Politicians are launching outlandish negative attacks and Americans have developed more negative views of the other party. But how connected are polarizing politicians and a polarized electorate? Mia Costa finds that political elites have more polarized views of the other side than the public but they still benefit electorally and legislatively from avoiding negative partisan attacks. Divisive rhetoric still breeds viral tweets, cable news appearances, and donations, but Americans mostly don’t like it or reward it. The polarizers just get more attention.
Guest: Mia Costa, Dartmouth
Study: How Politicians Polarize
Transcript
Matt Grossmann: If we don’t like polarizing politicians, why do we get them, this week on The Science of Politics. For the Niskanen Center, I’m Matt Grossmann. There seem to be more politicians who launch outlandish negative attacks, rather than working on legislation. But at the same time, Americans have developed more negative views of the other party. How connected are polarizing politicians and a polarized electorate? This week, I talked to Mia Costa of Dartmouth about her new Chicago book, How Politicians Polarize. She finds that political elites have more polarized views of the other side, but they still benefit electorally and legislatively from avoiding negative attacks.
Negative partisan rhetoric does breed viral tweets, cable news appearances and donations, but Americans mostly don’t like them or reward them. So, we aren’t in an inevitable cycle of electoral demands and polarizing politicians, the polarizers just get more attention. That still leaves a lot of open questions, but I think you’ll enjoy our conversation. Tell us about How Politicians Polarize. What did you find?
Mia Costa: Yeah, lots of things. I guess I’ll start with something that is really important to me, because I can’t write a book about how politicians talk titled Politicians Polarize without being transparent that they don’t actually talk. They don’t actually use as much polarizing language as maybe we think. So, one of the first findings that I want to emphasize is that legislators talk about policy a lot. I measure how much they talk about substantive issues compared to talking about in group, out group partisanship. And above and beyond, they’re talking about substantive issues. They’re engaging directly with constituents on legislation, on policy, on their work in Washington, which I guess that’s a good thing. That’s a kind of silver lining. And the study of polarization and elite rhetoric is that actually they’re talking about policy. But unfortunately, the partisan attacks they do make are associated with getting lots of attention and it gets more coverage online. It gets more coverage in the news media. It’s associated with more fundraising. So, it’s all we hear about.
So, it seems like there’s this context of polarizing language from elites and partisan, fearmongering, and partisan animosity, because that is what we hear about. Unless we’re engaging directly with our representative, we’re not really hearing in general about how on politicians’ platforms to talk with constituents. I look at specifically congressional newsletters and tweets. And we don’t really hear about how they’re talking about the issues. We do hear a lot about the partisan attacks. And so, there are incentives for it, even though I find that partisan attacks don’t win votes. People don’t like it. They don’t report that they like it, even when it’s coming from their own representative. And politicians don’t think that voters like it.
I have these elite surveys and experiments where I find over and over again that people who have run for office or that currently hold office, they think that insulting the other party or even just focusing on the other party, and focusing on division, and partisan conflict is just not going to be an electorally beneficial rhetorical strategy, especially compared to focusing on the issues. So, legislators do talk about policy the most. They do also attack the other side, but they don’t do it because it garners support from voters. I argue that they do it, because it is associated with other forms of influence, and power, and attention. And so, that’s not great.
Matt Grossmann: So, as you say, you find that legislators emphasize a policy for the most part. And that maybe it really is just these cranks that get all the attention, who are focused on denigrating the other side, but it doesn’t even work as a legislative strategy. It’s sort of the ineffective folks that are doing it. So, does that mean that we’ve just overblown this whole phenomenon and that we just have a few polarizing politicians?
Mia Costa: I mean, a little. I think it is a bit overblown. I mean, it’s hard to say, because we hear about the… You said maybe it’s just a few cranks. And it’s like, yeah, but we hear about them a lot though, and they have a lot of power. So, it’s overblown in the sense that it’s not as maybe widespread as it seems like. So, just to provide context. On average, any given legislator will talk about policy or important issues in their newsletters 97% of the time. And on average, they’ll mention substantive issues almost 800 times in any congressional session in their newsletters specifically. So, this is not just say in floor speeches when they make news appearances, which that’s even more. So, like 800 times, it’s like 770 times in any given congressional session. And they’ll only mention the other party in the newsletters 17 times on average.
That was the estimate I found, which is that’s a laughable difference. It’s really not that much. On Twitter, it’s not as stark, because only 35% of tweets I find talk about policy from Congress members, but that’s because there’s a lot of other stuff going on Twitter. It’s not because they’re just talking about the other party. It’s not because they’re just making partisan. They’re slinging partisan attacks, because again, in a congressional session, any given legislator on average talks about policy on Twitter, well, 600 times, 570 times. These numbers have been drilled into my mind. And they’ll talk about the out party 36 times in any congressional session. It’s not that much. This is literally a handful of times that we hear about. And the important thing though is that those instances of partisan attacks do get picked up.
I find that they do get picked up by the partisan news media, television news media, Fox News, MSNBC, CNN. And I do find that those instances on Twitter, they do go viral. They’re more likely to go viral. And the legislators that do it more than average are the ones that have more power. So, they are more senior. They’re in the leadership in the Chamber. So, it’s not as bad as it seems it is, but the fact that it seems it is bad is on its own a problem. It kind of shows the context of polarization and it does create a sort of feedback loop. It’s kind of like if a tree falls and no one is around.
This is like if politicians are talking about substantive issues and policy, but no one is paying attention to it, then does it make a sound? Legislators are doing their job when they’re crafting their communications to constituents. They’re talking about substantive issues, but that’s not really what gets covered. It’s not what feels true. And that on its own is important, too.
Matt Grossmann: But you also seem to be letting voters off the hook here. We’re not the cause of these politicians’ strategies. It’s not what we want. It’s not what we reward. And it’s folks that are not really facing much electoral competition who are engaged in this. So, it doesn’t really sound like this is coming from voters or even the most activist parts of the voting constituencies; is that right?
Mia Costa: Yeah, that’s right. And I’m thinking about this question, am I letting voters off the hook? And yeah, it is one way to look at it, like, wait, we don’t want this. Voters don’t want representatives who are just bashing the other side and inflaming partisan negativity. So, we don’t want this. We’re merely pawns. We’re powerless. But of course, it is letting them off the hook a little bit. It is letting people off the hook, because representatives are put there by people who vote them into office. But it doesn’t mean that everything representatives do is actively desired, like the two-party system and on competitive districts do mean that there’s a lot that politicians can do that voters don’t love. But it won’t cost them an election.
Non-competitive districts reduce electoral accountability. People in safe seats mean that they can cater to their base without broader appeal. And so, voters enable it, but the lack of competition insulates this behavior. That lets some people off the hook, because, of course, there’s some people who do get riled up from it. There are people who really are actively engaged with this, like the partisan fighting and just are really affectively polarized in the mass public. And that’s a real phenomenon. But I argue that that’s different than in an ideal world, like what would we want our representatives to do, and to focus on and to talk about? I think that’s different than what people tolerate and enable in a two-party system with uncompetitive elections. And so, you’re going to vote for your person, even if they’re exhibiting behavior that you don’t necessarily like.
You don’t like that they’re maybe sitting on Twitter and bashing the Democrats or Republicans, but that doesn’t mean you’re going to vote for the other guy on the other party. And so, I think there’s a little bit of a difference there between the ideal and the actual. And I do think it’s important to know, well, what would people ideal be, and that voters don’t reward this behavior from politicians. They don’t reward polarizing rhetoric at the voting booth. But on Twitter, are they going to hit retweet, especially if they themselves are riled up and have these affective feelings about the other party? Yeah. And so, it is a way of enabling the behavior and creating a reward system for it.
But it’s different than what we would want our representatives to actually be focusing on in office. I find that people just don’t respond actually well to it. So, it’s hard, because if you’re in a safe seat though, you can get away with a lot, especially if you’re a senior party leader or something and you have all this attention. You have the ear of donors. And so, using this sort of language gets you attention. That’s really powerful, because attention is what gets your name out there, gets the attention of the news media, gets the attention of donors, and so gets the attention from colleagues, from party elites, too. And so, it’s powerful language, even if voters wouldn’t prefer it, if they could design it another way.
Matt Grossmann: So, you also find that these political elites hold more negative opinions of the other party than do voters. And that might be a source of the issue, is that the people with higher animosity toward the other parties are the ones that are elected. But it also got me thinking about whether this is just a sincere view and whether it makes sense that if you’re very involved in politics, you’re going to recognize that partisanship is driving a lot of behavior. You disagree with the other side on a lot of different things. Is there another path where we have less people who are elected that don’t have this animosity? Or is this just kind of a natural outgrowth of the two parties really do disagree on most things, so the people who know that are not going to have a very positive opinion of the other side?
Mia Costa: I mean, I agree with you. And that is what I find. So, just to spell it out, I measure partisan affect among candidates and current officeholders. It’s mostly the local and state level I have data for. And compared to these mass level surveys. And find that maybe unsurprisingly, elites are actually more negative towards the other party than the public is. And at the time that I was running these analyses, we did not have good measures of elite affective polarization. Now there is some studies out there. There are [inaudible 00:14:49] looking at that. And I mean, I think what you said is exactly right, is that it’s not a far logical leap to think that the people who are in office are also more politically knowledgeable, and sophisticated, and partisan, and ideological. And so, therefore, they’re also going to be the ones that maybe have stronger feelings.
So, yeah, I think you’re right. I mean, of course, one of the challenges in studying as a representation scholar, in studying the traits of people who are in political power, is we don’t know whether it’s selection or socialization. Are the people who run for office already higher in partisan animosity, like we have just said? Or does being in office make them that way? People who run for office are driven to run because they have strong views. That would be the selection. They select into it. But also, like I had just mentioned is that there are other incentives and allowances for partisan attacks. And there is a context of polarization. And so there is also the possibility that the exposure to party conflict, the learned understanding of strategic incentives of institutional norms of pressure from donors and party elites, it’s possible that that political environment might also shape the attitudes and behaviors of political elites. So the traits observed after entering office, specifically levels of partisan animosity, may not reflect pre-entry levels perfectly. That’s hard to disentangle. I think though that it is a selection problem and so yeah, it’s hard to imagine any other way.
It’s hard to imagine that we’re going to incentivize a bunch of moderates who aren’t prone to a little bit more partisan conflict, or also, the show horse nature of politics, a lot of politicians are a little bit… They’re at least a little bit with the performance side of politics and campaigning. And so there is that selection there and so it’s hard to imagine that other types of people are really going to be recruited, especially to run for high office like Congress. But there are incentives, like Andy Hall has his book, Who Wants to Run?, about ideological extremity of candidate entry. And it’s all part of the same thing is that there’s an incentive structure for certain types of people to be in this position, and that’s going to be associated with qualities that enable the partisan rhetoric, the forceful othering. So that’s a hard nut to crack, definitely.
Matt Grossmann: But does this make you rethink any of how we talk about affective polarization in the public? Because it seems to me that, I guess, maybe this is an extreme version of the story we usually tell, but a lot of the story is, well, this is something different, it’s not really about disagreement, it’s about feelings toward the other people. And it is animated by hearing a bunch of things, but not necessarily connecting them to policy views. That doesn’t really make any sense if applied to politicians. That’s not true of politicians that they don’t know the policy stakes or aren’t engaged in actual ideological conflict with the other side. So to me, it makes me think, “Well, these people who have negative opinion of the other side is yes, they may not disagree on policy as much as politicians, but maybe the high affective polarization in the public that are increasing is just a product of them learning the similar kinds of lessons that the politicians learn, that the parties really do disagree on a lot.
Mia Costa: Right. And there’s a little bit of that debate going on right now in the literature or has been of, is affective polarization, at least in the mass level, underlined and driven by ideological differences, genuine ideological differences in ideological polarization or purely affective reasoning. Yeah, I mean, after working on this, I am more squarely in the ideology is driving this because when I first started working on this, this really started with my article called Ideology, Not Affect in the AJPS. And I only started working on that because I was reading all of that affective polarization literature and hearing a lot of the conclusions from that work that were like, “Oh, people really dislike the other side.” Maybe some of it is driven by their positions on the issues and thinking that the parties are different and have these genuine difference. And so therefore people are really just more motivated by this out party loathing than they are by the substantive issues.
Because there’s a little bit of this idea that even if it’s undergirded by views on substantive issues, it then morphs into its own thing where it’s really just about the affective feelings of the other side. And I came to that work again thinking of how this might influence perceptions of representation, wanting to study like, “Oh, does this mean…” If all this scholarship is suggesting that people are not really interested in the substantive issues, maybe that’s how it starts, but that it’s really about now these affective feelings and then that’s what should also be driving our views about representative behavior.
And in that article, that’s not what I found, above and beyond in my surveys and my experiments was people react most strongly to ideological congruence, to issue congruence, to issue alignment, and we’re not responding well to partisan rhetoric. And so I do think that that complicates the affective polarization story a bit because if this was solely just based on negative feelings and there wasn’t some sort of connection to ideology or substance of issues or genuine disagreements with the other party, then we would probably see that people were rewarding hostility.
But again, rewarding hostility from politicians is very different than just expressing or exhibiting hostility towards out partisans. So I think it complicates the affective polarization story, but it’s not inconsistent with it necessarily because people may feel a lot of animus towards Republicans or Democrats as voters or as politicians, but not, again, want representatives to capitalize on that. That might not be what they want to hear about. It reminds me of the book by Yanna Krupnikov and John Barry Ryan, The Other Divide. And their argument is the one divide that we hear about is between Democrats and Republicans, but the other divide is from people who are partisans and then everyone else who doesn’t want to hear about partisanship. And so I think there’s a lot of that.
I think my work suggests that the negative rhetoric reflects the elite incentives more than public demand and so that is a piece of affective polarization that sort of complicates whether or not it’s driven by genuine ideological differences because it does support a shift towards polarization rooted in policy sorting and partisan identity, but not just animus on its own. The public may be polarized but not necessarily cheering on negative representation. And the animus may exist on its own for some people, but like you said, we can’t reasonably make the argument that elites are ignorant of partisan or genuine ideological disagreements or differences and so if they’re also on their own expressing a lot of partisan hostility, that’s evidence in favor of like, well, that’s because they have genuine disagreements with the other side. And so I think that’s a big part of it.
Matt Grossmann: So as you mentioned, we have this idea long-running that there are show horses and work horses in Congress and obviously it predates a lot of the kinds of things you’re talking about, tweets and cable news bookings, but it might just be seen as just this is the contemporary form. There have always been incentives for some people to engage more in grandstanding than policy work. Some people have always taken that route. Now we have a new way to do it with cable news and social media, but it’s really the same old story. How much do you think that’s true versus there really is new incentives, new behavior?
Mia Costa: Yeah, no, I think it is the modern show horse, is that it is more performance driven, more… The modern show horse is going to be more performance driven, more media savvy with cable news, with online content, especially since I find a really strong relationship between negative rhetoric and visibility. And I find that, but that is a consistent finding in the literature, media scholars have found that for years, and so that’s really key for building a personal brand. And building a personal brand has always been important, of course, but it just looks different now. And so the viral content and outrage drive attention more than policy work more than the work horses. And so I don’t have data on this, but that can translate to influence elsewhere, the show horse. I have data on whether that translates to influence with fundraising and attention and party leadership, but there’s also probably other ways in which it translates to influence both in and outside the chamber.
So yeah, I mean, like you said, we have more modern and updated mediums for the visibility of the work of the show horses, like Twitter has a very different incentive structure and a different format and different audience than even the congressional electronic newsletters and then being on cable news versus being on network news, broadcast. And so we need to pay attention to the fact that even though there’s a lot of speech that doesn’t sound like it’s based in substantive representation, it doesn’t mean that it’s all partisan hate either. It’s just there is this idea that that’s what politics looks like now. We hear more about those show horses, this goes back to that initial finding that I was talking about. And it’s just like, I think it’s what you said, that this is kind of an updated version of the traditional show horse style. We just have new mediums for it, and so therefore it’s easier maybe. So it seems more common. It seems like a stronger force. It’s always been there, but because of the different formats, it’s going to have a new type of power to it.
Matt Grossmann: On the other hand, some listeners may be saying, hasn’t everything changed under Trump? He certainly came to power with a different social media style. There’s some evidence that some members of Congress preceded him and built on him and their own social media strategies. The second Trump administration seems to have brought in contrast to the first more anger from the Democratic base that there isn’t more direct confrontation of the things that Trump is doing. So are we in the process of changing where there might be both people pursuing a Trump led strategy or a Trump similar strategy in an opposition that wants more confrontation?
Mia Costa: That’s a good question. I mean, the Trump question is when I’ve gotten before, because I really struggled with how to, not struggled, but I really thought a lot about how to handle Trump in this book. It feels difficult to write a book about politicians talking smack about the other party and using insults on social media and not talk about Trump. I don’t really talk about Trump that much in the book, so I had to think about this a lot. Trump accelerated and normalized an existing trend, like the Tea party figures, Ted Cruz, Jim Jordan, whatever, we’re adopters of this style, a combative style really focused on othering. I think what changed under Trump is the scale and frequency, and especially within the Republican Party, the party wide uptake, because as we’re seeing now, Trump is such a party leader, and so Trump mainstreamed the personal attacks and the performative outrage is really central to the GOP brand.
But to be clear, I don’t find, when I’m measuring these forms of rhetoric on social media and in the newsletters, I don’t find an increase from the years before Trump and now or in general, I don’t find an increase. It goes up and down. It has to do with my ideas. When parties lose elections, when they lose the White House, then in the following congressional session, they talk a lot more about the other party, but then especially when they lose the majority in Congress, which makes sense. This follows other work on majorities in Congress is when you’re in the minority, you have an incentive to attack the other side because you don’t really have much power in the chamber. And so that’s what I find, it kind of goes up and down. And I don’t find that there’s a clear, steady increase in talking about the parties as opposed to policy over time. I don’t have data that goes back so far, but enough that we would maybe see that.
But it’s really hard because like you said, right now, Trump’s second term, there’s a lot that’s different with Trump’s second term. And so it’s really hard to say because the opposition, the forms of opposition that’s happening right now is a little bit of a different type of threat to the functioning of government. And so, there is a lot more, so it might look different. It might look different in terms of what oppositional rhetoric might look like, but I think in general it’s hard to say whether any one person like Trump totally created this playbook or just accelerated an existing trend. I think we could see that with better overtime data. But the data that I have, it’s not clear that there’s just an overall increase. It’s just a sort of, I think it’s what we’re seeing now is it’s becoming more normalized, and so therefore it becomes more tolerated. And so, it does, again, we hear about it more. The scale is different, the frequency is different, but it was an existing trend.
Matt Grossmann: And does what’s happening now with Democrats fit the pattern or make you, I guess, interpret the pattern differently? As you said, you get more negative rhetoric about the other party when you lose. On the other hand, I think Democrats right now would just say, “Well, we’re responding to what’s happening. There is a lot of action that Republicans are taking or threatening to take. And so, we’re going to talk more about them.” But I don’t know if we’ve gotten Democrats who would say they haven’t gotten more affectively polarized. They’re just reacting to what is actually taking place, actions by the other party.
Mia Costa: Right. And I can speculate. I don’t have data on whether now Democrats are following the same pattern as maybe I’ve seen in the other years that I have looked at, but I would assume so, that now the Democrats are focusing more on what the Republicans are doing and talking about the Republicans, talking about the Republicans as opposed to maybe talking about their own side or focusing more on the issues. But whether that is, because that just follows the trend that I’ve found before, which is that that that is usually what happens after losing an election or losing power in some form, or it’s because again, right now we are a little bit in an unprecedented era of the threats to the normal functioning of government and the Constitution and to democracy. And in a very party-wide way, the Republicans have been unified in some of that. And so, it feels a little bit easier therefore to have a party-wide attack against “Republicans” as opposed to just Trump.
I would love to measure in Democrats rhetoric, are they focusing more on Trump versus Republicans as a group? The concept and theory that I measure in the book is that that group-based partisan rhetoric, so specifically are Democrats talking about the “Republican Party” and Republicans and vice versa for Republicans. But I do think that right now it would be really interesting to see whether a focus on Trump is different than just talking about the GOP as a whole, especially because the ways in which Trump’s second term is throwing a wrench into everything. And so, like you said, Democrats would say, “Well, we’re just responding to some of these unprecedented actions with the White House.” And so, are they focusing on Republicans though, or is it really just Trump? I’m not sure.
And I think it’s hard to say again because so much that we are hearing about now like in the news where the first half of this in 2025 is like Trump and Elon Musk. And so, those are very specific people as opposed to the Republican Party as a whole. And so, I’d be super curious to measure that. I don’t know the answer, but I would guess that the incentives remain as what I’ve previously found, which is when you’re in the minority or you’ve lost a power or you lose the presidency, then you have the incentive to focus on the other side because otherwise what else are you going to do? So I think that is what we’re seeing.
Matt Grossmann: So a lot of what stimulated this literature was the rise in negative partisanship or affective polarization in the public. And you too begin with the politicians polarized in voting in Congress. The public voted polarized with views of the other party. But it seems like we still don’t really have a story of the connections between those two things. The public is not causing the elite polarization in your story. The politicians are affectively polarized more so, but not necessarily getting more so or doing rhetorical engaging in rhetoric that would increase affective polarization in the public. So maybe these are just two separate phenomenon. People have different views of what’s going on in the opposite party, maybe because learning more about the differences between the parties or some other reason. Elites are changing behavior in response to incentives from changes in the media environment to other things, but not necessarily that connected to what’s happening in the public.
Mia Costa: So hard to know what drives what. And especially with this question, I mean it goes back to the, is it ideology? Is it about the conversation we had about, well, what is sort of driving affective polarization? Is it genuine ideological differences or not? In the same way, it’s like this is a question about is it on the elite side causing the affective polarization on the mass side or vice versa? It’s probably both. My evidence shows a mismatch that politicians use more hostile rhetoric than the public seems to want. That doesn’t mean that elites are causing public affective polarization. I hint at that which I think is what gets at your question of just because the public doesn’t want that language, so therefore it sounds like I’m maybe then suggesting, well, if the public doesn’t want it, then maybe it’s starting with elites. And so, that would suggest that polarization on the elite side is driving affective polarization on the mass side.
But it could be these are two separate things happening. I don’t know if it would be wholly separate because elites are people too. This is something I say in the book that sometimes people don’t like. I think a lot of representation scholars really love to only think of elite behavior in terms of the strategic incentives and we can only study them as rational actors because we have this observable behavior and we can’t otherwise know, but we can know because they’re people. And so, my thinking is that a lot of what might drive behavior and perceptions and attitudes in them at the mass level also is going to operate at the elite level.
So yes, I think it’s likely that both voters in the mass public and political elites are responding to broader forces, whether that’s sorting for whatever reason, like partisan sorting. And so, therefore they are these just issue and ideological differences that are motivating an affective response, whether it’s media as a broader force and context, whether it’s identity cleavages, if the parties are sorting along these other lines, then racial animus is going to be tied up in partisan animus. So both are probably responding to these other broader forces rather than just driving each other. But we also know that elites shape discourse. Elites set the tone, amplify current frames that operate and help drive our understanding of politics, and elites model behavior of what’s the norm.
So while not probably the sole driving cause of public affective polarization, I think elites absolutely reinforce and legitimize mass level affective polarization even if it’s not the root starting cause, which we’ll never know. We’ll never really know that. We have ideas. It’s probably all of these things, it’s going to be a messy story of why people have polarized and that’s a messy story. So I think yes, I agree. I don’t think that just because I find that mismatch between politicians using rhetoric and the public doesn’t want it, that totally means that elites are the driving cause. But I think it’s suggestive and I think that they do reinforce it, because one thing I find is that people think the other side wants it more. And so, I think that’s all part of this feedback loop that just reinforces itself.
Matt Grossmann: So that mirrors some findings in political behavior research on polarization as you know and even some of the, what can we do about it research that you reference which is that if some of this is about misperception, maybe just letting people know that the other side is not any more into this than you are will make a difference. You think that’s right? Do you think there would be any change in behavior from people knowing the relatively low frequency of this kind of behavior on each side?
Mia Costa: TBD. I’m like, I have some work in progress that is at least on the public side, like testing whether this informational corrective would help change levels of partisan animosity in general or just change levels of how much people tolerate negative representation from elites. So I’m working on it. It’s something I’m interested in. I’m not super optimistic. I think, yeah, probably not a lot in terms of would it change things. Because the other incentives remain for that behavior from one of the side of the elites with the media attention, the base, approval from the base, fundraising, a lot of politicians don’t prioritize general election success if they’re in safe seats.
So on the elite side, learning that people don’t respond well to that, even though I find that they already think that, that changes unlikely on their side, because changing the norms of politicians like rhetoric and changing the playbook, which we know now is a little bit more common. It’s changing the norm is hard when personal rewards outweigh the risk like [inaudible 00:45:37] electoral risk.
On the public side, it’s like if people know that, “Oh, we’re all in this together. And even the Republicans don’t want their politicians to attack Democrats, and even Democrats don’t want their… This is not our fault. It’s not.” So would that help mass level affective polarization? Maybe, but I think it’s hard to do. I think it is possible in a short-term, give people that one shot bit of corrective information that that might move the needle a little bit. But whether that would have any long-standing or durable change, I think that’s hard because I do think the partisan animus is driven by genuine ideological disagreement. And so, even though they might think, oh, people don’t want negative representation, but that, I don’t know if that’s what’s driving affective polarization. If people still have genuine disagreements with the other side, then they’re still not going to like the other side.
Matt Grossmann: What about on the elite behavior side? You have been emphasizing that people have incentives to do this, but those incentives are cable news bookings and money and social media retweets. It seems like you do have a case that you could make here to politicians that they aren’t actually responding to what voters want. They’re not necessarily helping themselves legislatively or electorally. Those are the kinds of things that do sometimes get through to politicians to change behavior. But let’s say we got through the informational challenge of actually having them read the book or get the summary. Do you think that they already know that there’s this basic trade off or do you think that they are under the misperception that this is a winning strategy across the board?
Mia Costa: Yeah, well, [inaudible 00:48:01]-
Mia Costa: So I ran these experiments to see their reactions to this polarizing, partisan rhetoric and the effect of that rhetoric on elites perceptions of electoral viability and success, and it doesn’t seem like they’re under the impression that it’s a winning strategy. They’re really likely to report, at least. Of course, there’s a limitation of the data that I can only have what they’re willing to tell me, which is that they’re at least reporting that they don’t think it wins votes.
I think the problem is that, okay, voters don’t respond well to this rhetoric, but they don’t put their money where their mouth is because at the end of the day, they are more important things to base your vote on. Voters aren’t voting people out because they’re being mean to the other side. They might vote people out if they don’t agree with them on the issues, and especially in a general election. Again, you’re not going to party switch, and I think politicians know that, right? It’s like they know that that trade-off is one that is, that’s a worthwhile risk because you’re not going to lose enough vote, maybe you’re going to get more attention from donors and from the news media, but you’re not really going to lose your base, even though your base, I find that the primary voters, the strongest partisan supporters, still do not respond favorably to this type of rhetoric, but they don’t penalize it either. And I call it sort of like a selective tolerance, right? It’s like, “Okay, I will tolerate this behavior as long as you agree with me otherwise.”
And so I think there are incentives for it, and I think politicians know that. And so I don’t think on the elite side, it’s like a informational piece that needs to be done. I find that the voters in the public think that other people want this, and so therefore they tolerate it. So I think that’s where the intervention could move the needle. I think the public could be informed somehow that people don’t want this. People want to focus on the issues and that could move levels of effect, polarization or whatever.
But on the elite side, it’s like I find that elites actually are pretty accurate when it comes to their understanding of what voters want, but I think it’s just, again, the incentive structure outweighs it, and there’s not much electoral risk in deploying a strategy that voters sometimes don’t like. Because of that vector polarization, people aren’t going to vote for the other party, so you don’t have to do everything that voters want.
Matt Grossmann: And do we want them to? You have sort of seen this as a representation story. We want politicians to do what voters want or we want to at least see why they’re not doing that.
On the other hand, this is partly just transparency. It would be odd if you were spending all of your day fighting the other party and then you never talked about it. And in some ways, maybe what they’re hiding is they’re hiding the level of partisanship. They’re sending us these newsletters saying, “Everybody is working on veterans issues together,” and finding common ground, and then they’re spending all day actually voting against everything that the other party is putting forward and strategizing about how to win partisan initiatives. I don’t want to say that’s all of what they do, but it seems like if there’s a failure in transparency here is actually on the not as much partisan, negative partisan rhetoric as you would expect given their behavior.
What do you think? We have too little of this given that, or is this kind of a traditional failure of representation for what the voters want they’re not getting?
Mia Costa: Yeah, that’s such an interesting question because we don’t know what the right, or rather what the most accurate level of negative rhetoric or partisan rhetoric is to match what you’re suggesting is to match what politicians are actually doing. I don’t think there’s a way to, it’s not a one-to-one. If we think of the [inaudible 00:53:09] home style of part of your job in cultivating a home style as a congress member is explaining what you’re doing in Washington. So in that respect, yeah, maybe if all you’re doing in Washington is opposing the other side, then talking about it, focusing on the other party is actually transparently explaining what you’re doing in Washington.
What I would say is that that is still focused on the issues though. If what you’re doing in Washington is focused on opposing the other side in terms of their agenda, then that is a legislative agenda, and you can talk about it in that way. But I specifically measure distinctions between talking about the other party absent policy language. And so if you are just focused on, “Oh, the Republicans are bad for this country,” but you’re not really saying why, or if that’s based in a legislative agenda, then that is what I’m considering this negative partisan rhetoric that isn’t based in the substantive issues.
And so I think there would be a way to, if what you’re doing in Washington is blocking or oppositional or whatever it’s like, but you can still talk about that in ideological terms, in legislative terms, in substantive terms. More generally, it’s just a really interesting question to think, do we want politicians to do what they think the voters want or do we want them to just be more transparent in terms of what they’re doing or what they actually think? And in terms of, I mean, it’s just a large bigger question that is interesting to think about of how much should elites be, a little bit of the pandering should be pandering versus transparently testing what they’re doing.
Catering to voters, that is the mechanism we have for representation. It’s a blunt mechanism because elections are an imperfect way to mandate preferences. Something I always say to my students when I’m teaching about representation is like, why is it thought of as a bad thing when politicians cater to voters or are always campaigning or putting things in a certain light that voters would want? Don’t we want our politicians to pursue strategies that they would appeal to the most people, or don’t we want our politicians to appeal to people?
Of course, we don’t want them to be dishonest, but there’s also something to be said on its own for representatives caring what their voters think and wanting to pitch things in a way that the most number of voters would respond well to, especially when those strategies are probably otherwise good for representative democracy like in this case, focusing more on policy that more than substantive, less partisan othering.
So therefore, politicians in Congress are really just focused on being oppositional, but I still don’t know if we would think that it would be a good thing for them, for their rhetoric to then be only oppositional because that might have effects on democratic attitudes and polarization. And so we would still want them to talk about issues even if maybe they’re not performing well in Congress, but their rhetoric still matters on its own.
And so of course it’s more complicated than that. We don’t want, I don’t mean to suggest that politicians should just be catering to voters and being dishonest just to garner support, especially because elections, voting in general is not just a blunt mechanism for representation, but it’s a vastly unequal one, right? Some voters matter more than others than to people, to candidates and to representatives.
And so while pursuing broadly appealing strategies sounds democratic, of course, it depends on which voters are being centered. And so if it’s only primary voters and the base that are being centered, then we’re getting more polarized, more extreme politics than maybe what most people want. But what I’m trying to say is if what works quote unquote works electorally also leads to more policy focus and less partisan hostility, even just in rhetoric, then I think that’s a win for representative democracy.
Matt Grossmann: And anything you want to tout about what you’re working on now or anything we didn’t get to that you wanted to leave us with?
Mia Costa: So one thing I’m really interested in working on now, stuff we already touched on, which is how to shift those incentives for politicians specifically, especially if representatives and elites and candidates sort of know that voters don’t really want partisan attacks over a focus on substantive issues, then how do we shift their understanding of those incentives, if at all possible? Really just in general, what do elites make of this? I’m really interested in elite perceptions of representation, so continuing down that path.
And then a project that I’m also currently working on right now is on the mass public side, is the downstream effects of this behavior beyond just the attitude in all outcomes that I already have. So things I measure in the book already are voters initial responses to this type of rhetoric from politicians compared to policy alignment. I also have attitudinal measures of affective polarization, of course, and then what people think the other side wants and prefers, right? So do Democrats think that Republicans really prefer negative representation and vice versa?
But one thing I haven’t looked at yet is the behavioral outcomes, the downstream effects for the public. Is this sort of polarizing rhetoric firing people up? I find people don’t like it, but they tolerate it. But if you don’t like it, does that fire people up to be mobilized or does it burn them out to sort of sit it out? Because like we talked about, a lot of people just aren’t really into this sort of the partisan focus, right? They want to talk about the issues or they’re just disillusioned in general from polarization and from partisan conflict. And so one thing I’m working on now is collecting data on measures of civic engagement and participation in response to polarizing rhetoric from elites. So that’s an interesting direction that I’m working on now.
Matt Grossmann: There’s a lot more to learn. The Science of Politics is available bi-weekly from the Niskanen Center, and I’m your host, Matt Grossmann. If you like this discussion, here are the episodes you should check out next, all linked on our website, How Media Coverage of Congress Limits Policymaking, How Voters Judge Congress, How Party Leaders Change Congress, How Congress Communicates, and the Deterioration of Congress. Thanks to Mia Costa for joining me. Please check out How Politicians Polarize and then listen in next time.