Chicago Booth Review Podcast Is America’s Cultural Divide Growing?

Emir Kamenica: Liberals and conservatives have grown further and further apart over time. It's much easier for me to guess whether you're liberal or conservative from your answers to the question, the general social survey today than it was five years ago. And it was easier five years ago than it was 10 years ago, et cetera.

Hal Weitzman: Many of us have a general sense that as a society we're not just becoming more polarized politically, but we're also moving further apart culturally. What we buy, what we watch, how we spend our spare time, these consumer behaviors are also increasingly diverging, but are they? Welcome to the Chicago Booth Review Podcast, where we bring you groundbreaking academic research in a clear and straightforward way. I'm Hal Weitzman. Today, is America's cultural divide growing? Chicago Booth's Emir Kamenica has studied the cultural divide in America. While his research confirms a widening gap in social attitudes between conservatives and liberals, it also reveals that the culture gap is remarkably stable. How did he arrive at that conclusion? What does it mean for society? Emir Kamenica, welcome back to the Chicago Booth Review Podcast.

Emir Kamenica: Thank you for having me.

Hal Weitzman: Delighted to have you back. Last time we had such fun talking about what's newsworthy and what's boring I guess, that we had to have you back to talk about culture. You've done research on culture, cultural divides, how consumption can reveal all sorts of interesting things about people. I guess my first question is, you did this research with Marianne Bertrand from Chicago Booth, your colleague, our colleague. What are two economists doing studying culture? We typically think of culture as the world of sociologists or others. Why would economists be studying culture?

Emir Kamenica: Well, I think this paper really dovetails in a number of questions that economists have been thinking about for quite some time, and there are a lot of topics of relevance to economists that, I think, doing the kind of measurement when doing this paper and a lot of topics that are informed by this kind of measurement. So perhaps most obviously economists along with other social scientists, including political scientists, have been really interested in the topic of polarization about the extent to which liberal and conservative individuals in the US have been growing apart on various measures. And one thing that seems worth knowing in the context where we talk about polarization so much is to understand with a bit more nuance, the extent to which the lives and the common references and some kind of cultural unity is changing or not between liberals and conservatives. So in this paper we look at not only the extent to which people who self-identify as liberal and people who self-identify as conservative give different answers to politically charged questions such as whether abortion should be legal or whether government should spend more or less money on helping the poor.

We also ask questions like, what about liberals and conservatives choice of what TV shows to watch? What about their choice of what kind of food to eat? What about their choice as to how to spend their free time? What are their hobbies? And one might really want to know whether as these political views are diverging, and we certainly in our paper find that the perception that the political views as in our context evidenced by questions from the general social survey of the variety I just mentioned between liberals and conservatives are in fact diverging. So the data for that particular dimension of culture, if you want to think of it as that starts back in 1972 and from '72 until our sample lens in 2015, liberals and conservatives are growing further and further apart in their stated political views. And of course in different other dimensions as well.

For instance, liberals and conservatives in 2015 watch different things on TV. And one can, in this paper we provide a certain quantification about well, how different are the things that they watch and the approach that we take, which allows us to sort of take this highly multidimensional behavior like which TV shows you watch, and sort of measure the distance as just this given number. We say, suppose I just know there's an equal chance someone is a liberal or a conservative, so 50-50 chance, and then you tell me the TV shows that they watch with what accuracy can I now guess knowing what TV shows they watch, whether they're liberal or conservative. If the set of shows being watched is exactly the same, then obviously we're going to get the lower bound where my probability of guessing correctly whether this person is liberal or conservative remains at one half.

Suppose that the shows are completely disjoint. So no liberal ever watches conservative shows, no conservative ever watches liberal shows. Then the moment you tell me what shows you watch, I can tell would probability one, obviously the truth is going to be somewhere in between. In 2015, the truth is something like 0.7. Okay, if I know what TV shows you watch, I can correctly guess whether you're liberal or conservative with a 70% accuracy, which even by itself I think is kind of interesting. It is a quantifiable measure of the difference in the overlap of what kind of TV shows people watch. But really the primary focus of our exercise is taking this number like 0.7, my ability to guess whether you're liberal or conservative, how has that changed over time? Our data on TV shows goes back to the early '90s and you might think, well, it was a completely different world in the early '90s.

Fox News did not yet exist. The kind of shows that people watching, whether liberal, conservative now are sort of long gone out of popularity. So you could ask, okay, to what extent has this similarity of what liberals and conservatives watch changed over time? And somewhat shockingly we discovered it hasn't changed at all. So whether you look in 2015 or 2010 or early 2000s or mid '90s, you are around the probability that I can guess correctly, whether you're liberal, conservative based on what you're watching is always hovers close to 70%, which is certainly not what we sort of knew we would find or even expected we would find before we looked at the data, but that is what we find. So in that particular dimension of grouping people, liberals versus conservatives, and that particular dimension of culture, namely what TV shows you watch, while there is a cultural gap, that cultural gap has been very stable over a quarter-century.

And in the paper we look at other cleavages you might call them, so not just liberals and conservatives, we also look at rich versus poor, and men versus women, and whites versus non-whites, educated versus less educated. And then we look at many dimensions of culture, what TV shows you watch, what are your political attitudes, how you name your children, what kind of food do you eat, how do you spend your time. And with only a couple of exceptions, we find in fact that while there are cultural divides, while there are groups are different, the extent of that difference has largely been stable over time. But I sort of diverged and didn't finish to answer your question, why economists thinking about this, and why is of this interest to economics? I briefly talked about the issue of polarization, but there's other issues. There's issues having to do, for instance, with intergenerational income inequality. Think about a world where the rich and the poor who obviously have different income by definition, watch the same TV shows and eat the same food and have the same cultural references, compared to one where those are entirely different.

Think about what that means for ease with which say a kid from a low SES family that gets into a top college can sort of integrate and form friendships with the on average much more affluent fellow students. A world where there's a big cultural gap between the rich and the poor, so that we don't eat the same food, we don't sort of have the same character references from the TV shows we loved. It's a world where those social frictions are probably going to make it more difficult for children from low SES families to sort of fully integrate-

Hal Weitzman: SES being?

Emir Kamenica: Socioeconomic status.

Hal Weitzman: Okay.

Emir Kamenica: Sorry.

Hal Weitzman: To integrate, so you're saying that the cultural divide actually is in itself a barrier to social mobility?

Emir Kamenica: Certainly might be, yeah. And then there's a question of, as of course income inequality has gone up in the US, one might ask, was that accompanied by a widening gap in cultural distances between the top and the bottom quartile? And we find that on most dimensions, no, it hasn't. Of course rich and poor people behave differently, but no more so than in the past. And then lastly, if you think about gender differences or race differences in terms of how represented they are at in certain position society, you could imagine a world where it is just because people are humans and they're influenced by social relationships. It could be that a boss in a firm is going to end up being closer with employees who share the same interest, have the same cultural references.

So to the extent that given that we live in a world where many people in a position of power are white men, the extent to which white men have different interests than women or white men have different interests than non-white men, you might think that that's going to also play out in the ability of those from sort of less currently represented demographics to get promoted and kind of make the social aspect of the world, whether it's the workplace or their educational environment or whatever it is.

So navigating those places is different depending on the cultural.

Hal Weitzman: That's fascinating because economists think about frictions, don't they? Why doesn't meritocracy just work? So one of the frictions, it sounds like is culture, this reminds me that I should tell our readers that it is worth going back to your original paper and reading it because it's a beautifully written paper. And I believe the opening line is something like, "Half of America is watching Duck Dynasty if that show is even still going and half of it is doing yoga in the park or whatever." So just you capture there very nicely the divide which you're saying is not just about preferences but actually about prospects.

Emir Kamenica: Yeah. Actually the punchline of the paper is really about there being no substantial trend in cultural divides in most dimensions and most groups, but for an academic paper there is on top of this punchline, which is just you might think is not that exciting, but we find exciting just things have not been changing. The paper does include a lot of sort of tasty popcorn sized bits about what are particular cultural behaviors that distinguish rich and the poor, and liberals versus conservatives. So we find some delightful things like that actually purchasing Grey Poupon Dijon mustard in the late '90s is in fact a very strong predictor of being from the top versus the bottom quartile, which anybody was around for the time of the Grey Poupon commercials will find rather amusing. We can ask fun questions like, of all the movies that were in wide release, so about a hundred movies a year gets released in the US every year and for every one of those movies you can ask the question, how well can I learn, say whether you're liberal, conservatives, if I ask you a question, did you watch that movie?

The answer is always going to be between 0.5 and 1. 0.5 is where the likelihood of seeing the movie is the same, whether you're liberal or conservative. One is a theoretical upper bound, where there's a movie that every liberal has seen and no conservative has or vice versa. So you can ask fun questions in any year, what is the most divisive, so to speak, movie that was released and you discover things that are interpretable like Brokeback Mountain ranks very highly as a movie that liberals would watch and conservatives did not. So I learned a lot about your political leanings from knowing whether you watched Brokeback Mountain the year it was released.

Hal Weitzman: Right. It made me think that political campaigns are probably contacting Netflix to get your streaming history so they can determine how you're going to vote in the next election. But as you said earlier, I want to go back to this divide part because we do have the impression, don't we? That at one point, all of America was happily watching I Love Lucy or whatever, and now we just the worlds that just don't touch each other. You're saying that is somewhat the case, but it has been the case for a significant amount of time and it's not getting worse.

Emir Kamenica: Yeah. I mean our data goes back to early '90s, I think I Love Lucy was popular earlier than that. So starting from early '90s onwards, the extent to which people are watching the same thing, and of course there are shows that are not particularly indicative of your income or education or political leanings, but the prevalence of those shows relative to the shows that are in fact divisive if you want to call it, that divisive suggests that it's causing division. So I think divisive is the wrong word. The shows that are kind of diagnostic of your income or race or political leanings, their prevalence and their importance in the portfolio of consumption doesn't seem to have changed for over a quarter of century.

Hal Weitzman: I mean in a sense this isn't that surprising. People make products and they sell Grey Poupon, I don't remember the ad because I wasn't living in America there, but it's something to do with it's a fancy product basically.

Emir Kamenica: Yes. It's people are passing it between limos exactly.

Hal Weitzman: I see you have a butt like serving you your mustard. So that makes sense. But it's done as a joke, but it's also done partly seriously because people have this association that that is an expensive product and therefore worth whatever premium they have to pay to buy it. So products are designed for different people. I mean in America there was a long history of cigarette companies trying to sell menthol cigarettes to Black people, and specifically targeting that demographic. So sure enough there were higher instance of menthol cigarette smoking among in the Black community. So I mean that in itself shouldn't be that surprising. Similarly, with cultural products, they make TV shows for certain demographics, right?

Emir Kamenica: We're not finding that the presence of a gap is going to be surprising. We're really kind of interested in the extent to which that gap. First, let's quantify it. Let's get a number on it, so it's not just, well obviously there's a gap. How big is it? We provide a measure that lets you put a number, it's a number between 0.5 and 1 on every particular cultural gap and then ask how has evolved over time and find that with a couple of exceptions, and those exceptions are interesting in their own right, there hasn't been much change. So what are the exceptions? One exception, as I mentioned the outset is that in political attitudes, liberals and conservatives have grown further and further apart over time. It is much easier for me to guess whether you're liberal or conservative from your answers to the question, the general social survey today than it was five years ago. And it was easier five years ago than it was 10 years ago, et cetera.

So that's one place where we do not have lack of trends. And the other place we don't have lack of trends is how men and women spend time when, we're only looking at men and women who work full time, how men and women spend their time when they're not working. These are amongst the full working individuals. We find a lot of convergence from 1965 to 1995, where if you go back to 1965, there were certain activities where highly indicative of your gender. That gender gap is greatly reduced from 1965 to 1995. But somewhat remarkably, there's been no further convergence since 1995. So that's in 20 years of our data, the gender gap is remaining-

Hal Weitzman: In terms of activities [inaudible 00:16:12] whatever.

Emir Kamenica: ... in terms of activities outside of work. And this is interesting because the timing does parallel certain findings about gender gap in wages and then gender gap in labor force participation, which also saw a lot of gains that kind of stopped progressing starting in the mid to late '90s.

Hal Weitzman: If you're enjoying this podcast, there's another University of Chicago podcast network show that you should check out. It's called Not Another Politics Podcast. Not Another Politics Podcast provides a fresh perspective on the biggest political stories, not through opinions and anecdotes, but through rigorous scholarship, massive data sets and a deep knowledge of theory. If you want to understand the political science behind the political headlines, then listen to Not Another Politics Podcast, part of the University of Chicago podcast network.

Emir, in the first half we talked about this cultural divide I guess, and not that it's divisive. That was a good point that you made there about that. So it is a divide. It is somewhat stable in terms of our cultural consumption. But you talk in your paper about cultural capital. What's cultural capital and why is it important?

Emir Kamenica: Well, I mean we did in the first half talk about things that are closely related to cultural capital. When I talked about the ease with which a student from a low socioeconomic background can integrate with wealthier peers in colleges. That ease of assimilation, that ease of making friends, that ease of finding yourself at home at an environment that is not one that your family is as familiar with. That's precisely, I think, what people have in mind. The sociologists that have written about this had in mind when they talk about cultural capital. In order to do well in school, it's good to know the material, but it's probably helpful and there's been research on this done by others that it's also helpful to have friends and get along with your peers, and one's ability to do that does rely on this cultural capital which allows you to assimilate.

Same thing we talked briefly about bosses who might be a white man, having closer social connection with the employees that match them on culture. That's precisely so what an employee might be lacking is the cultural capital that allows them to sort of have the same kind of rapport with their boss.

Hal Weitzman: That's not static presumably. I mean you can be raised on Duck Dynasty and end up going to the opera.

Emir Kamenica: Well, of course it's not static. And one of the things that we are sort of measuring here is the extent to which it is. Your question specifically concerned about how the-

Hal Weitzman: I mean when people progress, if someone is successful, whatever their roots, they are-

Emir Kamenica: A great question-

Hal Weitzman: ... then access all sorts of things.

Emir Kamenica: You're pointing to what is a great question, which is the one of intrapersonal. So look at a person over time, especially a person maybe whose circumstances change. To what extent does their cultural consumption change? Our data does not allow us to speak about this at all, because we don't actually see particular individuals repeatedly over time what we have what is called a repeated cross-section, so we're not seeing the same person in one period of time, another period of time. A super interesting question that, I think, would be with if one can find the right data set to explore it would be the extent to which a given person's cultural capital, cultural consumption, habits, et cetera, evolve as their, say, socioeconomic status evolves or an extent in which people move from one city to another, the extent to which they sort of pick up the cultural norms of the place that they move to versus maintain the cultural behavior and cultural capital of their place of origin.

Hal Weitzman: Okay. I'm curious because you are covering quite a lot of time in your research, how do you account for changing products and TV shows and media titles over the years?

Emir Kamenica: It's a very good question. One could imagine wanting to eliminate changes that are due to what an economist might call supply side factors, like what are the products that are available, what are the TV shows that can be introduced. And the perspective that we take is that we don't see those as confounds in any way. We see those as some of the reasons why cultural gaps might have widened or gotten more narrow. So yes, it is a fact that the number of TV shows on offer is very different in 1992 than it is in 2000 or in 2015. We don't think that as something that we want to remove the effects of to the extent that, and we find that there was no extent, to the extent that the cultural gap widened because there's just many more TV shows now than there were in the past, though theoretically could go either way.

Well, we would find that to be a perfectly fine reason why the gap widened or narrowed. So we are embracing as just facts of the world, the fact that the world is changing in various dimensions, including the number of products offered and say given that background change, what is happening to the cultural gap? I mean same thing applies to income inequality. Income inequality is widened, so maybe that's going to lead to a widening cultural gap between the top quartile and the bottom quartile. We don't find that widening gap, but if we found it very well could be due to these structural changes in the background.

Hal Weitzman: But you accommodate the fact there's been this massive fragmentation in everyone is not watching I Love Lucy?

Emir Kamenica: Correct. And to the extent that led to people, cultural gap between the two groups becoming small, larger, we would've detected that change.

Hal Weitzman: Okay. So we talked a lot about culture. So let's just go to the political part of this and you talked about how you actually do confirm this suspicion or intuition that many people would have that we have greater political polarization in terms of our views. And it's interesting your sample ended, I guess the year that Donald Trump was elected the first time around.

Emir Kamenica: Yes.

Hal Weitzman: Fascinating coincidence there. But what do you make of that growing gap in social attitudes then that you confirmed, I guess?

Emir Kamenica: Well, what to make of it? It is certainly an interesting fact. It is a fact that I think is important to bring to the table along with many other facts and many researchers have brought to the table relating to behavior of politicians, the extent to which politician one party versus another are more or less likely to vote the same way on any particular topic. The facts that people have brought to the table about where politicians who are running for office position themselves vis-a-vis the preferences in the cultural space versus economic space. We're just trying to, along with many other researchers, try to bring facts to the table about the causes and consequences of polarization in many forms. And that ends up being one of the contributions of our paper, but not the one that we really focus on relative to most of the research that's much more squarely focused on polarization per se.

Hal Weitzman: To go back to our conversation we had on the last time we had it on the podcast about what's newsworthy, I guess the combination of the two is newsworthy, right? That on the one hand you are confirming polarization politically, but on the other hand there's no polarization in terms of cultural consumption.

Emir Kamenica: Yeah. You might've thought that as political views diverge, so do many other aspects of life. You might think that the way that liberals or conservatives spend their free time becomes more dissimilar or what they watch in the television becomes more dissimilar. We don't find that. So that is sort of a notable contrast with a very clear, just jumps out at you in the data, increase in the cultural distance and the distance in political attitudes between liberals and conservatives.

Hal Weitzman: Right. So to go back to again to our last podcast where we talked about newsworthiness, were you surprised by this trend where you confirm the political divisions or the polarization, but you found quite of a steadiness in what you call the cultural divide?

Emir Kamenica: I didn't have too strong a prior, so I wouldn't say I was too surprised. It's difficult to be too surprised if you're really quite unsure what you're going to find. I do think that it was fortunate in a way that what we found is so easy to summarize because we do look at these various cleavages, so liberal versus conservative, rich versus poor, educated versus less educated, men versus women, white versus non-white. And we look at all these different dimensions of culture, political opinions, how you name your baby, what TV shows you watch, what movies do you see, what kind of food that you eat, et cetera, what are your hobbies. So we could have gotten a hodgepodge of results, where like, well, for this particular combination of cleavage and dimension of culture, that's been increasing. This other one's been going up and then down, this one hasn't changed, this one went up and then down and then back up.

So we basically were looking at something like 24 different trends, and it could have happened that there was no easy way to summarize what we find. It was unexpected to find that in fact the summary is really easy with only a couple of exceptions, namely political attitudes between liberals and conservatives, and time use between men and women. With those two exceptions, pretty much everything else, the big punchline is there really has not been a movement in cultural gap regardless of which dimension of culture you look at and regardless of which two groups you're considering. So I do think I was pleasantly surprised that the answer is so sort of can be so neatly packaged.

Hal Weitzman: Right. You're pleasantly surprised because you could write a paper about it. And you could come on the Chicago Booth Review Podcast, and we're very glad that you did.

Emir Kamenica: Exactly.

Hal Weitzman: I have to say next time I eat Grey Poupon, which I now do buy, I confess or watch Duck Dynasty, which I don't, but I'm going to start doing that, I'll be thinking of your research. Thank you very much, Emir Kamenica, for coming on the Chicago Booth Review Podcast.

Emir Kamenica: Thank you.

Hal Weitzman: That's it for this episode of the Chicago Booth Review Podcast, part of the University of Chicago Podcast Network. For more research, analysis, and insights, visit our website at chicagobooth.edu/review. When you're there, sign up for our weekly newsletter so you never miss the latest in business-focused academic research.

This episode was produced by Josh Stunkel. If you enjoyed it, please subscribe, and please do leave us a five-star review. Until next time, I'm Hal Weitzman. Thanks for listening.