For a special edition celebrating the 100th episode of the Science of Politics, Matt talks with Ezra Klein about how well political science informs American politics and public policy. They discuss how political science has changed in the age of Twitter and the era of Trump and the roles of scholars and journalists using research in debates on climate, COVID, and race.

Guests: Ezra Klein, New York Times

Study: Why We’re Polarized


Matt Grossmann: The role of political science in American public life, this week on the Science of Politics. For the Niskanen Center, I’m Matt Grossman. We’ve come to the 100th episode of the Science of Politics, thanks so much for listening. To celebrate I invited Ezra Klein on to discuss how well political science informs American politics and public policy. He is a New York Times columnist, a founder of Vox and host of the Ezra Klein show. He’s been at the forefront of popularizing, integrating and critiquing political science research, including in his recent book, Why We’re Polarized, now out in paperback. We discuss how political science has changed in the age of Twitter and Trump and the roles of scholars in debates on climate, COVID and race. Here’s our conversation.

It’s been almost seven years since you wrote that the rise of political science was the best thing to happen in political journalism since you’ve been in it. So has it continued to rise? And to what extent has that remained a positive influence?

Ezra Klein: Ooh, it’s a good question. So I think a lot of the gains were there at the beginning, so I don’t think you keep having quite the same level of the curve. Let me be more critical of political science over the past couple of years, because I continue to learn a tremendous amount from it, but I’ve also become a little more measured in that judgment, which is simply to say, and you could make this critique of much of my traditional work as well, political science is extraordinarily valuable at describing what is, and it has a lot of trouble with what isn’t yet or also what ought to be.

And I think the great failure of political science over the past couple of years, it’s something that has, I think, properly increased a certain amount of my skepticism and others and that is believing in the boundaries of its own models a little bit too much. The tremendous collapse, for instance, of The Party Decides model in 2016, which I think led to a lot of people, including a lot of political scientists I was talking to in that era, very much underestimating the possibility of something like Donald Trump. So I drive tremendous benefit continuously from political science, and I think it is nothing but good that there is so much more political science in the journalistic and Twitter political conversations, but I also think it’s a, like anything else, it is one of many tools in the toolkit.

Matt Grossmann: So we do have a reputation for describing that things are broken and why they’re broken rather than necessarily sort of offering solutions for them. But in the past, you’ve sort of seen our structural theories as sort of adding to what the conventional wisdom looks like. So how valuable is it to say, this is a structural problem that will be difficult to solve without major institutional change that doesn’t seem to be forthcoming?

Ezra Klein: I think it’s extraordinarily valuable to correctly describe structural problems, whether or not you can solve them. I mean, obviously that’s a big part of my work, it’s a big part of my book. So my trust in that, and my belief in the importance of that, it really hasn’t changed. This is a question I will push to you though a little bit, because one of my intuitions is that the political science conversation has changed quite a bit since political scientists took to Twitter. So when I write that seven years ago, I’m really talking about political science in the age of blogs, in the age of [inaudible 00:03:28] blogging, the monkey cage and work you and others are doing.

And now when I think of where the political science conversation is, it seems less centered there, and this is true for a lot of different disciplines that converge around politics, and it is on Twitter. And I would say political scientists are very different on Twitter than they are in blogs and they are when I talk to them. So an interesting-

Matt Grossmann: Because they’re more activists or because they-

Ezra Klein: Yeah, because they’re-

Matt Grossmann: Yeah.

Ezra Klein: And also because of ones who are bigger voices are the ones who are more unsparing. Twitter compels, or at least incentivizes, a different kind of communication from all of us who are going to thrive on it than other mediums do. So I also think an interesting piece of this is that the political science conversation is a little bit different than it was then, and in many ways, a little less political science-y. This is not really about you, who I think manages to run a very, very, very valuable Twitter feed of just, here’s a bunch of papers and here’s a bunch of books, but I do think in general, the role political scientists are playing in the conversation has flattened on Twitter, much as it has for journalists and activists and to some degree politicians. So everybody’s just give and takes and is a little less differentiated in their expertise and in their role than was true 10 years ago, certainly than was true 25 years ago. But I’d be curious if you see it that way?

Matt Grossmann: Well, so let’s talk about some specific examples, because I looked through some guests that we’ve both had on our podcasts. So they include Leah Stokes, Jamila Michener, Frances Lee, Lee Drutman, John Sides, Lilliana Mason, Eitan Hersh. So tell me kind of what would cause you to bring political scientists to a broader audience? And is it that sort of line that they’re willing to kind of go beyond a little bit and serve these intermediary roles?

Ezra Klein: You dodged my question, Matt. Do you not give answers on this show, I’m the only one that gives answers?

Matt Grossmann: That is, I think, a piece of the answer is, is it that political science has changed or just that they’re sort of a niche of people who go beyond the research to be involved in a public conversation, and then that inherently kind of entails some activism.

Ezra Klein: Activism is, first, not the word I would use for any of this, it is more… Look, political science research, like a lot of complicated things, translates only so well to Twitter or to anywhere else. Some of those folks, I think, do it really well, and some of them, it depends on the day. I’ve had a lot of political scientists on the show, including the folks you mentioned there, and then obviously some others, like John Sides and [Baverk 00:06:11] and Tesler and so on. And I think the conversations we’re able to have on the podcast are just different, in form and in quality to what happens on Twitter.

But I think a bunch of the people you named are really good at Twitter. Like for instance, I think the work Leah Stokes does on Twitter is actually pretty invaluable. Her ability to bring the work she has done on climate politics to kind of continuous ongoing analysis of climate decisions is really, it’s… She’s one of the people who uses the platform extraordinarily well, some of the others that you’ve mentioned just don’t strike me as that active on Twitter. I actually don’t know if Frances Lee has a Twitter account or not. If she does, it’s not, to my knowledge, a heavily used one.

So I think that’s a complicated question, but it is obviously going to be true that the political scientists who I bring onto the show are the ones who I think of as the most talented public communicators and who have the most relevant researched ongoing issues. At the same time, that’s different than that being sort of the universal political scientists I, or other journalists, will benefit from.

Today, one of the things I’m working on is a column for this week that’ll be out, I’m sure by the time this podcast comes out, which is about the California recall election. And the ways in which California’s progressive era of governmental reforms, which were of course designed recalls, propositions, initiatives and so on and so forth, to increase democratic participation in the system, often end up doing almost the opposite, they become ways that small and organized groups can foil broader public will. And for that I’m relying very much on people like, Bruce Kane and Francis Fukuyama and his ideas of vetocracy.

And so, I don’t know for me, there’s still a tremendous amount of value in political sides, both the people who are hot in the public debate and not, but I do think that political science like everything else has changed in the place of a lot of that conversation is going on. I think it was true for economics, I think it’s true for journalism. If you want to be at the white hot heart of the conversation, it’s in a place that is necessarily a little bit less friendly to nuance and to complicated methodology sections.

Matt Grossmann: So you also relied a lot on political science and why we’re polarized, so talk about what you learned from that, but also what frustrated you about the answers you didn’t find?

Ezra Klein: There is no why we’re polarized, without the work I’ve done trying to learn from political scientists over the past 10 years. Why we’re polarized is very much an effort to merge traditional political reporting, where I talk to senators, and staffers, and presidents, and White House staffers and interest group leads and so on, with the broader structural analysis and the broader temporal analysis of political science. And so what I’m trying to do there, and what I try to do on a lot of my calls is merged as two types of reporting in a way I think you don’t always see. To my view, some of the weakness of pure political science is that it is thin on the way actors see their own actions unfolding and the pressures they’re under. But of course, the weakness of talking to those actors directly is they don’t always understand the structural context that eventually leads them to take the action they do. And so we’re all very good at rationalizing our ultimate decision, but we don’t always see what made that ultimate decision the one that made sense to us.

So I’m trying to bring that together and create, as I talk about it in the book, which is now out in paperback, and people should get, I’m trying to create a model for understanding the political system. In terms of the biggest frustration out of political science while doing the book, and I hope my present company will not take offense at this, the hardest chapter for me was the Republican and Democrats chapter, which was fundamentally a chapter about asymmetric polarization, and I would say even beyond that asymmetric radicalization. And I was really struck when I really dug into it, how much I had absorbed the idea that asymmetric polarization was a known and measurable phenomenon, which it is, and it’s true. And at the same time, there was not to me nearly as much in the way of causal analysis of that phenomenon than I had thought there was.

So some of the best work, I think, is work you and at Hopkins have done, but, as you know from the book, I don’t find it totally convincing, and I think it has some things that you don’t give enough credit to. But much more broadly, A lot of it was just very individualized. A lot of the work I saw out there on asymmetric polarization was tracking what Newt Gingrich did in the ’90s or what so-and-so did in the ’80s.

And to me, what that ended up missing was, well, why are these kinds of figures taking root and succeeding on the Republican side and not on the Democratic side? As I think I put it in the book, it’s focusing on the flowers and not on the soil. Different kinds of plants grow in different kinds of soil, and what you’re saying is that, repeatedly different kinds of Republicans are taking root than Democrats. So there is a difference between a party that has the same rational leadership, more or less than it has had since ’06, and a party that keeps deposing its own speakers and knocking out its leadership, like we just saw, of course, with Cheney.

And so I was struck by how weak the causal analysis based metric polarization were. And to me, it was useful because I ended up having to do a lot of work and I think came up with something valuable in terms of differing Democratic incentives, and then in a space I’m a little less confident in, and so I didn’t push it too hard in the book, but I think is true, even though I can’t prove it, different psychological qualities on the two sides. And particularly the fact that you have psychological sorting among whites and not among Blacks, or to a somewhat lesser degree, but still Latinos or Asians. It strikes me as really important.

So I appreciated in some ways that it forced me to do more original work, but for all the work has been done on polarization, that struck me as a pretty big blind spot and a weakness in the literature that people did not draw much attention to on the many years they spent telling me about it.

Matt Grossmann: So you’re a liberal commentator, in addition to being a promoter of research, and we are an overwhelmingly liberal discipline. How do you avoid confirmation bias? How do you work around it? And is there any responsibility we as scholars have for taking into account where we are on the political spectrum?

Ezra Klein: Well, I think it would be folly to say I avoid confirmation bias, I’m certain that I don’t. I try, right? I try to take counter arguments seriously and look at the methodologies, and I try to offer pretty fair readings of people’s views who are not on my own. So, I mean, that is one place where it’s valuable to me to not be fully within political science. As I said, I do a lot of political reporting. And so when I write about Mitch McConnell, I have spoken to Mitch McConnell’s staff, in an ongoing way for years to understand the way he thinks and the way he approaches things. When I write about the Trump administration, it’s not like I never spoke to anybody who worked for Donald Trump.

And so I’m not just within the political science approach on this, but the broader thing you’re pointing out is really hard. And I believe you’re working on a book about the diploma divide, if I’m not wrong? And so a book about educational polarization, and that strikes me as a real accelerant to this as the parties split by education. And I mean, post-grad, it’s just an overwhelmingly Democratic group now, that’s also going to show up more and more in the academic discipline. So there is no doubt in my mind that that influences and shapes political science. And on the other hand, it was easier. I don’t know what you’re going to do as a public and political scientist in the era of Donald Trump if you want to maintain a valuable identification as a Republican, and also be honest about what’s going on. As as identification with Trump and some of the things he says begins to become the nature of Republican identity, certain very tough choices about the truth and empirical work are forced upon people.

And so one of the difficulties for me, and this isn’t just in political science it’s even in just bringing people onto my podcast, a lot of the Republicans who are Republicans, who do not like the left, who do not like Joe Biden, they’re not really Republicans anymore. They don’t represent where the party is anymore. People who for years… Take somebody who I respect greatly, like a Yuval Levin, extremely influential within the House Republicans, certainly in the Paul Ryan era, who I think it’s fair to say it was a reasonably close confidant to Paul Ryan’s. And then as the party goes on, Yuval is somebody who is fighting for the future of conservatism, but I’m not sure that talking to him is giving me or giving my audience say, a fair representation of what the center of Republican Party is now thinking and doing. And so one of the difficulties with sorting in the way it has happened and then the Trump takeover and the way that has happened is that it has made it harder to maintain mixed compositions in empirically rigorous disciplines or in places where you want to have empirically rigorous voices. And I don’t really know what to do about that, but I’m also not going to a sit around and pretending or telling other people they should pretend that the Trumpified Republican party, that there are no costs to maintaining your membership in good standing.

Matt Grossmann: So how did we do in trying to understand democratic backsliding under the Trump Administration? There were several high-profile efforts. There was an increase in comparative and international relation scholarship on American politics. There are some people say we overreacted. Others say we under reacted. What’s your assessment?

Ezra Klein: I found it valuable. I don’t know how to rate that overreacted, under-reacted set of claims because of course, part of reacting sharply to a democratic threat, hopefully, at least theoretically, makes its full flowering less likely. But I think if you went back in time and you asked the people who said, “You’re under-reacting” to rate the likelihood of the events we saw after the 2020 election, they would have rated them very low and said, “That’s ridiculous. None of that’s going to happen.” And then it all happened.

It did not work, but it was a very out of sample event for modern American politics. Not of course for historical American politics, which I think is an important point, but for modern American politics. And so I’m not very friendly to the under-reacting crowd. It seems a little wild to me to look around in the current system and watch Republicans primarying other Republicans who did not buy into Trump’s post-election lies and say, “Ah, there’s no real threat to democracy here. You’re just overreacting.”

I think the rising comparative is really useful. I just did a piece. My July 4th column was talking to political scientists who were born and work outside of the US about how the American political system looks to them. I thought that was a useful piece and I enjoyed hearing their perspectives. And really, it made me a little bit more optimistic about our system, where I do think one thing… And maybe this speaks to the previous question too, Matt. I do think one thing that is underplayed in the way political scientists are talking about the current moment in democracy is that what we’re seeing is not mere democratic backsliding. It’s democratic polarization, right? There is a polarization. We are sorting, the parties of sorting in their views of democracy itself.

So one reason why, if you look at a list of how difficult it is to vote in different states, and we can argue about the methodologies of these, but you said there’s an example for this argument, why you will not just see a pure blue to red list is that voting has not traditionally been all that polarized and issue in sort of modern decades of American politics. Like for instance, mail-in voting, which is now so low that on the right, is something that took root primarily in more Republican states for years.

But what has happened the past couple of years is not simple backsliding. It’s also that the Democratic party has become committed in a way it has not been before to more democracy. I mean, obviously Democratic party has a deeper history in being a very anti-democratic party, right? I mean, it’s period as a Dixiecrat party, at least in large part. But in more modern times, it’s roughly pro-democracy, but it doesn’t try that hard at it. I mean, now they really are. Whether or not they pass HR1, whether or not they pass HR4, we will see, but they are committed to these things in a way that has not been true.

And so one of the things that I think the democratic backsliding conversation has missed is that unlike in a lot of the countries often used as examples there… Let’s not use like Nazi Germany here. Let’s just talk about I think things that are closer, like Turkey under Erdogan, Hungary Orban obviously people talk about all the time. You could say to some degree Russia under Putin. We could name a few more.

One of the things that tends to happen there is you have a party that actually is dominant, at least at the start, in terms of its public support. And then it uses its dominance among the public and among the machinery of the government to constrict the danger that elections pose to it. They have power and they use it to become a competitive authoritarian state.

That is not really the case here, and I think this is underplayed. What is happening is that Republicans are trying to do that as far I think as anybody can tell. At the same time though, Democrats actually hold more certainly national power and certainly national legislative and presidential power. And they are trying to use that power, again, may not succeed, to radically expand the boundaries and processes of democracy. I mean, if something like HR1 or even the Joe Manchin compromise version passed, it would be not the pre-Trump status quo ex-ante. It would be a completely new approach to election administration. I mean, a ban on partisan gerrymandering, election day’s federal holiday, this would not be new, but the revitalization of the Voting Rights Act. There’s a nationwide automatic voter registration of a sort. This would be a really big push forward.

And so the fact that we’re having a fight over democracy where at least one plausible outcome is a radical expansion of it is I think a little bit underplayed, in part because both pundits like me and the political science profession has been focused on the threat of Trump and it’s more loss averse then kind of imagining what are also the more positive possibilities of the moment.

Matt Grossmann: So let me ask a question about differences across disciplines that you mentioned, relying on psychologists, [inaudible 00:21:12]. And I know that you’re very familiar with economist’s role in public policy debates as well. So talk about political science relative to economists, sociologists, psychologists. Are they all moving in the same direction in terms of their public influence and their role in policy debates? Is there something political scientists can learn from the other disciplines?

Ezra Klein: I’m not sure I have a good answer to that, to be honest. My sense is economics has weakened quite a bit as a public player in the past couple of years. And not just, by the way, in the Biden Administration, but in the Trump Administration too. And so one, the economics discipline doesn’t speak with as clear voices as it did previously, but also it is not listened to in the same way that it was previously. I don’t want to say [Keynes 00:22:00] was right and economists are [inaudible 00:22:02], but they are much more secondary to political movements now than they were before.

And one reason for that, which I think is appropriate, is that one really significant critique to make of the economists, or the economists who operated within the political sphere, is that they did not understand political risks correctly. They maybe understood economic risks. We can argue which ones understood them correctly and which ones didn’t and then you get into whole arguments about boundaries of political system, but putting all of that aside, they did not understand the political risk that was being created by a weak economy and by a long period of undershooting full employment and wage gains. So I don’t want to say that to the whole reason for Donald Trump or other things, but I do think political risk is being taken more seriously than it has previously.

In terms of the other disciplines, I wouldn’t say I know exactly. I mean, I talked to a lot of sociologists. I find their work very, very valuable. Their work doesn’t tend to come with as much of a centralized political valence as public science. And by that, I don’t mean that they’re not liberal. Sociologists are very, very liberal. What I mean is that they are not organized in the same way to try to have a view on how American public policy or political coalition should change to be understood. And so there’s a lot of different stuff out there, but I do think the primary disciplines that I tend to work with are political science and economics. And I think that for them, frankly, like for everybody else, it’s been a hard couple of years. But in some ways I think probably compared to when I started in covering politics in 2005 professionally, I would say political scientists have raised up in status quite a bit. And I think economists have lowered in status quite a bit.

Matt Grossmann: And what about academics compared to think tank community or the more policy advocacy world? are there things that we could learn? Is there still a need for those kinds of translators of research for policy debates?

Ezra Klein: I would say in general, I think the think tank world has weakened quite a bit. That’s my impression as somebody who draws on a lot of that work. And I think there are a couple of reasons for it.

One is, again, the fact that there’s a flattening going on. So many more kinds of players, from academics on Twitter to journalists like me to all kinds of academic organizations, have encroached more on the think tank space than had been true before. There are just a lot of people who are not literally located in Washington DC, who are able to be in an ongoing, constant way part of the DC conversation. And so part of the value of think tanks was a localism to them. They are there. They have meetings. Like you can go to a lunch. And that’s just geography has become a little bit less important in politics. So that’s one.

My subjective impression, and I have not tracked this over time, but I think think tanks have ended up in a funding cul-de-sac where they get a lot of individual donations for individual projects that are not often the most valuable or important projects. And so if you look at the overall output of some of the major think tanks, you end up scratching your head a little bit about why it looks the way it does. And the answer is some rich people were willing to fund that particular constellation of outcomes.

And then I think that the same, as I sort of offered at the top of this, a bit of a critique of [inaudible 00:25:39], but also of myself for not thinking enough about the way things ought to be. I think that has afflicted think tanks are many of them as well. Before I came to Washington DC, or before I began blogging about politics, think tanks sounded like just places where the smartest people in the world went to think.

And in fact, they’re very bounded. They’re very much about what’s politically possible. They’re trying to husband their influence with whichever party they’re a little bit more aligned with. Polarization has made it harder for the think tanks to be influential. So their work is valuable. I mean, I’m not saying it isn’t and I in no way want to suggest that working within the constraints of the politically possible is somehow no longer a useful thing to do. It is necessary. It is done. It is important.

I just wrote a column on, I guess this wasn’t a think tank, but out of a kind of new school proposal for a guaranteed income. But I do think there’s a little bit more energy and this also reflects kind of social media engagement dynamics. And it also just reflects an opening of a political imagination over the past couple of years broadly.

There’s a little bit more of an interest in how the world should be, putting aside the question of what Congress can actually pass. And I that’s been a much harder place for think tanks to play in. And so some of the people who have been willing to move in that direction, who are often sort of more independent ideological entrepreneurs of smaller journals or whatever, have even a little bit of a think tank space of influence and Imagineering.

Matt Grossmann: And how about journalism? You obviously have been involved in creating the space of explanatory journalism. There’s a lot of complaints about the financial state, but in some ways it’s a golden age of the tanks industry as well. So where are journalists then using research and informing policy?

Ezra Klein: Well, I definitely think there are too many tanks. If you flip back a decade or two to the rise of blogging, blogging emerges in a time when there just isn’t that much political opinion, full-stop. This is a point I’ve made many, many times, but I’m 37. I’m not 1000. When I was growing up in Orange County, we got the LA Times, there was a nightly news. We didn’t really listen to conservative talk radio. My family was not a family that subscribed to political magazines. So basically the political opinion I routinely had access to was the op-ed page of the LA Times.

There was no political opinion, that is to say, to a first approximation. This was even before really the cable news space blew up in the way it later does. And now we’re just drowning in political opinion. And so I think it has made it harder to stand out and add value just for takes. I mean, there are people can do it because they’re that good, but there just aren’t that many of them. And so I still think there’s a tremendous role for well done reported explanatory work, which of course has an element of the take in it sometimes. Oftentimes even.

I think broadly speaking, the real problem in journalism is not at the national level. It’s the state and local level. I think that if what you’re looking for is national political commentary explanation, et cetera, there is, even right now with all the difficulties in the business models, there is probably more of it accessible than has ever been true ever before.

But if you are trying to understand what is happening in Sacramento, that’s not true. If you’re trying to understand what is happening in Minnesota politics, I’m not sure that’s true at all. If you’re trying to understand what is happening in San Francisco, the San Francisco Chronicle is weakened compared to what it once was, as are a lot of these papers. So I have real concerns about the state of journalism. The state of national level political journalism has issues, but it’s not at a point, in my view at least, of crisis. We all have our views on how our industry could be better and how frankly it could also be worse. But I think that state and local situation is pretty much a disaster.

Matt Grossmann: So we’ve just been through a year and a half of COVID to kind of be a case study of how all this came together. There were, of course, important social science regularities that were discovered during COVID, but there were plenty of complaints about moving too fast or making errors along the way as well. How did all of this come together in analyzing COVID? Did we produce useful information, and was it based on research?

Ezra Klein: Having just written a book where the entire point of it was that polarization and the broader ecosystem that drives it had locked people in political place and turned us into a polity that was scarily devoid of accountability or openness to change for new information. If you had told me what would have happened over… I mean, I released a book right before COVID. COVID interrupted the second half of the tour. If you had told me what would have happened over the next year, and then asked me where say President Trump’s approval rating would be, or where the American polity would be. I would have assumed much more change than we saw. I think I ran the numbers on this at one point. And if you look from the one year before Donald Trump took the stage at the RNC, that day. If I’m remembering this right, his approval rating went up one point and his disapproval went up three points.

That was it. The 2020 election looked pretty much like the 2016 election. I mean, there are differences, it was high turnout. It was pretty similar. I mean, the differences are important, I’m not taking them away, but it was pretty similar. So the level of lock-in, whatever you think of as the cause, whether you think it’s information ecosystems or just an unwillingness to reevaluate everything based on… Even pretty extraordinary that the level of lock-in was extraordinary. And I think it has to worry you. I mean, it just makes clear that even pretty extraordinary events, which have a tremendously significant role for confidence in them, are not going to be enough most of the time to lead to traumatic electoral changes.

Which creates a real accountability problem. Then add that in, my endless obsession with the way the Republican party is insulated from democracy by its geographic distribution and by gerrymandering. And you have a real, real issue. The idea that we were within a couple tens of thousands of votes of the election being thrown to the House of Representatives where Trump would have won. I mean, that should scare anybody.

Matt Grossmann: Well there was the view that this wasn’t a political science or social science question in the beginning, and the social scientists should stay out of it. But it certainly seems like a lot of the trends were right in the political science wheel house. The state government responses were very polarized, the public responses were very polarized. There was a lot of traditional leadership following. We’re still going through it with vaccines. I mean, does that suggest maybe there wasn’t enough political science in the conversation?

Ezra Klein: Maybe. Again, as somebody who tracked a lot of political scientists during this and talked to them during it, because I continued to cover politics during this period. I would say that a lot of people made the same mistake, which is the biggest mistake I talk about making in my book. Which is refusing to believe, or being unable to believe how much their models were going to hold. And so, again, there’s a difference between political science as some kind of abstract body of knowledge, and political scientists. There’s also, by the way, a difference between epidemiology and epidemiologists, public health and people who do public health. And maybe therefore I put this as a mistake that we make in journalism is letting the one stand in for the other. What a discipline says and what the people you’re talking to from a discipline say are different. [inaudible 00:33:50] when you’re facing something new. Political science did not have a tremendous number of pandemics that happened in modern polarized conditions to work off of.

And then the other thing is a lot of the data was tough. It is a case that Donald Trump’s polling looked much worse in key states throughout the crisis than what we saw on election day ended up proving out. And so a good political scientist who is working off the best data we have, which is, well, Trump looks like he’s going to get squashed in Wisconsin. So maybe he should have tried something different here, even though his overall approval ratings have been pretty stable. I would not have asked a political scientist to tell me the polls are wrong, that’s not in their expertise. But the polls were definitely wrong in Wisconsin. They were wrong in Pennsylvania. I mean, they were wrong in the same set of places. And so that’s a question. One of the very difficult questions for journalists, for political scientists, we all are relying on, at some point, someone else’s domain of data expertise to build a view of a complex world and figure out which of our ideas and which of our evidence-base is to apply to it.

And when it turns out that the people you are relying on have mid-modeled the situation, epidemiologists right at the beginning of COVID, pollsters through much of the election in the Midwest. I don’t know what I would ask of people. We’re trying to give you the best we can, but the best we can is within a world that is uncertain. And so I did a bunch of stories throughout the period of COVID talking to political scientists and pollsters about Donald Trump’s eerily stable approval ratings, right? The things I’m giving you are not things I thought of after the fact. This is my real-time journalism happening here. I mean, that’s all fine and well and good, but you only have the information you have at the time.

And the question of, what do you do if the polls suggest that Donald Trump collapse in key states, and then that doesn’t manifest? What do you think happened there, and how do you update that into your models? Should you never believe polls again? Should you say, well, nope, we have fundamentals models? I don’t know what the answer to that is. One other problem across political science during this period is, how do you look at economic fundamentals models during an economy has just been put into lockdown? Should you build off of that the same kind of recessionary dynamics we’ve seen in the past? Or should you say it’s something different? Different, as you know perfectly well, different forecasting models gave dramatically different answers depending on the moment they were taken. And so, I mean, again, what to believe from political science was an open question during this period.

Matt Grossmann: Yeah. I guess I didn’t see it that way in real-time, I thought political scientists were mostly clear that it was going to be a very close election and that the dynamics of the year made it more uncertain than in prior years. But maybe that’s not how it sounded, or maybe that just wasn’t an exciting story to relay. I think maybe that [crosstalk 00:37:02] second dynamic.

Ezra Klein: People constantly said it was going to be a close election. But there was a lot of, you have to both… What I heard happening constantly, from people on all sides of it, was, well the polls suggest a blowout right now, but you never know. And so that got reported a lot. But one of the things that I think is slightly different between the way I see this and you see it, is that I can’t call up political science and ask it a question. I have to call up a political scientist. And I have a pretty good idea of the literature, but also sometimes literature conflicts. Did I want the election forecasting literature here, or did I want the polarization literature?

This is not a critique of political science, I just don’t think political science had any more it could do during the past year than it could do. And I think it did what it could do fairly well. I mean, maybe this is not for this podcast exactly, I just don’t think it was an interesting time for a referendum on political science. I think the pandemic demanded other kinds of expertise to be understood.

Matt Grossmann: How about on racial prejudice and policing? We obviously went through last summer a moment when everybody was more interested and political science played some role in the public conversation. But how did social science generally do in informing that conversation and political science specifically?

Ezra Klein: That’s a really good question. I don’t know that I have a good answer to it. I sort of know about the policing conversation, where I don’t think political science has been as central in that debate. Probably because that is very much a debate about things that aren’t. Right? What would happen if we defunded the police? What would happen if we abolish the police? I mean, there is some information in political science about what would happen if you propose these things. And I think, more or less, that has panned out. But I don’t think that’s been a place where political science has the answers exactly. In terms of the broader debate though, look, I mean, this is a big part of my book too. I think something that has been present in political science, although not always well emphasized, is as Hakeem Jefferson likes to say, that race is the defining cleavage of American politics.

And I think that has come to be a much more broadly accepted view than was true five or six or eight years ago. But even if you look at political science that was done before this moment, I mean, I always find it very potent as a way of thinking about American political polarization. That for so long in the DW nominate system, you needed a second dimension of civil rights in order to explain the cleavages between the parties. I mean, that is something that, I don’t think that set of Poole and Rosenthal are thought of as crazy woke political scientists from a decade or two decades ago when they were starting to do this work, or whenever it begins. But I mean, right in there they show something that has become very, very, very broadly accepted. Which is that you cannot understand the nature of the coalitions and the nature of political polarization without understanding race as a central cleavage, certainly in mid century politics.

And that’s a very, very important part of my book, and my ultimate work. And I think that’s continued to be the case. I think that Lilliana Mason’s work is excellent here, and the work she’s now doing with Nathan Kalmoe has been really interesting. And there’s a lot of people doing fantastic stuff. Jennifer Richardson, who I think is technically a psychologist, but is a social scientist working on these issues. I think her work on demographic threat is fantastic. So in terms of that space, I think the social scientists, the empirical social scientists broadly have been terrific. The one place where I think there’s something very unsettled in political science is how to understand racial resentment measures. And there has been certainly an amount of internal debate in the profession about this, that I get the sense people don’t really want to have this fight. But is something that is a big asterisk in my head now, because I’ve both seen studies that seem to validate that the racial resentment measures are picking up on something important.

Ezra Klein: And also studies that questioned whether or not what it’s picking up on is as simple as racial resentment. And yet it’s a very standard way of testing this out. And so that’s a place where I think that there is a methodological debate in the discipline, that as I understand it, is reasonably unsettled. But whether it’s for reasons of concern about the politics, or whatever, strikes me as important. And I wish I felt that the best methodologists had come up with some kind of answer for me so I could be confident in what I was reading when I see things using those measures. [crosstalk 00:41:42] you think?

Matt Grossmann: I think the racial resentment measure incorporates both conservative ideological precepts and racial animus, and it was designed to do that. And it accurately points out that those two things are quite fused in modern American politics. But that means that you’re always going to have a debate about the extent to which it was formed by one or the other. And I think that’s going to be hard to solve historically, and incredibly hard to differentiate now because they’re so fused.

Ezra Klein: So then what do you think about studies or other things that use racial resentment as a stand-in for actual racial feelings, then use it to track how people are moving across a political system? Because if what you’re saying is true, then you can have somebody who they’re just conservative in their views. And you might just say, conservatism is just now a measure of racial resentment. But obviously a lot of conservatives don’t see it that way. And then it can blow out into other areas where maybe something, maybe a view. And I think this is a reasonable interpretation to take. Maybe use it, have their start in American politics as a way of excluding black people from public goods. Now they’re just conservative views and they’re weighing, as [inaudible 00:42:56] you would put it, of draining the pool for everyone.

But so then when you see a paper, because they come across your desk all the time, that starts with racial resentment measures to understand the way racial resentment is changing political decisions. Do you then say, okay, that is how racial resentment is changing political decisions? Would you say, that is how some complex mix of being a conservative and having resentful views about other race? What is your interpretation now?

Matt Grossmann: Well the latter. I mean, I think it’s unsettled again what the prime mover is. I think there’s also a lot of evidence that conservative views or Republican identification comes first, and then people adopt racial resentment views. I don’t think you can know for sure that one comes before the other, but it’s certainly true that there fused in contemporary American politics. So if you’re trying to pull this apart today it’s going to be hard. I think we can make progress historically in some cases of figuring out with panel surveys what came first, and then looking at the role of political rhetoric, infusing them. What mattered. I guess I’m more comfortable with the views are now fused, than maybe you are.

Ezra Klein: Fair enough.

Matt Grossmann: So climate, another pet issue of yours, and ours. It strikes me that this is an area where political science does the thing that you may not have liked at the beginning. We say what the problem is. We say it’s going to be hard to unpolarize the issue. If you want advice on convincing people, we tell you to use conservative arguments, use Republican spokespeople, things that might not be very viable. On the other hand, maybe that is the best we can do with the contemporary nature of the issue.

Ezra Klein: I think you may be over reading me [inaudible 00:44:54] I made at the top of this. I think this is a place where political science is doing a different thing I talked about, which is not being willing fully to believe-

Which is not being willing fully to believe what its own literature base tells us about the problem, which is to say it is politically solvable. For any of the set of goals that I think most of the people we are thinking of would hold as the level of de-carbonation need by the date we need it, the answer the political science gives you is it’s not fucking going to happen. It’s not going to come close to happening, absent some kind of technological miracle we’re not anticipating.

Then the question becomes, what then? If your answer to what then is try to have some republican spokesman, no that’s ridiculous. It’s like fine, give it a shot, but that’s not going to do it. A lot of people have tried that. I am struggling with this question right now for a column I’m thinking of writing, is the answer you got to really start just talking about [inaudible 00:45:50] because we are going to have two-ish or more Celsius degrees of warming, and the horrors that are going to follow from that is the answer like the new [inaudible 00:46:01] book argues that you’re going to begin to need extra legislative pressure blowing up pipelines, and this should be understood as a kind of violence that the president is committing on the future, and the future needs to begin striking back.

I don’t like even talking about that, but I think that the one thing that nobody seems to want to do in this conversation is just say almost certainly the legislative pathway is going to fail. And then what? I don’t like saying that either, but this is a place where I think there’s a “kind of” activist merger among political scientists who don’t want to sound hopeless to people. This is actually one place where I think the political scientists speak differently than you do, and maybe because they speak to me as somebody speaking to people who are broader, but political scientists never want to be hopeless. They believe in things too.

They want to believe we’re going to have a better future too. When you talk to them, you don’t kind of get sometimes the “hopeless take”, which is maybe the right one, which is it’s not going to happen. We might pass some bills, maybe some version of The American Jobs Act passed, but it’s not enough and it’s not going to pass in full form, and it’s just not going to happen quickly enough. So then what? It’s an interesting question for political science, also of course for society, like what do you do if politics is going to fail? What does political science do if politics is going to fail?

You might say it’s not politics failing, it’s politics succeeding because the republican party doesn’t want to do anything here, and outcomes are supposed to be a compromise between the pluralistic factions in our democracy. We’re getting in some way the sum total of what we want, but if you think that failing or succeeding should be based on something a little bit more objective like preserving some semblance of a climate in which human civilization has prospered, then it’s going to fail.

That is my take from reading a lot of political science. Given that take, the number of people who seem willing in the profession to answer that question, to even muse about that question, is very low in my experience, which I think is interesting. I don’t blame them. Again, I know why people don’t want to sound hopeless, and I don’t want to sound hopeless, but I don’t exactly think of it as hopeless. I do think that it is a case within the history of politics that sometimes the legislative pathway fails and people begin trying other things. The Civil Rights movement is an example of this. War, is of course, often an example of this, both civil and interstate.

There are all kinds of things one might begin to talk about, but I don’t find a lot of enthusiasm for talking about them.

Matt Grossmann: A lot of the social science that we’re talking about is uncertain, and we of course have gone through a replication crisis in psychology, a lot of complaints about social science. How do we and how do you go about explaining research evidence-based findings that nonetheless are uncertain and could change? How big of a threat has it been that some canonical studies haven’t held up, and the esteem of social science has suffered some as a result?

Ezra Klein:

You can tell me, because you know better, if I’m wrong on this. My understanding is this is much more of a problem within the broad psychological disciplines within political science. I get this question a lot, but political science has made me somewhat more skeptical about political psychology studies, including experimental ones, I will say that. That said, and we can talk about political psychology if you want, but a lot of what I use in political science is people running analysis on both totals.

Unless somebody made a mathematics mistake, which is of course possible, that’s not going to change exactly. The situation of American politics can change, and hopefully over time will, but I’m just trying to tell you what’s going on currently, oftentimes. What I think of using, say Frances Lee’s work on the president taking a position increases the chance of the party line vote, I’m not concerned that’s not going to replicate. A lot of the political science I use, backed up on my political reporting, I’m using political science to illuminate trends and dynamics I am seeing occurring right in front of me.

I have to think of that as a secondary form of validation, at least within the standards of journalism, if not the standards of literal peer review. It doesn’t worry me that much. I do hear sometimes about studies where everyone will be like, “Eh.” I think the political psychology literature is clearly getting at something very important, both tools that too crude for me to think they’re really reliable. I don’t really know what to do about that. I think there’s clear psychological sorting. I think all the stuff about openness to experience and some of these disgust responses and so on and so forth, it’s getting at something.

But with reasonably limited experimental manipulations, I’ve seen things that reverse results. Some of the things are cutting really, really tightly or narrowly, or you’re using measures that are… It’s interesting what they correlate, like the Authoritarian Index for Parenting. But what is it correlating to exactly? That is a place of political science or sometimes psychological work that I find both it’s clear that there’s a psychological substrate to our politics, and it’s clear that it’s telling us something, but at a narrow level I’m uncomfortable relying on it too heavily because I think our tools are simply too crude.

It’s just really correct, I think, and probably more important than we give it credit for in any one study I’m a little bit nervous about. For most in political science, for better or worse, y’all are not doing work that it’s so experimentally validated that it’s even open to replication. When I read a [inaudible 00:51:54] book about comparative politics, I’m not worried it’s exactly failed to replicate. We’re trying to draw some illumination out of history here. I sure hope the Nazis failed to replicate. I’m not going to call them up and yell at them and nobody else is taking powers of genocidal maniac despite the fact that he wrote that book.

Matt Grossmann: What’s next for you using social science to inform public commentary? Anything coming after why we’re polarized? Anything that political scientists should be doing to better inform public commentary?

Ezra Klein: I don’t know about that. For me, for this big world I try to covering in, and I got plenty on my plate right now, I’m thinking a lot about questions of democracy, and I think a lot about questions of blue state governance right now, both democracy at the high level but also the ways and reasons the blue states often fail to govern as well as you would think at the lower level by putting the history and legacy in some of the progressive [inaudible 00:52:56] forms and actually creating a bit of a divided soul on the left in terms of “We want to do a lot with government, but you’ve created a huge amount of veto points in government because you’re worried about special interests using it as a tool of their own power.”

In terms of political science, I continuously think, and you are somebody who does a tremendous job trying to battle this back, I continuously think it is so much harder than it should be to follow the state of the research, like why I cannot go somewhere that is run by APSA and see the set of papers and working papers, and so on that are coming out that week. It will just baffle me forever. And categorize them, and search them, and so on.

There’s a Matt Grossmann Twitter feed, and then there’s individual people yelling about stuff on Twitter, or more valuably, giving their research on Twitter. But there’s just not nearly as much of a digest or repository as I would think there would be. Continuously, that is what I think should exist, a better centralized resource to keep me up-to-date on what’s going on and then to search it. Some people are great at things at this. Women also know stuff. The site that brings to view lets you search by specialty the work of one political scientist. It’s fantastic.

Why there’s not one of those for political science too, where I could do some more searching and I could see who might be able to talk about my sub-specialty that I’m looking for that week, or again just follow the research that people think is worth seeing that week, aside from again your Twitter feed, I don’t know. That seems to be something the profession could put a couple hundred thousand bucks into a year and the problem would be solved.

Matt Grossmann: Maybe we’ll take a page from your podcast and have you recommend three political science books that have influenced you.

Ezra Klein: Sure. I always recommend, I guess I’ll do the couplet here, Francis Lee’s two “Beyond Ideology” and then “Insecure Majorities”, is it? I think have both influenced me tremendously. Lilliana Mason’s book, “Uncivil Agreement” is terrific and really changed how I thought about identity. Then a third that I just think is fantastic is Robert Mickey’s “Paths out of Dixie” to talk about some of the work political science has done on race. I think Micky’s a political scientist. I could be wrong.

Matt Grossmann: Yeah.

Ezra Klein: Yeah, he is. I just loved that book. It was fantastic. That also gave me a really different understanding of what was going on mid-century American politics. I highly recommend that one.

Matt Grossmann: There’s a lot more to learn. “The Science of Politics” is available bi-weekly from the Niskanen Center, and part of the Democracy Group Network. I am your host, Matt Grossmann. If you liked this discussion, you should check out my conversation with Brenda [Nyan 00:55:37]. Our episode on “When and Where Climate Policy Can Succeed”, and our panel on “Whether Racial Attitudes Elected Trump”. A special thanks to my engineer for 100 episodes, Alejandro Gillespie, and the lead producer at Niskanen, Kristie Eshelman. It’s been a great run so far. [inaudible 00:55:52] for joining. Please check out “Why We’re Polarized” and then listen in next time.