because it's better to win with someone you kind of like versus, have yet another Tory government in charge. The issue with America is that the parties have done such a good job of maximizing, like, retaining their ideological commitments, for the most part, while positioning themselves for, like, a 50-50 shot. Like, they're viable, like, the wrong party and Democrat party, they are viable in every presidential election. They're, you know, at 240 electoral votes, and they just got to find a way to flip Pennsylvania. So it doesn't seem like there are any major electoral ramifications for being extreme. Whereas if you had a period where one party just came in and dominated, the other party presumably would have to move a little bit closer to the center. That's what happened with Bill Clinton also in 92, or at least that's a story you can tell. I happen to believe it, where Democrats just got...
Chris:
[39:10] Blown out of the room in 80, 84, 88, and you start to lose three elections in a row. You start looking in the mirror a little bit. And, yeah. So the hard thing there is that if you're – there isn't – it doesn't seem like there's enough motivation for either of the parties to, like, unilaterally disarm. Yeah. Because they're viable. Right? We don't – I mean, this election – and I hate saying it because it's such a trite thing that people say, and you hear it in elections, like, all my life, a few of the mills that we knew. Well, you just don't know. It's a coin flip or it's really unsure. It's too close to call. Meanwhile, like 2012 was not too close to call.
Chris:
[39:50] But it's like something reporters can say and no one really calls it on them. But this one really is. I mean, Nate Silver had a good post about this yesterday. I actually happen to like Nate Silver a lot. He's like, no, like 50-50 can represent ignorance or 50-50 can just represent the true state of the world.
Chris:
[40:04] And it seems like that 50-50 represents the true state of the world. And that's dangerous for polarization reasons because when the party that eventually does lose, there's not a huge reason for them in the aftermath of it to think, okay, we got to go back and reconsider everything. We need to do movement. We need to change this. We need to make reforms. We need to reconsider our positions. They're going to lose it by this much. And that makes it pretty easy to tell yourself, well, we actually need to be stronger. We need to be more committed and louder and more aggressive. Like if Trump does win and Harris loses, everyone will blame this kind of messy transition between Biden and Harris. And so you can come up with excuses to write off and say, we're going to continue doing what we're doing. We're not going to change anything fundamentally because this weird fluky thing happened in 2024 that's not going to happen in 2028. Um, and no doubt Republicans will tell, or, or, you know, the, that, that we did win the election, but it was just stolen from us. Um. Just like in any kind of relationship, you can always find reasons to justify your actions, and that is probably what's going to happen here.
Brent:
[41:18] Yeah. Do you think the political elite, the actual party members in office are the ones moving first, or do you think the actual masses, the citizens are becoming more polarized first, and then electing more polarized candidates?
Chris:
[41:35] Yeah. I don't think we have a great answer to that. And I mean, the trite answer is, well, a little bit of both. And okay, yeah, that's true. But I mean, I think kind of the answer that I tend to put the most faith in is that, you know, there's this group that's somewhere in between public and elites, that they're not members of Congress, but they're way more active and involved than the average citizen. citizen. Um, so these are party activists, uh, who it's, it's, it's, it's difficult. You know, sometimes they have official titles with the party, you know, they're the committee person for the democratic party of Yolo County.
Chris:
[42:14] They, but sometimes they, a lot of times they don't have official titles, but they're the ones who donate. They're the ones who show up for primary. They're the ones who volunteer. They're the ones who show up to party caucuses. Um, and oftentimes they're the ones who run for office themselves, or at least find people that will run for office, recruit candidates and things like that. And I think that a lot of the puzzle lies at that intersection of this nether region between voters and elites. And I'm certainly not, it's not like I came up with that idea. A lot of people have offered it. But I think that it is a little bit overlooked, um, their role in getting this process started. And you can, I mean, you point to all kinds of examples, Republicans in the sixties where the Republican party became increasingly dominated by conservative activists, um, anti-tax activists were a big part of this, but also, you know, um, groups that were concerned about, uh.
Chris:
[43:11] Things like social change and abortion and more socially liberal lifestyles, religious traditionalist. And then you have on the Democratic Party in the late 60s, right, the rise of the new left that was much more aggressive on things like equality, right?
Brent:
[43:33] Civil rights act.
Chris:
[43:34] Exactly, yeah. And beyond just class terms, right? Democrats have always been about equality, but used to be much more of a focus on economic equality class differences versus more of the identity-based ones of the late 60s and beyond, and also things like the environmental movement and that. So, you know, a lot of that story really comes down to these activists, many of which were like kind of written off at the very beginning of like, well, they're cranky, kooky, they represent 10% of the population. That's true. But they are unduly influential in a hybrid system where party primaries matter a whole lot. You know, 50% of the electorate votes on even the best of elections. Many elections are, you know, 40, 30%. So they have an outsized role. Yeah. So, I mean, I think it's very activist driven. It's probably a little bit more elite driven than mass driven. But voters seem to be perfectly willing to go along with it, right? It doesn't seem like there's a – I think there is a penalty to be paid, but it's slight for extremism, and it's not enough for –, people like Mitt Romney or whatever, like you're going to get voted out of a, if you don't retire, you can vote it out.
Brent:
[44:47] Lynn Cheney, an example of that.
Chris:
[44:49] Like you're much more in danger of losing reelection by moving to the center
Chris:
[44:53] than the extremes. And so that's the incentives.
Keller:
[44:56] And could you explain how varying levels of political sophistication impact, first of all, like actual participation in politics, but also ideological convictions and maybe like the not centralization, but the unification of these ideological beliefs in politics?
Chris:
[45:11] Yeah, yeah. So it is a pattern that is virtually always there, and usually it's quite strong, which is the more you care about politics, the firmer you are in your beliefs. So the more ideologically consistent you are, so all of these things, knowledge, ideological constraint, passion, level of participation, all of those things hang together really well. There can be differences between them. They're not perfectly correlated, but they're all really strongly correlated, and that's been true as long as we've been tracking this stuff. And the huge downside there is that we have this – what we would love, what the founders envisioned was this combination voter who is highly informed and cares about politics but is also neutral. We just don't see evidence of a whole lot of that. People that are meaningfully swing voters, but also have very large amounts of political information and education. Sophistication is kind of what we can wrap all that up in. And so, yeah, that's where you see these interesting kind of results. A lot of people say depressing results of the more informed you are, the more prone you are to misinformation. That you're actually more gullible because it's not really that you're more gullible. It's that you're.
Chris:
[46:32] You have more skin of the game. You have greater convictions to one side or the other, and so you're more likely to immediately dismiss anything that's negative against your side or for the other side, as well as accept information that benefits your position. And people that are more informed seem to use their powers for filtering rather than objective evaluation, and that is also just a basic story of human – we like being told we're right. We really don't like being told we're wrong. So, you know, this idea of getting voters more informed is great. But if that information comes at the cost of extremism, which we know that it does, it's kind of a pick your poison thing, right? Would you rather have this moderate, squishy electorate that doesn't really care how much or the super engaged, you know, patriotic way, you know, Uber involved citizens that hate each other and won't listen at all to the other side? And of course, we have some interesting hybrid of that right now. But yeah, it does seem like we just empirically do not find evidence of that voter that, by the way, we always assumed did exist. The founders assumed that that voter would exist and that that would actually be the dominant style.
Chris:
[47:43] This was a huge crisis that came out when we started first, like in modern survey research, when you started doing polling in the 1940s, 1950s, good polling, where 40% of people can't name the vice president. Um, you know, 60% of people can't even name a single Supreme Court justice or, or even a branch of government and all like that. And so, and then you say, well, okay, that's okay. As long as the 40% of people who can do that, um.
Chris:
[48:11] A fair-minded, well, it turns out, no, that 40% is disproportionately on one side or the other. And it makes sense that, right, if you care a whole lot about politics, you probably have, you know, I don't care a whole lot about NFL football because I don't have a team. I care a whole lot about NFL football because I do have a team. And that's going to motivate me. So, yeah, this seems like it's a real, like, fundamental, inescapable trade-off that exists and democracies that can be managed. It's not apocalypse game over, but it definitely is not the rosy story that we were sort of taught. And it still kind of carries over in civics education when you're talking about democracy and all the stuff that's great about it of like fair-minded citizens can come together. Well, yeah, true, but a lot of asterisks on that.
Brent:
[48:59] Yeah, and I definitely think now with social media, it's easier to be told you're right more than ever before. But on that note, like the fair-minded citizen, do you think they are just becoming apathetic because they see both sides just going so far to each respective side that they just kind of disengage from politics, even if they might be informed on what is happening within politics?
Chris:
[49:17] Yeah, there's some evidence of that. I mean, you know, levels of information are lower. So the disengagement, you know, once you start not caring, you also tend to not follow politics as much. But there is still variation in information. You know, there are people that still do know what's going on. They can name the vice president, for example. They may not be uber involved. And, yeah, I try to make this point about – because I believe swing voters are a thing, you know, like a meaningful force in politics. It's not a huge amount of voters, but it's still large enough to make a difference. But, you know, people – the usual story we say with swing voters is, oh, yeah, there's – they don't exist anymore because the parties have such clear stances. Well, they definitely declined, and the parties having clearer stances has made it easier for people to sign. But right if you know we're trying to decide where to go to dinner and we hate both options right if it's a choice between like a you know a subway uh with um uh with a food violation.
Brent:
[50:15] And like a um.
Chris:
[50:16] You know an only mushrooms restaurant.
Brent:
[50:18] I hate mushrooms so.
Chris:
[50:19] Well that's very clear like i know what those choices are like one has a health violation and the other i'm eating mushrooms but that i'm still on the fence because like i don't like either one.
Brent:
[50:28] Of them yeah.
Chris:
[50:29] And so i'm still a swing voter. And that clarity has not helped me at all to make a decision because both are out of my purview. So yeah, I think that's often overlooked. That story is a little bit too simple when we say no swing voters because the party brands are clearer. That can help some voters, but for some voters, that's actually going to have a, making them just hate both sides that much more.
Brent:
[50:54] I think that's why I kind of believe that's why RFK took some traction, especially with like younger people that we've talked to in our groups. Like it's, we all kind of hate the Republicans and the Democrats because we see what like the politicians are doing as like, they're kind of doing it with just different colors on. And RFK kind of came in as like a separate voice to a lot of people that kind of like, he's listening to both sides. Whether or not you agree with every stance he takes, It just kind of felt like some sort of like middle ground for once. And I think that that's where I feel like personally, I see a lot more just disengagement, whether or not you're informed. But then like when an option does come up, that's somewhat in the middle. You're just like, oh, wow, that's refreshing. But it's interesting to see that data doesn't really fully back that. Or does it? It does.
Chris:
[51:42] I mean, there's definitely a lot of people like that are, that just hate both sides. That is definitely true. They tend to, though, be less informed.
Brent:
[51:54] When you say informed, do you mean like knowing politicians' names and like those types? Is that the easiest metric to track or is it like, okay, this is policy that they provided, like this was the outcome of it, or this is like the stance I want to see the country go, whether or not policy has reflected that yet?
Chris:
[52:10] Yeah, there's a whole debate on that, and you exactly sort of foresee that exactly. We usually do – the easiest metric is these like factual questions, which a lot of people take issue with as far as like, well, who cares if you know who the Speaker of the House is? Like you might still be caring about news and politics. For me though, it's – and others like when we do other kinds of survey instruments that get a little bit more involved, like for example, just letting you have the floor, open-ended questions.
Brent:
[52:40] You know, what.
Chris:
[52:41] Do you see as – like what do you think about the Republican Party? What do you think about the Democratic Party? And then you rate how sophisticated that is, which is tough, but you can come up with some measures there. It turns out that they're all really correlated with those knowledge questions. So yeah, there is nothing special about knowing the name of the chief justice of the Supreme Court. It just turns out that if you are actively following news, you're more likely to know that. And so it's not perfect, but it's a quick, easy, and pretty reliable way to gauge who's paying attention, who's following.
Brent:
[53:15] The use of these kinds of factual questions. And so people tend to.
Chris:
[53:19] They do get separated, which makes complete sense. If you care less about something, you follow it less. Yeah. And so it's, it's again, that, that, that ugly trade off that comes in. It doesn't mean that those people are idiots or, you know, total like political ignoramuses. They, they oftentimes can do a whole lot with a little, but, but it is, it is certainly less informed than the extremes.
Brent:
[53:42] Yeah.
Keller:
[53:43] Yeah. And looking at some of the reasonings, for polarization,
Keller:
[53:46] kind of like at more of a ground level, not leading us from the elites. Like, could you explain discussion networks and kind of the role that plays in continuing to drive people further to different sides and the difficulties involved with kind of changing those networks? Because oftentimes they are geographically based or they are based in occupation. The ability to enter isn't, at least to me, not abundantly clear. And so the path forward also isn't really clear.
Chris:
[54:14] Yeah, so you may be familiar with Robert Putnam, who's written a few things, the most famous of which was Bowling Alone. So I think that came out in 98, 99, and the whole thing was about the decline of what he called social capital in the United States. And the perfect sort of indicator of that was bowling leagues, that in the 50s and 60s, routine that people would be involved in the, would go down in Davis and Sacramento or whatever and do bowling. It sounds corny now.
Brent:
[54:45] But I guess there is a bowling now. Downstairs.
Chris:
[54:50] I mean, who doesn't love a game of bowling, but like those kinds of things. And then that's why I always love Putnam so much was like that fits in perfectly with conflict extension. If you're joining a bowling league, you have no like reason that that would make you more democratic or Republican. So in 1952 in Milwaukee, you're, you're part of the local group that on Tuesdays meets there and, you know, downs a couple of Schlitzes and bulls or whatever, like you're having a network where you're entering into a network that's going to be almost inevitably heterogeneous. It's going to be mixed. There's going to be Democrats and Republicans. And so, you know, as, and you mentioned this too, like with the, with social media, that certainly a side of this is an important aspect of this as well. It's just easier. You can either view it as easier to get yourself into your own echo chambers, or it's more difficult to find exposure to outside groups.
Chris:
[55:45] Both are at play. Both those things can be true and seem to be true. It's kind of funny. Putnam was a huge deal when he came out. People don't talk about that as much. I mean, maybe, at least in political science, maybe that's because it was so effective that we just sort of got over that. But I mean, to me, that's a crucial part of this entire story. It has to do with the composition of the discussion networks and the fact that you had all of these, you had so many more routes to cross cutting interactions, uh, then, then, then you do now from a media point of view, but also just in your daily lives. Um, uh, that's, that's, it's going to be a lot easier to see the other side as, as sort of sinister and casted in these good versus evil, right versus wrong terms, which is, you know, much simpler and more convenient, but less accurate than what's really the case, which is a bunch of shades of gray that you would hopefully see if you went to your bowling league and, those opportunities become less and less.
Brent:
[56:51] Yeah. Do you think the siloing of a lot of our social interactions is reflective of political polarization? Or do you, because especially nowadays with corporations, ESG gets thrown around a lot where it's like, this is a top-down, we have to enforce these rules. That creates the new culture in that company. Or if you're getting more and more polarized as a populace, I'm going to feel like I need to spend my time not at the bowling alley, but doing this grassroots organization or something like that. Do you kind of see that the polarization of politics happening first and then all of our kind of social interactions kind of follow that same trend?
Chris:
[57:30] Yeah, I think it's a positive feedback loop, which is kind of a cop-out answer, but I really think that that is the case, that one reinforces the other. I mean, it's sort of like moving, right? Where do you want to move? You're going to move somewhere across the country. Well, what's happened with sorting and, again, conflict extension is that, you know, rural versus urban has become a proxy for all kinds of stuff, which in a way it always has, right? Rural is going to be more hunting. You're going to have more of a gun culture. But, you know, it's also just like, you know, tied in with all of these different, you know, religiosity versus non-religiosity, the kinds of social attitudes, the kind of interactions, the kind of, you know, one of the best questions that we have to ask is a battery. It's called cosmopolitanism. And it's a bunch of questions like, have you eaten sushi in the last year? Or how many times have you eaten sushi in the last year? Or how many times have you eaten either, I think it's Indian, Indian, Korean, and something else. It doesn't even matter the specifics because, like, what that's getting at is really, like, you know, Davis versus, I don't know, California very well. But, you know, you go inland or whatever, and they ain't eaten sushi.
Brent:
[58:41] Yeah.
Chris:
[58:42] Where I'm from, the South. uh and take all kind of all kinds of areas where there's nary a sushi or an indian uh restaurant to be seen and so like okay what does that matter well it matters because that that piles up with each other and so if you are in that environment you're of that inclination um that tracks with all these other political indicators well that makes moving to an area without sushi or indian, availability of those foods, but because it's an indicator for a whole bunch of other stuff that people, again, are lining up more consistently. So social sorting, geographic sorting, political sorting, really, really hard, maybe virtually impossible to exactly draw an error from one or the other. But I think that it's important enough, at least, that it's a feedback loop. One feeds into the other. As social sorting happens more and more, that produces geographic sorting that's clear. That produces political sorting that's clear because people can use the fact that, oh, an area is 90% Republican. Well, I can make a whole lot of inferences about what kind of restaurants they're going to have in that area.
Brent:
[59:47] Yeah. And that was honestly the thing my sister talked about where she's moving now. She lives in Chicago. She's like, just crosses out cities because they don't have enough food diversity that she likes to eat. That's funny.
Keller:
[59:58] Is there an argument to be made around like political homogeneity more at the group level, not at like the country level, but social group level that it increases
Keller:
[1:00:08] political discourse and therefore there is a benefit to it? Or how would you think, even though we are more polarized now, has our degree of conversation around politics, is it better or is it more in the limelight than in the past?
Brent:
[1:00:24] Or more active too.
Keller:
[1:00:25] Yeah.
Chris:
[1:00:26] Yeah. It's a great point because, yeah, this is the thing, by default, talk about polarization as it's something bad, right? It's something that we need to fix. um in political science there was a famous um report in the uh the 1950 i think it was 1950, uh that the american political science association our big science organization uh put out uh that that said we have a crisis and the crisis is the parties are too similar to each other and they're not giving voters meaningful choices uh and that's why voters don't know as much as they should and aren't able to make meaningful choices is because the parties have the parties are too depolarized, So it's pretty extraordinary to go back and read that as, you know, exactly the same kind of freaked out tones that we have with polarization now, but totally 180 degrees. So, yeah, I mean, that's one view of polarization is like, well, it may be unpleasant, but that doesn't mean it's bad or certainly not all bad, that that may just be the natural order of things in a big democracy like the United States with a whole bunch of different traditions, racial groups, backgrounds, religious views, you know, local and regional interests and wealth levels. And how much agriculture you have, and is it a more agrarian kind of tradition. It's a big, diverse, complicated country.
Chris:
[1:01:48] It's kind of inevitable to some extent that the politics are going to reflect those sorts of differences, and they should if it's a democracy, if we believe that the majority should rule and that there's wisdom in that or at least some value in that. We need to have meaningful arguments about the difference.
Chris:
[1:02:07] And so, yeah, this idea that we're fighting over values, well, that's like inherent in any, you know, politics is nothing but a constant tradeoff of values. Do you want more equality? Do you want more security? Do you want more liberty? And you can always gain one at the expense of others. There's no answer to that, right? There's no final destination. It's a perpetual, you know, a series of conflicts over coming to what the majority prefers. So, yeah, that is another view. It's not, I mean, it doesn't, it still leaves the negative symptoms of polarization intact, but it also says, well, that's just kind of the cost of doing business.
Brent:
[1:02:52] Yeah. And then kind of stepping back from polarization towards like the middle of the swing voters, could you describe some of the work you've been doing there to identify swing voters, especially when they could be so pivotal in deciding an election? Yeah.
Chris:
[1:03:06] It's tough. You can do it – the problem is that swing voters kind of by their nature defy easy explanation because what they are is a mix of some liberal opinions, some conservative opinions, some demographics that would put them closer to the Democratic Party, some demographics that would put them more in line with the Republican Party. And so, these are the unsorted out there. And so, by definition, it's really tough. I mean, obviously, the media tries to come up with certain indicators that aren't always bad. They can be certainly better than just like a random flip of the coin to suburban mothers or whatever. Um, uh, but like, I, I think of that more as like species or subspecies of swing voters rather than, uh, of course, rather than swing voters themselves. Um, and, and the problem is, is that each of those groups is also like swing voters isn't very large. Now, when you're narrowing it down even further, you're talking about, you know, exotic, uh, you know, there's 32 penguins of this.
Brent:
[1:04:10] Of this state in the world and like, you know.
Chris:
[1:04:15] You know, that they don't go extinct or whatever. So like, yeah, that's the real frustration that I and others that try to dig into swing voters routinely encounter. On the other hand, I think it's kind of nice, right, that there's a reassuring story of that, that to the extent that you're going to get fair-minded voters, that probably is the most realistic path of a bunch of people who are not dead center, but rather are a mix, a mixed bag of left and right opinions. And you just sort of have to hope that in their decisions, right, that there's some wisdom that they're able to weigh up the pros and the cons and select the better option, even if no option is ideal for them.
Chris:
[1:05:07] So I put a lot of faith in swing voters. I mean, if people, political scientists get kind of, I think, way out there in these enormous do or die debates about are they 5% or are they 15%? I don't know. Split the difference as far as I care and say it's 10% to 12%. And by the way, it's not even a binary like you are a swing voter or not. Some people are more inclined to be swing voters in the other. So it's a, it's a granular thing anyway. Um, but whatever, it doesn't matter. The elections now are, you know, that if it is 75,000 votes in Pennsylvania, there's a, certainly seems like it's kind of careening towards, um, there's definitely 70,000 plus swing voters in.
Chris:
[1:05:47] Or whatever. And that's, I mean, that's really so much of the strategy with moderate campaigns and, you know, you sort of see Harris's strategy with this and that, you know, bringing around Lynn Cheney and speaking to Republicans who were disaffected and, you know, kind of the Nikki Haley voters and focusing very heavily on January 6th is that she thinks that's the most promising source of swing voters are alienated Republicans and she may well be right. But, you know, There's a few swing voter groups, right? Some people push Harris more to speak to the white working class and talk more about things like labor unions and economic differences than January 6th and abortion. It seems like she and her campaign have decided that it's a more promising route to go after issues like abortion on January 6th that are leading a lot of Republicans or at least right-leaning independents.
Chris:
[1:06:43] You know, that pushes them somewhat in the direction of Trump, but they can be scared away. If the defining issue is January 6th for them and or abortion, much more likely to switch over to the Democratic Party. So I think that's probably actually a reasonable strategy. It's just impossible to say whether that's the right one or not until after. But you kind of see the differences in swing voters and who they're talking to and what their strategies are and like harris's campaign i don't know exactly what their data is showing but whatever it is it's it's giving them enough confidence that they think there are more votes in those republican circles than are in you know the the white working
Chris:
[1:07:22] class that used to be democrat 40 years ago that were part of labor.
Brent:
[1:07:25] Unions and all that wouldn't that be where walls comes in.
Chris:
[1:07:28] Yeah yeah so yes uh that that is going to be more appealing to that group um and like Yeah, him donning the hunting vest. And so definitely Democrats still do that. It usually turns out terribly.
Brent:
[1:07:44] Like Kerry had that orange vest.
Chris:
[1:07:47] It just didn't look right. It didn't look normal. Yeah, and so, yeah, you don't have to – you can do multiple things. Where attention and time is most scarce, though, it seems like more emphasis. And we'll see. I mean, they're going to try to do it all. They're going to bring out Eminem and Obama to Detroit.
Brent:
[1:08:07] So like, just like Trump's bringing out Elon.
Keller:
[1:08:09] Exactly.
Chris:
[1:08:10] Yeah. Um, so you don't have to select one audience, but to me, there's a lot of evidence that, you know, in this, in this most valuable of real estate last two weeks of an election, They just feel like they can get more traction, but not 100% that that's the case. And I'm certainly not 100% that that's a good strategy or bad strategy. Yeah, we'll see. But that's always the debate, especially the Democrats have had, is do you try to – those people who were voting Democratic in the 1970s, 1980s, are they gone forever or can you get them back? Bill Clinton kind of got them back. Not kind of. He did get them back. But Obama was only partially successful in getting them back, and he was by far the most successful. You know, Gore wasn't able to get them back. Kerry wasn't able to get them back. Hillary Clinton certainly wasn't able to get them back. Joe Biden did a little bit, but, you know, at the margins.
Keller:
[1:09:02] Yeah, so we shall see. And stepping out a little bit more broadly in terms of, like, modeling politics in a general sense, how close are we with the data we have and the computation power we have to creating accurate models i guess particularly on elections that can consistently give us accurate estimations of where we're going to go and if we were to do that how do you envision that changing at the citizen of our dynamic and our relationship with politics.
Chris:
[1:09:28] Yeah i you know i mean so you know i study public opinion i mean that's it's uh that's my livelihood and you know just my experiences with surveys and also just in life. I mean, human behavior, there is just a random element to us. Whether it's truly random, gets into philosophical free will, rabbit holes that we shan't go down, but for all intents and purposes, it's random, right? We just do quirky stuff. So, you know, I just fundamentally believe that no matter how much information we're able to collect on any given individual, that there's still going to be the spark of randomness or just something that's beyond our perspective.
Chris:
[1:10:09] Don't even need to characterize what it is. It's enough to say that it defies any sort of, aggregate pattern where we can use information about how other people with exactly the same profile would act and you, for whatever reason, act something differently. So, you know, I tend to not put a lot of stock in kind of that. I mean, that's always a concern, right, of course, with AI and machine learning and just big data and data science more generally is that are we going to lose sort of the art? I don't think we can, even if we were able to collect a lot of information. And of course, the other fundamental thing is just that, at least with survey research, that data is getting more and more garbage as it goes on. So certainly other stores of information, big tech, all of that's out there. But I can believe that that's eerie, but I am very confident it's not perfect. And with elections, right, you have these multiple things you need to predict. You need to predict is someone going to vote or not, and then, of course, who they're going to vote for. And so there's multiple failure points, and that always is – when you think about error and prediction errors, there's multiple ways to be wrong.
Chris:
[1:11:27] Obviously, we've gotten better, with developing and using more sophisticated tools to come up with more predictive power, but I don't think we're ever going to be perfect. And again, I think it just is all contingent on how close the race is. I just don't think we're ever going to get to a system that can get it down to 50,000.
Brent:
[1:11:48] 75,000 voters in Pennsylvania.
Chris:
[1:11:50] Um, now if there's a 20 point margin or whatever, then yeah,
Chris:
[1:11:55] we'll be able to predict that at.
Brent:
[1:11:56] Least the direction perfectly.
Chris:
[1:11:57] But, um, I, I just doubt where it will ever have that. Was dangerous statements to make.
Brent:
[1:12:06] Yeah. That blank will never happen. Do you, have there been anything like we're two weeks out from the election? Like, has there been any like data points that you've seen that have been like really interesting, like really surprising so far in this last cycle?
Chris:
[1:12:18] In this one?
Brent:
[1:12:19] Yeah. Um, like for reference, one thing I've heard was like the young black population is like now switching towards Trump at a way they're estimating like 48% where they've typically been 80-90% Democratic voters before? 40% now switching over to Trump and the Republicans? Has there been any key, whoa, this is very different findings that you've seen?
Chris:
[1:12:44] Yeah. If it gets 40%, that would be... You're definitely right. It's in that rate. Some polls have put it that high. I would be shocked if it manifests itself as that. But your point's well taken that is surprising, although kind of think about, some trends that are present as far as like who constitutes the Democratic Party and who constitutes the Republican Party and just the kind of cultural styles of both as well as the kind of gender divisions that are more and more evident in the other parties. So I think that's surprising, but it's not totally shocking. I mean, what's been turned on its head, and this isn't new, but it is really shocking to me how quickly it's occurred and how drastic it's been as this complete inversion of regular voters, it always used to be that if you had a poll that were likely voters that did a screen for, is this person actually likely to show up? That was always going to favor Republicans. If you did registered voters, your population was everyone who could vote. That was going to favor Democrats. That was like an iron law forever that the likely voter polls were going to be two or three points in the direction of Republicans.
Chris:
[1:13:57] Not only has that narrowed it's flipped like in polarity that uh with with trump specifically it's the uh it's the irregular voters uh that that uh that that that break for him quite substantially and the the regular voters uh the the high propensity voters that are breaking for democrats i mean that that is just that is a total polarity flip.
Brent:
[1:14:22] Is that 2016 it flipped or like 2024 2020.
Chris:
[1:14:25] Yeah 2016 uh it it It flipped, but it's continued to expand 2020, presumably 2024, at least the surveys that I've seen so far suggest it's going to widen a little bit more. And again, it's that positive feedback loop of these parties get tied in. I mean, it's exactly concurrent with it used to be that college educated voters were going to be more Republican because they made more money. And it was the high school voters and less that were going to be the core of the Democrat Party. It was moving in that direction before 2016, but 2016 actually flipped it, and then that – in that positive feedback loop has continued to expand because the Republican Party has more closely tied itself with this rural, more traditional platform, this very populist platform that appeals to especially white working class, but more generally just working class voters with lower levels of educational attainment. And then Democrats are including the party of the sophisticated coastal urban elite. Davis, California, right? You have people that have petitions against light pollution and.
Brent:
[1:15:38] You know, anywhere else.
Chris:
[1:15:40] In the country, they'll punch you in the face or something. Which is, Davis is charming and quirky and so on. But this is like, talk about not representative.
Brent:
[1:15:49] Yeah.
Chris:
[1:15:49] And so, you know, it's only a matter of time before that like manifests itself. If you have the party of the little guy that is, you know, having bills, you know, is, you know, doing plaques for trees that were knocked down during a storm or whatever, like, that's not going to resonate. And so that tension has been there and been present for a while. Trump really just kind of seems like.
Brent:
[1:16:15] He was the.
Chris:
[1:16:15] Very pivotal figure that broke that open. And so I just – I guess I'm still – so it's not really new, but I'm still kind of blown away by that because we always treated those things as just like iron laws.
Brent:
[1:16:26] Yeah. Yeah.
Chris:
[1:16:28] Likely voters, the consistent voters are Republicans. Democrats are all about can they get off the base. And that's still present to some degree, but it's – I mean just empirically, it's been turned on its head. So seeing the continuation of that is surprising. But I mean, your example is a really good one, even if it's, you know, black males are voting at 20 or 25 percent for Republicans. That's a decided shift.
Keller:
[1:16:57] And as we wrap up here, do you have any advice to students, particularly those interested in the intersection of statistics and politics and kind of the way to navigate this field?
Chris:
[1:17:05] Yeah, I mean, maybe it's trite, but I mean, don't fear it because I mean, as someone who's like, I don't come from any kind of amazing math background or ability level, I did it because I saw that this was cool and answering questions that I cared about. And, you know, you figure out a way. So like you start approaching it, if that's your motivation, that's great. If you happen to love math, then that's absolutely fantastic. Whatever gets you more comfortable working with statistics. The other thing is that statistics, I mean, there's a huge, like, right brain part of that as well. Like, there's a whole creativity in interpreting data, interpreting what the story that the data are telling you, which I think is a huge value of, like, a social science degree, including a political science degree, is that you're not just a, you know, churning the cranks on data. You're having to interpret it. And there's artistry and creativity that are involved in that.
Chris:
[1:17:57] You can use that for good or ill, but the fact is it's there and we have to do it. We have to give some interpretability to these numbers and what it's trying to tell us. So with statistics, obviously there's math, but I actually don't think that's the hardest part of statistics. I think it is more of that kind of the abstract and the creative interpretation, that you're giving to your results where you have all of this stuff going on, all these, everything's affecting everything, each other else, errors running every which way. And you have to, and you have data, which is great, but it's just, you know, it's just a fire hose of numbers. You have to sort of make sense of humans, which are very complex, and we deal with each other in complex ways.
Chris:
[1:18:45] Trying to actually get leverage from the data is more rewarding, but it's also more challenging. And so... I don't know if that's encouraging or maybe depressing.
Keller:
[1:18:56] But like the, the.
Chris:
[1:18:58] The, the math is hard with statistics, but it's probably not even the hardest part. So, you know, a lot of people let that be an impediment. I let it be an impediment for a while too. Um, you, you can work through the math though. I guarantee it. Um, you may not be doing, you know, writing proofs and lemmas and, and all that other stuff. I don't, I don't really do any of that either. Uh, you don't have to. Um, uh, but, uh, so yeah, statistics is just, it's too, it's too valuable commodity. It's the most valuable thing that you're going to get quite just in a practical manner of speaking. The most valuable thing you're going to get from a political science degree is your ability to think creatively and interpreting data and understanding uncertainty and unknowns and probabilistic kind of thinking. And that is something very valuable that you will get out of a political science degree. But it does require you to do not an overwhelming amount, but a little bit of math. Um, and you know, it's, it's, uh, it's doable and it's worth it. I think.
Brent:
[1:19:53] Perfect. Awesome. Thank you.
Chris:
[1:19:56] Right.
Keller:
[1:19:56] Yeah.
Chris:
[1:19:56] Yeah. How long have you been doing this?