Do We Understand Members Of The Other Party?

Alright, here we go. Anthony's got it. Oh my god, I'm Anthony Fowler. I'm the old dude, that. And I'm Ethan Brenner de Mesquita and this is not another politics podcast. So we talk a lot on this podcast and in America about partisan animus and the idea that we have just divided into our teams and not only do we disagree with each other on politics but that spills in over into all sorts of domains of life we no longer live in the same places. We don't have the same kinds of jobs. We don't have the same kinds of education. And in a bunch of research there's I guess evidence that we report to hate one another. But I think one sort of deep underlying question if you think that people profoundly dislike it one another is is that because they disagree or is it because they just they just can't understand one another. They don't really know what the other person is like and they have in their head a caricature of one another. And so do Republicans have an accurate view of Democrats and hate them anyway or do they think Democrats are you know some three headed beast that they hate because of that? Do you know what? What do you think? What do I think? I did hear this kind of conversations in in with the science where people do report surveys or people who are really very expressing very negative views about the other side. And the interesting part is that this animal seems to be growing. And I did hear about this one explanation that people do do hate each other because they do misperceived themselves. And one explanation for that is that maybe we watch more partisan media maybe we are now in our echo chambers and we really don't have an opportunity to learn about the other person. We just don't know what the other person is thinking. As we mentioned we live in different neighborhoods. We work in different jobs. And there's a lot of research documenting this increasingly in the workplace. So it is possible that we have nice and nice information on the other side. And as a result we form opinions that are not reflective of who the other side is. But Anthony you spoke to someone who said I have an idea of how to actually figure it out whether we do have accurate opinions about the other person. As you can see it's not easy to figure it out because if we just ask people they can say whatever they want to say. I did I spoke with Adam Astriani who's a really interesting person. He's an independent researcher with a background in psychology. He's a very gifted writer. He writes a lot about philosophy of science in general. He writes a lot about problems in psychology. But we interviewed him because he had written this really interesting research paper on exactly this topic. And I think he came up with a clever way to study this question that nobody had thought to do before. So rather than just asking people what their perceptions were and asking people what their feelings were, he asked them to try to mimic somebody from the other party. And then he asked other people to see could they tell when someone was being honest or not. And the whole idea was people really misperceived the other party. They won't be able to mimic members of the other party. Maybe we should start kind of at the beginning. There is a big academic debate about the extent to which Democrats and Republicans understand each other or misunderstand each other. Tell us a little about that literature and why you thought that was an interesting topic to explore. So there's all this interest in like well why do Democrats and Republicans hate each other and how might we help them hate each other less? One branch of that idea has been well maybe they don't even know who the other side is. Wouldn't it be really convenient if in fact these two sides only hate each other because they imagine some caricature of the other side in their minds and that's really what they hate. And so there are all these papers that came out showing that oh look each side really doesn't know who the other side is or what they like or what they want. Like obviously directionally they know that Democrats are more likely to support abortion and Republicans are more likely to support gun rights. But the extent to which these things are true they overexaggerate and they even overexaggerate how much the other side hates them. So if you ask a Democrat like how do you feel about Republicans are like not great don't like them but then they're like how does Republicans feel about you they're like they hate my guts they want me dead when you ask the Republicans they're like I don't really I don't like them but they don't say I hate their guts I want them dead. We felt like that's all very interesting but the way that these questions have been asked is all pretty similar you just like ask people to bubble things in on a form essentially. And there's some issues with that like it's easy to respond expressively I think the other side is really evil and bad when really what I mean is I don't like them there's no real skin in the game here people aren't incentivized to be accurate most of the time and the task isn't very involved right you're just like pressing a button. So what if you had some skin in the game and the task was more complicated than just you know clicking a button on a likeered scale on a screen. And we do talk on this podcast a bit about what we call partisan cheerleading which is exactly the kind of thing you're just talking about that we get these survey reports we don't know people are paying close attention to the surveys or giving their genuine preferences maybe they're just giving the kind of the partisan answer or the fun answer. So let's get some skin in the game let's talk a little bit about how you do that. What is what is the touring test? So the original touring test was this idea that Alan Turing came up with in the 1950s and his setup was you know if you're a human and in one room there is a human typing answers to your question your two you can talk to them over basically a chat box and in one room there's a human answer your questions in another room there's a machine answer your questions you can't see them if you can't tell the difference between the human's answers and the computer's answers for all intents and purposes like well don't we have a thinking machine or isn't that machine human or or whatever we thought this could be a useful measure of to what extent do two sides understand each other so if you were talking to a real democrat and a republican pretending to be one or a real republican or a democrat pretending to be one and you couldn't tell the difference then it would suggest that the person who went undercover the person who's faking knows something important about what it means to be the other side at least enough to convince a copartisan we didn't have people talk we didn't think people were good enough actors to pull us off but what if you just had them type out a response we had a bunch of people who were democrats or republicans write a statement 100 words at least 100 words many people wrote more than that based on the prompt I'm a democrat because or I'm a republican because and they're randomly assigned to each of these prompts so half of people are telling the truth and half of people are faking so half of half of people are saying what they really believe I'm a democrat saying why I'm a democrat or a republican saying why I'm a republican and the other half of people are democrats explaining why they're republicans and republicans explaining why they're democrats so we have this big set of true and false statements are there any interesting things that you notice about what people do when they're telling the truth and when they're mis you know representing the other side and lying yeah the I think the most interesting thing is people took this task seriously so this would have been a really easy opportunity for partisan cheerleading to your point we did incentivize people we told them we're going to take your statements and then show them to someone else we're going to tell them the whole set of the experiment they're going to try to tell whether you're telling the truth or faking we'll pay you a bonus if people think that your statement is true and so people have some skin in the game but still a democrat pretending to be a republican could have been like I think every toddler should have an uzi and republicans pretending to be democrats could have been like I think everybody should be gay and and like they they could have done that and like a handful of people did that but out of the more than 800 responses we had they they were really good and convincing and serious okay great so you so you show these to people you have them predict which ones are real which ones are fake and everyone is incentivized the writers are incentivized yes and the readers are incentivized whatever interesting misperceptions there are out there about the other side this would be a good opportunity to pick them up well you what do you find what are the headline results here the writers do really well and the readers the people trying to pick out the true statements from the false ones are at chance and what's especially interesting is that people don't do better when they're reading statements that claim to be from their own party so it's one thing to say okay on average I couldn't tell the difference between fake and real but when it comes to my side I should definitely be better at picking out the fakers on my side then on the other side because I should know a lot more about my side and that's not what we see which suggests that like actually there's a lot of shared knowledge across both sides that it may well be true that there's plenty of misperceptions I still believe that but clearly these sides understand each other well enough to go undercover at least for a little bit and put on a convincing enough performance that someone reading it can't tell the difference between the real thing and a fake is it possible that it is just inattentive reading that the readers didn't care enough or pay close enough attention and that's why they're doing badly they're doing badly because not that the writers were good but that the readers were valid in a sense yeah you might think so for one thing we also pay the the readers based on their performance so they also have some skin in the game if they just click the right button they're going to get paid more but also you know we have a version on the internet that people can try for themselves and when you do it it immediately becomes clear that this is pretty hard yeah I have to admit that I myself I was skeptical when I read the findings of the study and I did I did follow through the links and I did do this exercise I read a I read a big sample of them and I thought for sure I'm a professional political scientist I started I think for living and not only am I a professional political scientist I've actually worked on research projects where we ask people to explain why they are the party they are and we code the responses so I have a lot of experience with these kinds of responses with the real ones and I thought for sure I would be able to pick out the real ones and I was pretty much also no better than chance so that made me much more convinced that this was something real and and if you if you're still skeptical of the result maybe you're just skeptical of my reading ability as well but I was I was really struck by how good the fake responses were or good or bad I was just struck by how similar they were to the real ones I suppose you mentioned a few heterogeneity tests like people were no better within their own party, you have a few other ones that are interesting as well. You might think I certainly would have thought that people who are more extreme partisans or extreme ideologues, they would be worse at this because they're worse at empathizing with the other side. What do you find there? Yeah, we similarly thought like if you're a real nut job for your party, probably you're going to do worse at writing because you probably have a more warped perception of what the other side is, or maybe you're just a more warped person in general, maybe you're worse at expressing yourself, didn't find anything there. We also didn't find any difference in terms of, okay, if you're a real nut job for either party, then when you read statements, you're probably going to do worse because you're probably in a more biased media environment like something made you more extreme in the first place, maybe that's going to make it hard. We can't find any differences there either. We were also surprised to find when people try to predict their performance on the writing side, they don't consistently over or underestimate, they just misestimate. You might think that everyone's going to go in thinking like, "I'm going to nail it," and then they do poorly. It actually turns out people have no idea how well they're going to do, so there's some lack of self-knowledge there. Then we also looked at like, "Okay, what about the confidence of the people reading the statements? Are they sure about these? Or is it really that like I'm just guessing?" What it turns out is people feel very confident when they think a statement is true, and they feel much less confident when they think a statement is false. This is consistent with a model that's been found in other lie detection literature that people basically default to thinking that a statement is true, and they only flip to false when they reach a certain threshold of evidence. It would make sense that if this statement looks pretty good, I feel confident, and I feel confident that it's true. Then only when there's something off about it, do I start feeling less confident, and then I flip to false. There's a bunch of additional results that we could throw into one being, "Okay, what kind of mistakes do people make?" You could be at 50% accuracy if you say true all the time. You could get there if you say fake all the time. You could get there if you respond randomly. What kind of mistakes are people making, and the kind of mistakes that they're making is that they say true too often. They're too credulous, which is also consistent with the more general lie detection literature that that is the mistake that people make more often is calling a fake thing true than calling a true thing fake. Just to be clear, people were told that half of them were fake. It's not like they were trying to be Bayesian's, but had the wrong prior or something like that. It's something else. We told them that there would be a mix. We didn't tell them, because we didn't want them to be like, "Oh, I've seen all of my real ones so far, so now we need to save fake." Is this the kind of thing that you might think that AI and computers would be able to pick up on? Even if people are bad at reading these things, maybe there really are subtle differences between the true and false statements that some more sophisticated procedure could pick up on? We checked this because we wanted to know if a computer can't even tell the difference between a real statement and a fake one. Maybe there isn't any signal there at all. Maybe this task was too easy, and it did actually turn out to be pretty hard to get the computers to see the differences. My co-author on this project, Chris Nichols, tried a bunch of different things. It didn't work. Layton, Deer Clay analysis, this bag of words, technique, random forest models. In the end, it was only at the time a pretty advanced technique of a lasso regression that was able to tell the difference. He later came back, and now there's a thing called "Burt" that does even better, but the headline being basically, "Yes, it is possible for machines to tell the difference." However, not all machines. We also fed these to chat GPT-4. Here are some statements. Here's the whole setup. You're going to tell the difference between true and fake. It was really convinced by all statements. It basically just thought everything was true. I think of a piece with the psychofency that people have been picking up in large language models recently. Can you look at those lasso or "Burt" results and figure out what are the things that the computer is picking up that would allow you to understand what subtle differences there are? A little bit. I'm more hesitant about these results because there's always this worry about overfitting or like, did it just do a really good job of picking up a few of them? But it does look like real Republicans, for instance, talk more about tradition than Democrats do when they're pretending to be Republicans. Real Democrats talk more about inequality than fake Democrats do. Republicans pretend to be Democrats. But this is something that I think we're still not sure about. But those are two possible differences. I want to give you a chance to speak more broadly about the state of academic research because I know you have a lot of interesting thoughts on this topic. Maybe we could start by talking about the context of this paper, which is this is a series research paper, but you've posted it on your substack. It doesn't look exactly like a normal research paper. It doesn't have a long appendix with lots of extra details, but there's data and there's code online and you're putting it out there and you're talking about the details. I assume you're not trying to publish this in a journal and get it peer-reviewed. So how do you think about what you're doing when you do the research and why are you going to kind of taking a different road than other researchers are? Yeah. I think of this as diversifying the ecosystem of science. I think we're in this weird moment where the university system, which has its pluses and minuses, has a monopoly on the production of science and scientists and the problem with a monopoly is that it's homogenous, that it can't optimize for everything. And so there's things that you can't do well within it. It's hard to explore weird ideas that aren't going to be amenable to being published or funded. And so I think what we need is a wider ecosystem of different kinds of institutions and people so that those little niches can optimize for the things that this big chunk in the middle can't do. So I was trained in a psychology department as an experimental social psychologist. I did that for a while, but then I had this feeling of like, whenever I write a paper, I feel like I immediately get stupider. And I didn't feel like I had a bunch of intelligence that was like an excess of smarts to begin with. And eventually I was like, why am I doing this? When I write directly for the public, I feel like I can use everything that I have. And I could still hold myself to the same standard. As you said, all the code data materials are there for the people who like want to build on this or dig into it, which is to be fair, something that you can't really say about most papers in the field. Don't even do that. And so now I'm doing the science exactly the way that I think it should be done from my tastes and abilities and goals. And I think there should be more of this. I don't think everybody should be a crazy person in the woods, but I certainly think that we need more of them. And we're in a weird historical moment where we have far fewer than we used to have that used to be only weird people in the woods really. And now, and we've centralized and professionalized it. And that came with a lot of pluses, but it also came with a lot of minuses. And I think we need to diversify to recapture some of the pluses that we've lost and some of the minuses that we've gained. And I know you're mostly a psychologist in terms of the topics that you study. You know, you're covering clinical psychology here with this paper. And I know you have a lot of thoughts about the field of psychology as a whole and what is doing well, what is not doing well, and what it should be doing. I wonder if you have maybe our listeners would be interested in some of those thoughts. And if you have any, do you think some of those lessons are transferable to political science? I think we'd be particularly interested to hear that. Yeah, I think the real thing that's holding psychology back right now is we are a pre-paradigmatic science, which is just a way of saying that we don't have a consistent model of the world. We don't have a way of organizing the things that we know and being able to tell like, or missing something here. And we don't have a way of assimilating new knowledge as it comes in. Rather than like building an edifice of knowledge, we have like a bunch of stones that nothing gets put on them. And so we have an ever-expanding field of, you know, brick walls that are one brick high. And I think creating that paradigm is going to look really weird. And that's why it's not going to happen inside academia because it requires throwing out a bunch of stuff that maybe we think to be true right now or putting things together that we don't think go together right now. Just like every paradigm shift we've seen in the past, it usually comes from an outsider. It's usually resisted by people who like the way that things are going now. And it usually doesn't make sense with a lot of the things that we already know. It explains part of things really well but leaves other things unexplained. I think psychology has to undergo that transformation if we're going to become a productive science. I think there are some like intriguing proposals for what that paradigm could be. Some of my friends have just put out a book called The Mind in the Wheel that's about a cybernetic take on psychology that I think it could be that but mainly we need more of that kind of thinking if we're ever going to go anywhere. I don't know as much about political science but my impression is that it's in a similar state where like what is it? Like what constitutes the core of it? There isn't a periodic table of it, I imagine. And not that a paradigmatic version of it wouldn't necessarily have a table but it would have a typology and ontology of these are the things that our world is made out of and these are the rules by which they operate. Those are the two things that make up a paradigm. And if we can't answer that question I think we're really going to be grasping in the dark and it's going to be possible to discover some things that are true and useful just like alchemists did but it took a belief in elements and atoms and the things that they can and can't do in order to really crack this thing open and start making it productive. My sense is that a lot of your criticisms would apply more broadly to the social sciences and to political science but there is probably more theory in political science than there is in psychology. And you'll find lots of wide disagreements among political scientists but some political scientists they work on, they have kind of a workhorse model of electoral accountability and they say these are the key ingredients of how electoral accountability works. And then let's go out and test to see if those if those key forces are playing the role they're supposed to be playing and let's test the You know, test the predictions of that month. model and you know there's a bunch of stuff built around that that turns out to be really useful so that if you show me some new electoral system in Sweden that I've never seen before I can think about it in a productive way and I can tell you what institutions might work better or worse and so far. So there's some of that kind of thing going on which I think I wouldn't use the word paradigm but I think is moving in that productive direction but there's also a lot of just one off like look at this fun wacky thing that we found without any attempt to like piece it together or work toward a bigger goal. So I don't know I think I think we're in a maybe we're in a slightly better state than psychology but not by much and I don't know. Yeah I mean we certainly I think face the same challenges of when you are studying people it becomes really easy to deal in abstractions that feel real but aren't real like in psychology an abstraction is like depression you know we talk about it as if like it's a thing whereas like it's really the way you describe how people fill out a depression questionnaire like it doesn't refer to anything in particular even though we all know what we mean when we say it but I think this is why like we haven't gotten better at treating depression in the past 50 years we don't have a mechanistic understanding of what is it to really get better at fixing the system we have to be able to say like what are the pieces that make up this system and which one is broken such that you have this thing that looks like depression and when you can do that you might realize oh what we call depression now is actually 17 different things that correspond to different malfunctions within the system like over here that like this when this part breaks you feel depressed in this way which is actually totally different from what happens when this part breaks but right now I think we're kind of at a state where you know you take your car to the shop and they're like your car has broken downness and you're like well what does that mean it's like well it doesn't work it's like well there's many many different things that could cause a car to break and each of them is fixed in a different way and if you don't have a theory of like what is the engine made out of how do the parts work together then you're never gonna be able to fix the car so I imagine in political science there there are two things that sound like they're real but actually turn out to be abstractions that if we're gonna make progress we have to dig down into like what really is like what is the nature of this system what's it made out of and where on the blueprint does that abstraction live yeah no it's great that's great it's really fascinating okay yeah we'd love to maybe we'll have to have you back maybe maybe we'll make you read more political science and then get you to put it we're doing well and what we're doing badly if you're getting a lot out of the important research shirt on this show there's another University of Chicago podcast network show you should check out it's called entitled International Lawyers, Claudia Flores and Tom Ginsburg have traveled the world getting into the weeds of global human rights debates on entitled they use that expertise to explore the stories and thorny questions around why rights matter and what's the matter of rights subscribe to entitled part of the award-winning University of Chicago podcast network so when I was reading the paper at the beginning I was convinced by this idea not indeed if I don't know anything about you I can't mimic being cute so like if you think about being a spy or you know pretending that you're you're part of the other group that makes a lot of sense but I'm trying to understand whether it makes sense in this context so people were asked to write a little paragraph of why are you a democrat or why are you a republican and they were taught try to pretend to really be a republican or really be a democrat and if I do well on this job is this a proof that I really know you or is this a proof that I know what you would write if you were asked to write what it means to be a republican in a way that's not distinguishable from the outset whether it's fake or not yes I mean it means you you know you have a good sense of what somebody might plausibly write maybe you don't really understand what's going on in their hearts or in their heads but I think even that is is still pretty useful to be able to understand what somebody would write in a question like that I think like one interesting thing along these lines so you're you have in mind that people might even if I'm an extremist and a real republican I also might moderate in what I'd say because I think to myself if they see a really extreme thing they're going to think it's maybe a caricature and so I think an interesting kind of extension that you might think of to try and get at this is to see how different the paragraphs would look like if you said to people I want you to your paragraph is going to be read by republicans I want you to write a paragraph to try and convince them that you're a republican or you're a paragraph to tell me exactly what you think about yourself so I've done exactly that in a research project where I've just adopted a simple thing of just and there was no incentives or anything like that we just asked people open-ended you just said you're a democrat or you just said you're a republican why why do you think of yourself as a democrat and typically people didn't write paragraphs we let them write whatever late they wanted and very often it was just one sentence or two but by and large they don't come across like extremists they they talk a lot about policy they talk a lot they they sometimes sometimes mention political candidates and recent performance and so forth they'll say like oh I I used to you know used to be a republican but I really don't like Trump that kind of thing but they don't come across as particularly extreme they'll they'll often say oh I've got this like mix I've got conflicted views I've got a mix of views but I kind of I kind of shake out this way and they're they're writing for researchers maybe you know maybe they're anticipating what they think that we want to hear or something like that either of course you're always thinking about your audience when you write something but you might think you know we hear all this reporting of partisan cheerleading and surveys and so forth and I think there is good evidence of that in a lot of cases but in these cases and in atom studies and in in the separate data I've looked at people don't write overly extreme or partisan things when you give them a chance to expound upon them so I'm not surprised like I actually truly people are not you know completely crazy and completely extremists like I we have a lot of conversations about this on this podcast what I'm just saying is put your research aside I think the experiment even though you know it it really sounded cool at the outset like I don't think it was really designed to figure out whether I truly know the other side because from that experiment I don't know like I don't know what the other people who the other people are everyone was instructed to play a beauty contest that beauty contest again am I saying it right I think you know this book is as a secondary role of correcting my English and and so I don't know everyone wrote what I guess I would write if I were told like you know what's the topic what do you think a typical person would write when it's structured in the typical way that's true now I guess maybe we should maybe we should go back to the details so the people when the people are writing the truth they're just told just told explain why you yourself are a democrat or a republican right and then when people are writing you have to think about that that's what they're asked to do and then you have to think about what the reward structure is for them and they get they get rewarded they get rewarded if their false paragraphs are perceived as true and true perceived as true everyone's awarded if whatever they write with this true false as far as I understood that's what that was my understanding too and so that that could create these incentives to kind of always always appear kind of a reasonable moderate if you are worried that people will interpret anything that looks extreme as more likely to have come from the caricatures and the real thing so even so even if it's true that people are misrepresenting their views when they're when it's when they're writing about their true partisanship in a way to make it look more realistic we're still kind of asking in this research project Adam and his co-authors are asking when they're when they're a member of the other party do they write paragraphs that look like members of that party would write when they were trying to mimic members of their own party so we're still learning something like do people have a good sense of what members of the other party would write and do they and do they write similar kinds of things so which which does get at this bigger question yeah I mean it's really misperceived Joe I think it does show that Democrats can at the level of a paragraph engage in the level of empathy required to answer the question why might a human being oppose gun control oppose abortion rights and want lower taxes and which is kind of the three like if you like those are kind of the three topics that come up and I don't know I read about 20 of them that those are kind of the three big topics I saw like people can empathize at the level of saying like why might a reasonable person well those use what kind of one sentence reasons would they give for that in a convincing way which I think does suggest that what's going on to the extent that a part of it's an analyst is real what's going on is not that I cannot imagine the kind of reasons you would give for your positions but rather that I think that your positions are somehow reflective of I mean there could be a couple things going on one is that they were reflective of a worldview that I find loads I'm saying yes you are in favor of more research distribution because you don't think people should be responsible for their own actions and you don't care about economic growth and you don't believe in freedom and then I'm going to label that communism and then I'm going to say I hate you but it's not that that that first sentence was right like in the right there's some progress there I think in terms of if we ask like of the possible reasons why we might have partisan animus is it just like we just naturally in the same way that I don't actually like let's I hope my father's not listening but like you know I don't actually hate the red socks I grew up a Yankees fan I don't actually hate the Packers but like when I'm at a bearers game and Aaron Rogers is saying I own you I feel like I hate them briefly if you ask me do you hate the Packers I was like yeah being a Packer is the worst kind of thing to be right now that's not that's not real that's just like in the same subway car, and the white socks, and the cups, and standing in front of each other, and they are peaceful, and so they don't really hate each other, but if you go to Europe, and it's like Barcelona playing with Real Madrid, those people truly hate each other. What sport are we talking about? I don't even know. Yeah, I hate people who play soccer here. I think there's a third possibility, and I think you hinted at this, but you sort of moved away from that. Imagine a Democrat nowadays, who really thinks that Trump's immigration enforcement is guided, and really feels deeply about huge injustice being done, because people are sent, people who have been in this country for 20 years, let's say. And haven't committed any crime, they are sent to El Salvador to rot in prison, even though they are from Venezuela. So suppose you're a person like this. You might dislike those who vote for Trump, because you might feel like if you take a step back and think about it, you should be against this policy. This is inhumane. But if you are asked, how do those people think about themselves? Obviously you're not thinking, oh yeah, those people walk around and say, oh, I'm so happy that all those people are sent to El Salvador to rot, because I hate them, and they should all be rotting in prison. They think about themselves differently, they think about themselves, oh, I really am against open borders. And the reason why this Democrat I'm describing might dislike Republicans is not because she truly believes that they are thinking of hateful thoughts, but because she plans them for not thinking more deeply, that behind the slogans, you know, they are human beings who are suffering. So I think it's possible that if I describe the other person, I will use reasonable language, because I think they would use reasonable language. But I still blame them for holding these beatings, because I think they are not thinking about themselves. I don't think there's anything inconsistent with that and the paper. I think what he's trying to say is there are lots of paths from where we are to hating one another. And we're going to try and rule one of them out. And the one we're going to try and rule out is we have, or you're totally unable to empathize with a human being holding the positions that the other side holds. And we're going to rule. I think there are claims in the literature along those lines. Yes, I think if you ask most scholars of political behavior, they would say exactly that. They would say, partisans really don't understand each other. They can't empathize with each other. Democrats think that Republicans just hate all foreigners. If I asked you to write a paragraph explaining why they think that. If they read it, they would think. I think I can do it. I think I can. I think I know them reasonably well. But I think I think that is a genuine view. And that is the view that they're trying to rule out. And I think if there were a long literature view for this paper, it would probably engage with that more deeply. But I think there are lots of claims along those lines that now I think they should have to grapple with this new evidence, which is. In some simple survey measures of like, do you think all Republicans are hateful bigots who don't like immigrants? Of course, people, lots of Democrats say yes. But then when you follow up and say, what would a Republican actually write? But I don't think they would rule it out. So again, I told the story that consists of evidence. But there's not a story in which I truly believe those Republicans are awful. And they truly want those people to write in prison. But I know they were asked to write a statement that would be judged by someone whether the statement is true or correct, you know, true or real fake. And I know that no one in the right mind is going to write. I want people to write in prison. Yeah, so that's interesting. Imagine, so I think, right, you have in mind like there's imagine two different worlds. One, where Republicans support Trump's immigration policy, because while they lament the inhumane behavior, they recognize it as in some way a necessary condition for what they think is a good policy goal that there's no way to the policy goal of control over the border that doesn't run through some in humanity. And they wish that wasn't true. And they feel bad about that thing. But they also say life is full of trade offs and on balance, I'm going to make this trade off. And another where they genuinely hate people from certain cultures and they want the inhumanity, those two different worlds. And your theory as I take it is those two worlds, now I'm a Democrat and my theory about Republicans, the reason I hate them is I'm pretty sure we're in the second world. But when asked to write an essay neither in both of those worlds, they don't say, well, what's going on is I hate people from Latin America. They don't like to expel what they say, it's very important to have control over the border and there has to be order and there's currently there was chaos under the Biden administration. And of course, there's lamentable things that happen with any policy mistakes are made and some lines are gray and whatever. And so, and then the Democrat asked to mimic them rights that sentence and now we haven't adjudicated. And now everyone says, oh, I can't tell the difference between Republicans and Democrats, but we have not ruled out whether the Democrat is right or wrong in their theory that deep down the Republicans hate the foreigners. That's the kind of idea, but exactly that's the, that's the gist of it. So we both can't tell which of you that Republicans have from seeing the essays, nor can we tell whether the Democrat, what view the Democrats actually have of the Republicans because they're trying to mimic and we can't tell them whether the Democrats are right or wrong. I think what's funny is like your theory, your theory is kind of the opposite of the standard. Like your theory, I just say, I just don't know. Okay, but your hypothesis here is like people are extreme in their hearts and moderate in their speech. Like I think actually like, and everybody knows that you don't know whether they're extreme in their heart or moderate in their hearts, but you know they're going to be moderate in their speech. And then you try, and so then you, yeah, and you suspect they're extremely right, you can't find out, whereas like lots of the standard discourse is like actually like, you have people say all sorts of crazy stuff, but like if you, if you get to the heart of it, like people are going to reasonable. I don't know what I'm just saying from this experiment, I don't know which theory is correct. It could be, it could be that you, you know, it could be that I hate you for the fact that you just speak this way, like even though I know that you have more moderate. Like anything could be true. Yeah, it could be. We did, my co-authors and I try to related kind of, I mean, it's a little bit, it's a little bit different, but we actually asked a bunch of scholars, a bunch of political behavior scholars to say, to try to predict how regular people would explain their partisanship. And then we also asked them like, now what do you think are actually the, you know, the causes of their partisanship as opposed to what they'll say. And there are some important differences, at least the political behavior scholars themselves, they anticipate some differences along those lines. They think what people are going to come up with, you know, reason sensible, like they'll talk a lot about policy, but they don't really mean it. What really motivates their partisanship is something more like group identity, things like that, but the differences weren't huge actually, but they anticipate some difference. And I'm sure that's true for the general public as well, if you could sort of ask them like, what is there, what, you know, ask a Democrat like, what would our public and say, and what are they really thinking, and maybe they would give two different answers, and that would kind of interesting. I don't know how you would actually. Yeah, I mean, so. That's a good question. How do you educate it from like, if you were, if you were to ask me, what do I think is going on in this experiment, I think I would say, people know that people don't perceive themselves as extra like most of the people don't walk around hating other people don't want the right. I want all the babies to be here. Wait, but if they don't go around hating other people, they're genuinely not extreme. I mean, I think they don't actually hate each other. When I think about other people, I, you know, suppose I'm a Republican, do I really think that all those Democrats want to run around and kill all the babies? Like, obviously, I don't think this way. Like, I see this people every day on the street, but they are not like, you know, going and with a stick hitting children on the head. But I truly believe they are okay with late-term abortion. And I truly believe if they really weren't good people, they would have stopped and they would have thought about this and they would have said, this is wrong, this is immora, and that's why I hate that. I hate them because they are deep down immora people. But, you know, I have no evidence from the survey, but this is the case. That's just like my one of my, I mean, I think, I mean, I think I think the hate is overblown in this literature. Like, people don't really hate members of the other party with, you know, I'm sure there's, there are exceptions, but I do hate the Packers, I'm sitting here. To the extent that there is some level of partisan animosity, I think there's pretty good evidence that it is, it is motivated by actual policy distribution of which there's plenty in American society. Even though a lot of people are somewhere in the middle, there are certainly plenty of people who have different values than other people. And that's partly, that's largely what, it's not some, like, purely kind of naive, like, they're on the other team, they're the Packers, and I'm the Bears, and I hate those people. I don't think that's really what's going on here. And so I think you're probably right, there really are some people with fundamentally different values on the abortion question. And they, maybe they can or can't empathize, I don't know, but they just really, really dislike people who hold the whole different opinion on that question. Do you have anything wrong with that? Like, that's, I don't know, that's the way what it means to live in a large heterogeneous society. Do you think, I don't know how we have evidence on this, but like, I think that I feel, are you going to help me? So, yeah, like some, yes, some, some nonsense theory. Okay, but like, I'm going to say one thing you agree with, and then one thing you're going to make, make, make, make you want to throw me off the podcast. Yeah, exactly. Okay. So politicians across the board are pretty awful. That's the part you're going to agree with. Yeah, okay, right. So, you know, the Trump is administration genuinely is taking bribes through cryptocurrency and by administration, genuinely head from the American public is mental infirmity. We are as human beings maybe inclined to see the good and the people we disagree with and the bad, oh, sorry, the good and the people we agree with and the bad and the people we disagree with kind of thing. And so it is pretty easy whoever you are to like see the evil in the other party because there's plenty of it to find if you just go and look, is that crazy? I'm sure there's lots of evidence and psychology to support this kind of thing that people are good at, you know, this kind of confirmation bias kind of thing. People are good at, I already know I like that person so I can find arguments in favor of liking that person and vice versa. My interpretation on the politicians is a little bit different from yours, which is I think rightly or wrongly, perhaps wrongly, most Americans just don't care that much about the bribery and the corruption and the lying and the cheating and the kinds of things the politicians do. What they really care about is policy. They care about taxes, they care about abortion, they care about health care and they care about the actual policies implemented by the government. And when those policies are aligned with you, you're willing to make excuses for that side because the thing you really care about is the policy. So do most Americans really care whether the president is getting a blowjob and the Oval Office, the answer is not really and how do I know that? It's because when it's a politician who's delivering policies they like, they say, "Oh, it's fine." And when it's a politician, they don't like they say, "It's terrible." And it's not that they're like, "It's not that they're picking." It's really that they're deciding on the candidates for other reasons and then, of course, the rhetoric goes around the stuff they don't, you know. - Yeah, I agree with that. I totally agree with that. I just wonder then if you ask them, how do you feel about the other side when they're thinking about the other side, it's easy for them to identify hateful things. I don't know. But I agree, there's no, it seems completely clear that we are willing to forgive. I know this can come out in a while, but the New York Meryl election notwithstanding, we're willing to forgive very bad behavior by people who are delivering policies we agree with. And totally, I mean, and yes, and hyper-sensitive to bad behavior by people who are delivering policies we disagree with. - To what extent is it really important how Americans think about vote just from voting for the other party? You know, when I think about where my parents live, there is a lot of polarization, political polarization, and half of the village vote in one way, half of the village vote in the other way. But they are all neighbor, and they, like, when they talk about each other, they just absolutely hate this idea that the other half have voters for the other party. But, you know, like they help each other, they live together, and-- - That sounds like socialism. - This has happened. You know, when the cows and the ditch. - I mean, I think what you're describing could apply to lots of towns in America. - I'm concerned yet, yeah. - And I think, and I think a big part of it is, there's actually a lot of moderates that live in these towns, and they might end up slightly on one side or the other in terms of deciding who to vote for, and the choices are very stark. And so they really strongly just like the candidates, they don't, you know, so I don't strongly dislike my neighbor who voted for the Republican, but I really don't like the Republican candidate. And vice versa, he really, you know, he really doesn't like, he likes me fine, but he really doesn't like the candidate, I vote for it, et cetera, you know, that kind of thing. I suspect that's going on a lot of ways. - Are we supposed to talk about the philosophy of science part of your interview with him? - We don't have to, we can. Do you have, what do you have to do? - Unless you have, you want to have my life. - I did laugh. I mean, I think I had more, I had the most genuine laugh I've ever had doing one of these interviews. He said he found himself getting dumber in the process of writing academic papers. - Oh, I did hear him say that. - He's advertising them for, you know, for editors. - So he's got these two different things. He's got one about like the sociology of science and he's got this other about psychology being a pre-paradigmatic science, where I think I don't know enough about psychology to know I'm sure there are problems with psychology as there are with political science, as there are with physics. I think this distinction he makes between, he has this, so he says a couple things. So one, he says like psychology doesn't have a paradigm by which you at least interpret him to mean like they're engaged in kind of barefoot empiricism, which I'm not sure is what he meant. But if he did, obviously I'm not a fan of barefoot empiricism. But then he said this thing about he made a distinction between depression and atoms. And he's like, atoms are real. Like atoms are an explanatory category in physics and they're real in a way that like philosophers of science used to use the phrase they cut nature at the joints is like I think the kind of thing he has in mind that there's, he has a kind of correspondence theory of truth, right, that like there's this object and the model's called atoms and there's an object in the world and there's a mapping. I think that's a crazy view of what atoms are. But put that aside for a moment. Do you believe in atoms? I believe all of science is engaged in building models and some models are more stable and more pragmatically useful than others. And when I say I believe in something, what I mean is that they are part of a model which proved really useful. I don't think there is a thing. And like I don't think there's natural kinds of that. So do I believe in mammals? Yes, I believe in mammals are a useful category for thinking about the structure of evolution, et cetera. Do I believe mammals are real in the sense that like you could not have categorized, you could not have created another model of whatever that had different category, whatever. No, I don't like, there's bleeding in between, like the world is more complicated and that it's all model building and it's all pragmatics is how I think about these things. And so then he says, two things, one of which-- Let's just say Adam's, the atomic theory, is it really powerful? Super useful. If you want to ask me, has physics made more progress in solving its problems than political science has in solving its problems, the answer is 100% yes. Of course they have, but not because of a correspondence theory of choice. OK, then he says this thing about depression. He said two different things about depression. So when he says all depression is for psychologists, and I don't know if this is true because I'm not a psychologist, is the answers to a survey. And if that's true, I'm 100% with him on the-- that's an old logical positiveist position that there is no distinction between the way we measure a concept and the concept itself. That's the kind of standard. That's also in the background of some parts of revealed preference theory that all preferences are is behavior. I don't believe that. That's a thing that a decision theorist people sometimes say. And what do we mean when we say, as an economist, what do I mean when I say you prefer apples over oranges? I mean, I gave you a choice between apples and oranges and you chose the apple. The all that preference is is a description of behavior. All that depression is is a description of the answer to the surveys. I think that's crazy. Preferences are, I think, the way I at least think about them, they don't have to be a mental state, but they are an object in the model themselves. They are not simply a measurement. And if psychologists have convinced themselves the old 1950 style naive logical positiveist position that all depression is, is the measurement of depression, they're really going in a counterproductive direction. That said, I don't think this position, like he then said this other thing, which is depression as measured by the survey, may actually be 20 different things going on in the brain. And again, of course, that is almost surely true. That doesn't mean that depression is fundamentally different than atoms as an explanatory category. It's a model. And it should be evaluated against its pragmatic usefulness, its explanatory power. And not all atoms are the same thing either. Like different atoms are made up of different things going on underneath. And certainly not all mammals are the same thing, but mammals is still a useful category for answering certain kinds of questions. So I think this idea that there are real things and not real things as opposed to objects in models, which are then evaluated against their explanatory power, their pragmatic power, and their stability as explanatory concepts in the science, is just like it's a kind of naive view of the way science progresses. And so 100% psychology has not made as much progress as physics. Political science has not made as much progress as physics. But the idea that atoms and depression are fundamentally different kinds doesn't seem right to me. They are both model objects that we use for explaining the world and want to prove more useful and lasting than the other. I agree with all of that. I think there's a lot more agreement that you would have if you had a longer conversation with Adam, I suspect. And I'm mostly interpret his writing on this topic. And he uses different words than we would use 'cause he's in it for field. But I think I mostly interpret him as saying, here's why we need more theory. And here's why theory could be proved useful. - And sorry, look, can I stop you? I hear it from a co-host who complained last week that I snuck in, and being pay for all to the podcast. I think that's my bottom line. (laughing) - I've never said theory's not useful or we shouldn't have theory in this podcast. I just complained about the bait and switch last time. (laughing) - I mean, there's the only way I can bridge your mind. - But of course, yeah, of course we should have, of course we should have theory in social sciences. And I think his view is, there's a bunch of psychologists and certainly this is true of political science and lots of other fields. They're just going around finding someone off thing that's kind of cute and fun. And they're not thinking about what they're doing or why they're doing it or what it's in service of. They're like, here's a fun finding because I had this particular measure and I had this other measure. And he's saying, we should write down. And we'll be wrong a lot of the time, but we should try to write down like, here's a grand theory of how these things are tied together. And then we should go test implications of that. I don't know if you want to go. grand theory. And I think like the thing is like categories like depression are a step in that direction. I would say like that's the part where I like I think categories like depression are an attempt to say like there's a bunch of different things but we can unify them in a structure that are for perhaps all responsive to the same kind of drugs or all responsive to the same kind of therapy or all whatever and then we can we can start to think about why might that be? Why are they all responsive to the same kinds of drugs? Maybe there's similar things going on in the brain maybe even I don't know like I again I'm way outside my depth here but I'm just saying like the barefoot empiricism critique seems to be intention with the like depression is this not natural kind category that we've used to lay over lay over a bunch of different brain phenomena or whatever that seems like actually the direction you want to go one direction you might want to go another you might want to have a neuro physics or something I don't know but I tend to be a little bit skeptical of that kind of reductionism but I don't know is that my deal? Ethan what's your bottom line? I think my bottom line is that we should launch a not another amateur philosophy science podcast where we can say all sorts of ill-informed things about philosophy of science and amused professionals what's my other bottom line? I think okay so you know I think like this project of trying to think about a new strategy for crossing off the list of potential explanations for why we talk about and behave the way we do in our political sphere to try and cross off that list that we just fundamentally misunderstand each other it's like that's I think that's a good question it's an interesting it's an interesting idea I am convinced by viola that I should be less convinced this has done that than I then I then I came into this conversation thinking I do think it is possible that when given this task people are trying to mimic what they think people expect a reasonable person to say and that they worry that if they say they're true more extreme beliefs they will look like a caricature from the other side so I think there's further work to be done I think work along the lines of like suppose you were only communicating with like write an essay that other Republicans will recognize as a Republicans like or that other Democrats will recognize it might be an interesting next step to try and see if you can get people to be revealed more about themselves but I think it's an interesting project you mentioning a long interview that people you know in the past we're running around widely and trying to turn the ledding to gold up actually you can turn led into gold it's physically possible do you have any do you have any of this simulated gold here in the room or any ledding to gold he talked about he was giving this example of like psychologists are alchemists and it's like it's not like the alchemists didn't make any progress the alchemists in fact made a fair bit of progress even though they were working on like with non-natural kind categories you can still make progress inside non-natural kind categories which should have led him which should have led him to think oh there isn't the thing in natural kind it's all pragmatics and like you work within something and make some progress and then you come up with a new thing but anyway all right good job you all right you can turn led into gold I just yeah I just want to emphasize I the reason I picked this paper for us to discuss is because I think this is a this is a relatively new and creative way of setting a long-standing question so I like that idea I think I think we should be more creative and how we ask some of these questions you know this there have been all these grand claims about partisan animosity and partisan misperceptions and and I think this was a completely new angle to come with this question from and I think it did we did learn something useful I was generally surprised by the result I'm not I mean I'm not shocked that like people have a better understanding of the other side than you might think but I was generally surprised that people were as good as they were thanks for listening to not another politics podcast if you like what you heard please leave us a review if you don't like what you heard please send an angry email to one of my co-hosts our show is a podcast from the Harris School of Public Policy and is produced by Matt Hodepp and Leah Seasreen